From: root Date: Tue, 31 Jan 2006 20:50:48 +0000 (+0000) Subject: *** empty log message *** X-Git-Url: http://git.openbox.org/?a=commitdiff_plain;h=454e5b425706c4085717a1fb3584fbd1911c8c39;p=dana%2Furxvt.git *** empty log message *** --- diff --git a/Changes b/Changes index 012fd9dc..deb963f2 100644 --- a/Changes +++ b/Changes @@ -15,8 +15,6 @@ DUMB: support tex fonts THINK: struct passwd *userinfo; /* defined in pwd.h + if ((userinfo = getpwuid(getuid())) != NULL) + shell = strdup(userinfo->pw_shell); /* retrieve current shell */ TODO: xim "how to" faq entry -TODO: reorganize faq -TODO: scrollstyle -next, raised/sunken, maybe extra pixels(?) w. halfshadow - changed interpretation of [alpha] colour prefix. - +option now really sets the option to default, instead of using the @@ -25,6 +23,7 @@ TODO: scrollstyle -next, raised/sunken, maybe extra pixels(?) w. halfshadow the colour components. - plain scrollbar works better with -sr. - fixed half-shadow scrollbar look. + - reorganised the FAQ into multiple sections. 7.5 Tue Jan 31 15:15:43 CET 2006 - further improvements to the careful mode detection, and font width diff --git a/README.FAQ b/README.FAQ index ee7da8f5..2845c4eb 100644 --- a/README.FAQ +++ b/README.FAQ @@ -1,122 +1,85 @@ FREQUENTLY ASKED QUESTIONS - The new selection selects pieces that are too big, how can I select single words? - If you want to select e.g. alphanumeric words, you can use the following - setting: - - URxvt.selection.pattern-0: ([[:word:]]+) - - If you click more than twice, the selection will be extended more and - more. - - To get a selection that is very similar to the old code, try this - pattern: - - URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+) - - Please also note that the *LeftClick Shift-LeftClik* combination also - selects words like the old code. - - I don't like the new selection/popups/hotkeys/perl, how do I change/disable it? - You can disable the perl extension completely by setting the - perl-ext-common resource to the empty string, which also keeps - rxvt-unicode from initialising perl, saving memory. - - If you only want to disable specific features, you first have to - identify which perl extension is responsible. For this, read the section - PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For example, to - disable the selection-popup and option-popup, specify this - perl-ext-common resource: - - URxvt.perl-ext-common: default,-selection-popup,-option-popup - - This will keep the default extensions, but disable the two popup - extensions. Some extensions can also be configured, for example, - scrollback search mode is triggered by M-s. You can move it to any other - combination either by setting the searchable-scrollback resource: - - URxvt.searchable-scrollback: CM-s - - The cursor moves when selecting text in the current input line, how do I switch this off? - See next entry. - - During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this? - These are caused by the "readline" perl extension. Under normal - circumstances, it will move your cursor around when you click into the - line that contains it. It tries hard not to do this at the wrong moment, - but when running a program that doesn't parse cursor movements or in - some cases during rlogin sessions, it fails to detect this properly. - - You can permamently switch this feature off by disabling the "readline" - extension: - - URxvt.perl-ext-common: default,-readline - - Why doesn't rxvt-unicode read my resources? - Well, why, indeed? It does, in a way very similar to other X - applications. Most importantly, this means that if you or your OS loads - resources into the X display (the right way to do it), rxvt-unicode will - ignore any resource files in your home directory. It will only read - $HOME/.Xdefaults when no resources are attached to the display. - - If you have or use an $HOME/.Xresources file, chances are that resources - are loaded into your X-server. In this case, you have to re-login after - every change (or run xrdb -merge $HOME/.Xresources). - - Also consider the form resources have to use: - - URxvt.resource: value - - If you want to use another form (there are lots of different ways of - specifying resources), make sure you understand wether and why it works. - If unsure, use the form above. - - I can't get transparency working, what am I doing wrong? - First of all, transparency isn't officially supported in rxvt-unicode, - so you are mostly on your own. Do not bug the author about it (but you - may bug everybody else). Also, if you can't get it working consider it a - rite of passage: ... and you failed. + Meta, Features & Commandline Issues + My question isn't answered here, can I ask a human? + Before sending me mail, you could go to IRC: "irc.freenode.net", channel + "#rxvt-unicode" has some rxvt-unicode enthusiasts that might be + interested in learning about new and exciting problems (but not FAQs :). - Here are four ways to get transparency. Do read the manpage and option - descriptions for the programs mentioned and rxvt-unicode. Really, do it! + Does it support tabs, can I have a tabbed rxvt-unicode? + Beginning with version 7.3, there is a perl extension that implements a + simple tabbed terminal. It is installed by default, so any of these + should give you tabs: - 1. Use inheritPixmap: + rxvt -pe tabbed - Esetroot wallpaper.jpg - rxvt -ip -tint red -sh 40 + URxvt.perl-ext-common: default,tabbed - That works. If you think it doesn't, you lack transparency and tinting - support, or you are unable to read. + It will also work fine with tabbing functionality of many window + managers or similar tabbing programs, and its embedding-features allow + it to be embedded into other programs, as witnessed by doc/rxvt-tabbed + or the upcoming "Gtk2::URxvt" perl module, which features a tabbed urxvt + (murxvt) terminal as an example embedding application. - 2. Use a simple pixmap and emulate pseudo-transparency. This enables you - to use effects other than tinting and shading: Just shade/tint/whatever - your picture with gimp: + How do I know which rxvt-unicode version I'm using? + The version number is displayed with the usage (-h). Also the escape + sequence "ESC [ 8 n" sets the window title to the version number. When + using the rxvtc client, the version displayed is that of the daemon. - convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm - rxvt -pixmap background.xpm -pe automove-background + Rxvt-unicode uses gobs of memory, how can I reduce that? + Rxvt-unicode tries to obey the rule of not charging you for something + you don't use. One thing you should try is to configure out all settings + that you don't need, for example, Xft support is a resource hog by + design, when used. Compiling it out ensures that no Xft font will be + loaded accidentally when rxvt-unicode tries to find a font for your + characters. - That works. If you think it doesn't, you lack XPM and Perl support, or - you are unable to read. + Also, many people (me included) like large windows and even larger + scrollback buffers: Without "--enable-unicode3", rxvt-unicode will use 6 + bytes per screen cell. For a 160x?? window this amounts to almost a + kilobyte per line. A scrollback buffer of 10000 lines will then (if + full) use 10 Megabytes of memory. With "--enable-unicode3" it gets + worse, as rxvt-unicode then uses 8 bytes per screen cell. - 3. Use an ARGB visual: + How can I start rxvtd in a race-free way? + Try "rxvtd -f -o", which tells rxvtd to open the display, create the + listening socket and then fork. - rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc + How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc. + rxvt and rxvt-unicode always export the variable "COLORTERM", so you can + check and see if that is set. Note that several programs, JED, slrn, + Midnight Commander automatically check this variable to decide whether + or not to use color. - This requires XFT support, and the support of your X-server. If that - doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't - there yet, no matter what they claim. Rxvt-Unicode contains the - neccessary bugfixes and workarounds for Xft and Xlib to make it work, - but that doesn't mean that your WM has the required kludges in place. + How do I set the correct, full IP address for the DISPLAY variable? + If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled + insecure mode then it is possible to use the following shell script + snippets to correctly set the display. If your version of rxvt-unicode + wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) + then the COLORTERM variable can be used to distinguish rxvt-unicode from + a regular xterm. - 4. Use xcompmgr and let it do the job: + Courtesy of Chuck Blake with the following shell script + snippets: - xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ - -set _NET_WM_WINDOW_OPACITY 0xc0000000 + # Bourne/Korn/POSIX family of shells: + [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know + if [ ${TERM:-foo} = xterm ]; then + stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not + echo -n '^[Z' + read term_id + stty icanon echo + if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then + echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string + read DISPLAY # set it in our local shell + fi + fi - Then click on a window you want to make transparent. Replace 0xc0000000 - by other values to change the degree of opacity. If it doesn't work and - your server crashes, you got to keep the pieces. + How do I compile the manual pages on my own? + You need to have a recent version of perl installed as /usr/bin/perl, + one that comes with pod2man, pod2text and pod2html. Then go to the doc + subdirectory and enter "make alldoc". - Isn't rxvt supposed to be small? Don't all those features bloat? + Isn't rxvt-unicode supposed to be small? Don't all those features bloat? I often get asked about this, and I think, no, they didn't cause extra bloat. If you compare a minimal rxvt and a minimal urxvt, you can see that the urxvt binary is larger (due to some encoding tables always @@ -157,7 +120,7 @@ FREQUENTLY ASKED QUESTIONS a minute of startup time, including the hundreds of warnings it spits out), it fares extremely well *g*. - Why C++, isn't that unportable/bloated/uncool? + Why C++, isn't that unportable/bloated/uncool? Is this a question? :) It comes up very often. The simple answer is: I had to write it, and C++ allowed me to write and maintain it in a fraction of the time and effort (which is a scarce resource for me). Put @@ -190,222 +153,221 @@ FREQUENTLY ASKED QUESTIONS No large bloated libraries (of course, none were linked in statically), except maybe libX11 :) - Does it support tabs, can I have a tabbed rxvt-unicode? - Beginning with version 7.3, there is a perl extension that implements a - simple tabbed terminal. It is installed by default, so any of these - should give you tabs: + Rendering, Font & Look and Feel Issues + I can't get transparency working, what am I doing wrong? + First of all, transparency isn't officially supported in rxvt-unicode, + so you are mostly on your own. Do not bug the author about it (but you + may bug everybody else). Also, if you can't get it working consider it a + rite of passage: ... and you failed. - rxvt -pe tabbed + Here are four ways to get transparency. Do read the manpage and option + descriptions for the programs mentioned and rxvt-unicode. Really, do it! - URxvt.perl-ext-common: default,tabbed + 1. Use inheritPixmap: - It will also work fine with tabbing functionality of many window - managers or similar tabbing programs, and its embedding-features allow - it to be embedded into other programs, as witnessed by doc/rxvt-tabbed - or the upcoming "Gtk2::URxvt" perl module, which features a tabbed urxvt - (murxvt) terminal as an example embedding application. + Esetroot wallpaper.jpg + rxvt -ip -tint red -sh 40 - How do I know which rxvt-unicode version I'm using? - The version number is displayed with the usage (-h). Also the escape - sequence "ESC [ 8 n" sets the window title to the version number. When - using the rxvtc client, the version displayed is that of the daemon. + That works. If you think it doesn't, you lack transparency and tinting + support, or you are unable to read. - I am using Debian GNU/Linux and have a problem... - The Debian GNU/Linux package of rxvt-unicode in sarge contains large - patches that considerably change the behaviour of rxvt-unicode (but - unfortunately this notice has been removed). Before reporting a bug to - the original rxvt-unicode author please download and install the genuine - version () and try to reproduce - the problem. If you cannot, chances are that the problems are specific - to Debian GNU/Linux, in which case it should be reported via the Debian - Bug Tracking System (use "reportbug" to report the bug). + 2. Use a simple pixmap and emulate pseudo-transparency. This enables you + to use effects other than tinting and shading: Just shade/tint/whatever + your picture with gimp or any other tool: - For other problems that also affect the Debian package, you can and - probably should use the Debian BTS, too, because, after all, it's also a - bug in the Debian version and it serves as a reminder for other users - that might encounter the same issue. + convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm + rxvt -pixmap background.xpm -pe automove-background - I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation? - You should build one binary with the default options. configure now - enables most useful options, and the trend goes to making them - runtime-switchable, too, so there is usually no drawback to enbaling - them, except higher disk and possibly memory usage. The perl interpreter - should be enabled, as important functionality (menus, selection, likely - more in the future) depends on it. + That works. If you think it doesn't, you lack XPM and Perl support, or + you are unable to read. - You should not overwrite the "perl-ext-common" snd "perl-ext" resources - system-wide (except maybe with "defaults"). This will result in useful - behaviour. If your distribution aims at low memory, add an empty - "perl-ext-common" resource to the app-defaults file. This will keep the - perl interpreter disabled until the user enables it. + 3. Use an ARGB visual: - If you can/want build more binaries, I recommend building a minimal one - with "--disable-everything" (very useful) and a maximal one with - "--enable-everything" (less useful, it will be very big due to a lot of - encodings built-in that increase download times and are rarely used). + rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc - I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe? - It should be, starting with release 7.1. You are encouraged to properly - install urxvt with privileges necessary for your OS now. + This requires XFT support, and the support of your X-server. If that + doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't + there yet, no matter what they claim. Rxvt-Unicode contains the + neccessary bugfixes and workarounds for Xft and Xlib to make it work, + but that doesn't mean that your WM has the required kludges in place. - When rxvt-unicode detects that it runs setuid or setgid, it will fork - into a helper process for privileged operations (pty handling on some - systems, utmp/wtmp/lastlog handling on others) and drop privileges - immediately. This is much safer than most other terminals that keep - privileges while running (but is more relevant to urxvt, as it contains - things as perl interpreters, which might be "helpful" to attackers). + 4. Use xcompmgr and let it do the job: - This forking is done as the very first within main(), which is very - early and reduces possible bugs to initialisation code run before - main(), or things like the dynamic loader of your system, which should - result in very little risk. + xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ + -set _NET_WM_WINDOW_OPACITY 0xc0000000 - When I log-in to another system it tells me about missing terminfo data? - The terminal description used by rxvt-unicode is not as widely available - as that for xterm, or even rxvt (for which the same problem often - arises). + Then click on a window you want to make transparent. Replace 0xc0000000 + by other values to change the degree of opacity. If it doesn't work and + your server crashes, you got to keep the pieces. - The correct solution for this problem is to install the terminfo, this - can be done like this (with ncurses' infocmp): + Why do some chinese characters look so different than others? + This is because there is a difference between script and language -- + rxvt-unicode does not know which language the text that is output is, as + it only knows the unicode character codes. If rxvt-unicode first sees a + japanese/chinese character, it might choose a japanese font for display. + Subsequent japanese characters will use that font. Now, many chinese + characters aren't represented in japanese fonts, so when the first + non-japanese character comes up, rxvt-unicode will look for a chinese + font -- unfortunately at this point, it will still use the japanese font + for chinese characters that are also in the japanese font. - REMOTE=remotesystem.domain - infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" + The workaround is easy: just tag a chinese font at the end of your font + list (see the previous question). The key is to view the font list as a + preference list: If you expect more japanese, list a japanese font + first. If you expect more chinese, put a chinese font first. - ... or by installing rxvt-unicode normally on the remote system, + In the future it might be possible to switch language preferences at + runtime (the internal data structure has no problem with using different + fonts for the same character at the same time, but no interface for this + has been designed yet). - If you cannot or do not want to do this, then you can simply set - "TERM=rxvt" or even "TERM=xterm", and live with the small number of - problems arising, which includes wrong keymapping, less and different - colours and some refresh errors in fullscreen applications. It's a nice - quick-and-dirty workaround for rare cases, though. + Until then, you might get away with switching fonts at runtime (see "Can + I switch the fonts at runtime?" later in this document). - If you always want to do this (and are fine with the consequences) you - can either recompile rxvt-unicode with the desired TERM value or use a - resource to set it: + Why does rxvt-unicode sometimes leave pixel droppings? + Most fonts were not designed for terminal use, which means that + character size varies a lot. A font that is otherwise fine for terminal + use might contain some characters that are simply too wide. Rxvt-unicode + will avoid these characters. For characters that are just "a bit" too + wide a special "careful" rendering mode is used that redraws adjacent + characters. - URxvt.termName: rxvt + All of this requires that fonts do not lie about character sizes, + however: Xft fonts often draw glyphs larger than their acclaimed + bounding box, and rxvt-unicode has no way of detecting this (the correct + way is to ask for the character bounding box, which unfortunately is + wrong in these cases). - If you don't plan to use rxvt (quite common...) you could also replace - the rxvt terminfo file with the rxvt-unicode one. + It's not clear (to me at least), wether this is a bug in Xft, freetype, + or the respective font. If you encounter this problem you might try + using the "-lsp" option to give the font more height. If that doesn't + work, you might be forced to use a different font. - "tic" outputs some error when compiling the terminfo entry. - Most likely it's the empty definition for "enacs=". Just replace it by - "enacs=\E[0@" and try again. + All of this is not a problem when using X11 core fonts, as their + bounding box data is correct. - "bash"'s readline does not work correctly under rxvt. - See next entry. + How can I keep rxvt-unicode from using reverse video so much? + First of all, make sure you are running with the right terminal settings + ("TERM=rxvt-unicode"), which will get rid of most of these effects. Then + make sure you have specified colours for italic and bold, as otherwise + rxvt-unicode might use reverse video to simulate the effect: - I need a termcap file entry. - One reason you might want this is that some distributions or operating - systems still compile some programs using the long-obsoleted termcap - library (Fedora Core's bash is one example) and rely on a termcap entry - for "rxvt-unicode". + URxvt.colorBD: white + URxvt.colorIT: green - You could use rxvt's termcap entry with resonable results in many cases. - You can also create a termcap entry by using terminfo's infocmp program - like this: + Some programs assume totally weird colours (red instead of blue), how can I fix that? + For some unexplainable reason, some rare programs assume a very weird + colour palette when confronted with a terminal with more than the + standard 8 colours (rxvt-unicode supports 88). The right fix is, of + course, to fix these programs not to assume non-ISO colours without very + good reasons. - infocmp -C rxvt-unicode + In the meantime, you can either edit your "rxvt-unicode" terminfo + definition to only claim 8 colour support or use "TERM=rxvt", which will + fix colours but keep you from using other rxvt-unicode features. - Or you could use this termcap entry, generated by the command above: + Can I switch the fonts at runtime? + Yes, using an escape sequence. Try something like this, which has the + same effect as using the "-fn" switch, and takes effect immediately: - rxvt-unicode|rxvt-unicode terminal (X Window System):\ - :am:bw:eo:km:mi:ms:xn:xo:\ - :co#80:it#8:li#24:lm#0:\ - :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ - :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ - :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ - :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ - :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ - :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ - :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ - :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ - :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ - :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ - :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ - :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ - :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ - :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ - :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ - :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ - :vs=\E[?25h: + printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" - Why does "ls" no longer have coloured output? - The "ls" in the GNU coreutils unfortunately doesn't use terminfo to - decide wether a terminal has colour, but uses it's own configuration - file. Needless to say, "rxvt-unicode" is not in it's default file (among - with most other terminals supporting colour). Either add: + This is useful if you e.g. work primarily with japanese (and prefer a + japanese font), but you have to switch to chinese temporarily, where + japanese fonts would only be in your way. - TERM rxvt-unicode + You can think of this as a kind of manual ISO-2022 switching. - to "/etc/DIR_COLORS" or simply add: + Why do italic characters look as if clipped? + Many fonts have difficulties with italic characters and hinting. For + example, the otherwise very nicely hinted font "xft:Bitstream Vera Sans + Mono" completely fails in it's italic face. A workaround might be to + enable freetype autohinting, i.e. like this: - alias ls='ls --color=auto' + URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true + URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true - to your ".profile" or ".bashrc". + Can I speed up Xft rendering somehow? + Yes, the most obvious way to speed it up is to avoid Xft entirely, as it + is simply slow. If you still want Xft fonts you might try to disable + antialiasing (by appending ":antialias=false"), which saves lots of + memory and also speeds up rendering considerably. - Why doesn't vim/emacs etc. use the 88 colour mode? - See next entry. + Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? + Rxvt-unicode will use whatever you specify as a font. If it needs to + fall back to it's default font search list it will prefer X11 core + fonts, because they are small and fast, and then use Xft fonts. It has + antialiasing disabled for most of them, because the author thinks they + look best that way. - Why doesn't vim/emacs etc. make use of italic? - See next entry. + If you want antialiasing, you have to specify the fonts manually. - Why are the secondary screen-related options not working properly? - Make sure you are using "TERM=rxvt-unicode". Some pre-packaged - distributions (most notably Debian GNU/Linux) break rxvt-unicode by - setting "TERM" to "rxvt", which doesn't have these extra features. - Unfortunately, some of these (most notably, again, Debian GNU/Linux) - furthermore fail to even install the "rxvt-unicode" terminfo file, so - you will need to install it on your own (See the question When I log-in - to another system it tells me about missing terminfo data? on how to do - this). + What's with this bold/blink stuff? + If no bold colour is set via "colorBD:", bold will invert text using the + standard foreground colour. - My numerical keypad acts weird and generates differing output? - Some Debian GNUL/Linux users seem to have this problem, although no - specific details were reported so far. It is possible that this is - caused by the wrong "TERM" setting, although the details of wether and - how this can happen are unknown, as "TERM=rxvt" should offer a - compatible keymap. See the answer to the previous question, and please - report if that helped. + For the standard background colour, blinking will actually make the text + blink when compiled with "--enable-blinking". with standard colours. + Without "--enable-blinking", the blink attribute will be ignored. - Rxvt-unicode does not seem to understand the selected encoding? - See next entry. + On ANSI colours, bold/blink attributes are used to set high-intensity + foreground/background colors. - Unicode does not seem to work? - If you encounter strange problems like typing an accented character but - getting two unrelated other characters or similar, or if program output - is subtly garbled, then you should check your locale settings. + color0-7 are the low-intensity colors. - Rxvt-unicode must be started with the same "LC_CTYPE" setting as the - programs. Often rxvt-unicode is started in the "C" locale, while the - login script running within the rxvt-unicode window changes the locale - to something else, e.g. "en_GB.UTF-8". Needless to say, this is not - going to work. + color8-15 are the corresponding high-intensity colors. - The best thing is to fix your startup environment, as you will likely - run into other problems. If nothing works you can try this in your - .profile. + I don't like the screen colors. How do I change them? + You can change the screen colors at run-time using ~/.Xdefaults + resources (or as long-options). - printf '\e]701;%s\007' "$LC_CTYPE" + Here are values that are supposed to resemble a VGA screen, including + the murky brown that passes for low-intensity yellow: - If this doesn't work, then maybe you use a "LC_CTYPE" specification not - supported on your systems. Some systems have a "locale" command which - displays this (also, "perl -e0" can be used to check locale settings, as - it will complain loudly if it cannot set the locale). If it displays - something like: + URxvt.color0: #000000 + URxvt.color1: #A80000 + URxvt.color2: #00A800 + URxvt.color3: #A8A800 + URxvt.color4: #0000A8 + URxvt.color5: #A800A8 + URxvt.color6: #00A8A8 + URxvt.color7: #A8A8A8 - locale: Cannot set LC_CTYPE to default locale: ... + URxvt.color8: #000054 + URxvt.color9: #FF0054 + URxvt.color10: #00FF54 + URxvt.color11: #FFFF54 + URxvt.color12: #0000FF + URxvt.color13: #FF00FF + URxvt.color14: #00FFFF + URxvt.color15: #FFFFFF - Then the locale you specified is not supported on your system. + And here is a more complete set of non-standard colors described (not by + me) as "pretty girly". - If nothing works and you are sure that everything is set correctly then - you will need to remember a little known fact: Some programs just don't - support locales :( + URxvt.cursorColor: #dc74d1 + URxvt.pointerColor: #dc74d1 + URxvt.background: #0e0e0e + URxvt.foreground: #4ad5e1 + URxvt.color0: #000000 + URxvt.color8: #8b8f93 + URxvt.color1: #dc74d1 + URxvt.color9: #dc74d1 + URxvt.color2: #0eb8c7 + URxvt.color10: #0eb8c7 + URxvt.color3: #dfe37e + URxvt.color11: #dfe37e + URxvt.color5: #9e88f0 + URxvt.color13: #9e88f0 + URxvt.color6: #73f7ff + URxvt.color14: #73f7ff + URxvt.color7: #e1dddd + URxvt.color15: #e1dddd - Why do some characters look so much different than others? + Why do some characters look so much different than others? See next entry. - How does rxvt-unicode choose fonts? + How does rxvt-unicode choose fonts? Most fonts do not contain the full range of Unicode, which is fine. Chances are that the font you (or the admin/package maintainer of your system/os) have specified does not cover all the characters you want to @@ -422,7 +384,7 @@ FREQUENTLY ASKED QUESTIONS e.g.: rxvt -fn basefont,font2,font3... - + When rxvt-unicode sees a character, it will first look at the base font. If the base font does not contain the character, it will go to the next font, and so on. Specifying your own fonts will also speed up this @@ -432,60 +394,68 @@ FREQUENTLY ASKED QUESTIONS base font, as the base font defines the terminal character cell size, which must be the same due to the way terminals work. - Why do some chinese characters look so different than others? - This is because there is a difference between script and language -- - rxvt-unicode does not know which language the text that is output is, as - it only knows the unicode character codes. If rxvt-unicode first sees a - japanese/chinese character, it might choose a japanese font for display. - Subsequent japanese characters will use that font. Now, many chinese - characters aren't represented in japanese fonts, so when the first - non-japanese character comes up, rxvt-unicode will look for a chinese - font -- unfortunately at this point, it will still use the japanese font - for chinese characters that are also in the japanese font. + Keyboard, Mouse & User Interaction + The new selection selects pieces that are too big, how can I select single words? + If you want to select e.g. alphanumeric words, you can use the following + setting: - The workaround is easy: just tag a chinese font at the end of your font - list (see the previous question). The key is to view the font list as a - preference list: If you expect more japanese, list a japanese font - first. If you expect more chinese, put a chinese font first. + URxvt.selection.pattern-0: ([[:word:]]+) - In the future it might be possible to switch language preferences at - runtime (the internal data structure has no problem with using different - fonts for the same character at the same time, but no interface for this - has been designed yet). + If you click more than twice, the selection will be extended more and + more. - Until then, you might get away with switching fonts at runtime (see "Can - I switch the fonts at runtime?" later in this document). + To get a selection that is very similar to the old code, try this + pattern: - Why does rxvt-unicode sometimes leave pixel droppings? - Most fonts were not designed for terminal use, which means that - character size varies a lot. A font that is otherwise fine for terminal - use might contain some characters that are simply too wide. Rxvt-unicode - will avoid these characters. For characters that are just "a bit" too - wide a special "careful" rendering mode is used that redraws adjacent - characters. + URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+) - All of this requires that fonts do not lie about character sizes, - however: Xft fonts often draw glyphs larger than their acclaimed - bounding box, and rxvt-unicode has no way of detecting this (the correct - way is to ask for the character bounding box, which unfortunately is - wrong in these cases). + Please also note that the *LeftClick Shift-LeftClik* combination also + selects words like the old code. - It's not clear (to me at least), wether this is a bug in Xft, freetype, - or the respective font. If you encounter this problem you might try - using the "-lsp" option to give the font more height. If that doesn't - work, you might be forced to use a different font. + I don't like the new selection/popups/hotkeys/perl, how do I change/disable it? + You can disable the perl extension completely by setting the + perl-ext-common resource to the empty string, which also keeps + rxvt-unicode from initialising perl, saving memory. - All of this is not a problem when using X11 core fonts, as their - bounding box data is correct. + If you only want to disable specific features, you first have to + identify which perl extension is responsible. For this, read the section + PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For example, to + disable the selection-popup and option-popup, specify this + perl-ext-common resource: - On Solaris 9, many line-drawing characters are too wide. - Seems to be a known bug, read - . Some people use the - following ugly workaround to get non-double-wide-characters working: + URxvt.perl-ext-common: default,-selection-popup,-option-popup - #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) + This will keep the default extensions, but disable the two popup + extensions. Some extensions can also be configured, for example, + scrollback search mode is triggered by M-s. You can move it to any other + combination either by setting the searchable-scrollback resource: + + URxvt.searchable-scrollback: CM-s + + The cursor moves when selecting text in the current input line, how do I switch this off? + See next entry. + + During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this? + These are caused by the "readline" perl extension. Under normal + circumstances, it will move your cursor around when you click into the + line that contains it. It tries hard not to do this at the wrong moment, + but when running a program that doesn't parse cursor movements or in + some cases during rlogin sessions, it fails to detect this properly. + + You can permamently switch this feature off by disabling the "readline" + extension: + + URxvt.perl-ext-common: default,-readline + + My numerical keypad acts weird and generates differing output? + Some Debian GNUL/Linux users seem to have this problem, although no + specific details were reported so far. It is possible that this is + caused by the wrong "TERM" setting, although the details of wether and + how this can happen are unknown, as "TERM=rxvt" should offer a + compatible keymap. See the answer to the previous question, and please + report if that helped. - My Compose (Multi_key) key is no longer working. + My Compose (Multi_key) key is no longer working. The most common causes for this are that either your locale is not set correctly, or you specified a preeditStyle that is not supported by your input method. For example, if you specified OverTheSpot and your input @@ -496,416 +466,452 @@ FREQUENTLY ASKED QUESTIONS In this case either do not specify a preeditStyle or specify more than one pre-edit style, such as OverTheSpot,Root,None. - I cannot type "Ctrl-Shift-2" to get an ASCII NUL character due to ISO 14755 + I cannot type "Ctrl-Shift-2" to get an ASCII NUL character due to ISO 14755 Either try "Ctrl-2" alone (it often is mapped to ASCII NUL even on international keyboards) or simply use ISO 14755 support to your advantage, typing to get a ASCII NUL. This works for other codes, too, such as "Ctrl-Shift-1-d" to type the default telnet escape character and so on. - How can I keep rxvt-unicode from using reverse video so much? - First of all, make sure you are running with the right terminal settings - ("TERM=rxvt-unicode"), which will get rid of most of these effects. Then - make sure you have specified colours for italic and bold, as otherwise - rxvt-unicode might use reverse video to simulate the effect: + Mouse cut/paste suddenly no longer works. + Make sure that mouse reporting is actually turned off since killing some + editors prematurely may leave the mouse in mouse report mode. I've heard + that tcsh may use mouse reporting unless it otherwise specified. A quick + check is to see if cut/paste works when the Alt or Shift keys are + depressed. - URxvt.colorBD: white - URxvt.colorIT: green + What's with the strange Backspace/Delete key behaviour? + Assuming that the physical Backspace key corresponds to the BackSpace + keysym (not likely for Linux ... see the following question) there are + two standard values that can be used for Backspace: "^H" and "^?". - Some programs assume totally weird colours (red instead of blue), how can I fix that? - For some unexplainable reason, some rare programs assume a very weird - colour palette when confronted with a terminal with more than the - standard 8 colours (rxvt-unicode supports 88). The right fix is, of - course, to fix these programs not to assume non-ISO colours without very - good reasons. + Historically, either value is correct, but rxvt-unicode adopts the + debian policy of using "^?" when unsure, because it's the one only only + correct choice :). - In the meantime, you can either edit your "rxvt-unicode" terminfo - definition to only claim 8 colour support or use "TERM=rxvt", which will - fix colours but keep you from using other rxvt-unicode features. + Rxvt-unicode tries to inherit the current stty settings and uses the + value of `erase' to guess the value for backspace. If rxvt-unicode + wasn't started from a terminal (say, from a menu or by remote shell), + then the system value of `erase', which corresponds to CERASE in + , will be used (which may not be the same as your stty + setting). - I am on FreeBSD and rxvt-unicode does not seem to work at all. - Rxvt-unicode requires the symbol "__STDC_ISO_10646__" to be defined in - your compile environment, or an implementation that implements it, - wether it defines the symbol or not. "__STDC_ISO_10646__" requires that - wchar_t is represented as unicode. + For starting a new rxvt-unicode: - As you might have guessed, FreeBSD does neither define this symobl nor - does it support it. Instead, it uses it's own internal representation of - wchar_t. This is, of course, completely fine with respect to standards. + # use Backspace = ^H + $ stty erase ^H + $ rxvt - However, that means rxvt-unicode only works in "POSIX", "ISO-8859-1" and - "UTF-8" locales under FreeBSD (which all use Unicode as wchar_t. + # use Backspace = ^? + $ stty erase ^? + $ rxvt - "__STDC_ISO_10646__" is the only sane way to support multi-language apps - in an OS, as using a locale-dependent (and non-standardized) - representation of wchar_t makes it impossible to convert between wchar_t - (as used by X11 and your applications) and any other encoding without - implementing OS-specific-wrappers for each and every locale. There - simply are no APIs to convert wchar_t into anything except the current - locale encoding. + Toggle with "ESC [ 36 h" / "ESC [ 36 l". - Some applications (such as the formidable mlterm) work around this by - carrying their own replacement functions for character set handling with - them, and either implementing OS-dependent hacks or doing multiple - conversions (which is slow and unreliable in case the OS implements - encodings slightly different than the terminal emulator). + For an existing rxvt-unicode: - The rxvt-unicode author insists that the right way to fix this is in the - system libraries once and for all, instead of forcing every app to carry - complete replacements for them :) + # use Backspace = ^H + $ stty erase ^H + $ echo -n "^[[36h" - I use Solaris 9 and it doesn't compile/work/etc. - Try the diff in doc/solaris9.patch as a base. It fixes the worst - problems with "wcwidth" and a compile problem. + # use Backspace = ^? + $ stty erase ^? + $ echo -n "^[[36l" - How can I use rxvt-unicode under cygwin? - rxvt-unicode should compile and run out of the box on cygwin, using the - X11 libraries that come with cygwin. libW11 emulation is no longer - supported (and makes no sense, either, as it only supported a single - font). I recommend starting the X-server in "-multiwindow" or - "-rootless" mode instead, which will result in similar look&feel as the - old libW11 emulation. + This helps satisfy some of the Backspace discrepancies that occur, but + if you use Backspace = "^H", make sure that the termcap/terminfo value + properly reflects that. - At the time of this writing, cygwin didn't seem to support any - multi-byte encodings (you might try "LC_CTYPE=C-UTF-8"), so you are - likely limited to 8-bit encodings. + The Delete key is a another casualty of the ill-defined Backspace + problem. To avoid confusion between the Backspace and Delete keys, the + Delete key has been assigned an escape sequence to match the vt100 for + Execute ("ESC [ 3 ~") and is in the supplied termcap/terminfo. - How does rxvt-unicode determine the encoding to use? - See next entry. + Some other Backspace problems: - Is there an option to switch encodings? - Unlike some other terminals, rxvt-unicode has no encoding switch, and no - specific "utf-8" mode, such as xterm. In fact, it doesn't even know - about UTF-8 or any other encodings with respect to terminal I/O. + some editors use termcap/terminfo, some editors (vim I'm told) expect + Backspace = ^H, GNU Emacs (and Emacs-like editors) use ^H for help. - The reasons is that there exists a perfectly fine mechanism for - selecting the encoding, doing I/O and (most important) communicating - this to all applications so everybody agrees on character properties - such as width and code number. This mechanism is the *locale*. - Applications not using that info will have problems (for example, - "xterm" gets the width of characters wrong as it uses it's own, - locale-independent table under all locales). + Perhaps someday this will all be resolved in a consistent manner. - Rxvt-unicode uses the "LC_CTYPE" locale category to select encoding. All - programs doing the same (that is, most) will automatically agree in the - interpretation of characters. + I don't like the key-bindings. How do I change them? + There are some compile-time selections available via configure. Unless + you have run "configure" with the "--disable-resources" option you can + use the `keysym' resource to alter the keystrings associated with + keysyms. - Unfortunately, there is no system-independent way to select locales, nor - is there a standard on how locale specifiers will look like. + Here's an example for a URxvt session started using "rxvt -name URxvt" - On most systems, the content of the "LC_CTYPE" environment variable - contains an arbitrary string which corresponds to an already-installed - locale. Common names for locales are "en_US.UTF-8", "de_DE.ISO-8859-15", - "ja_JP.EUC-JP", i.e. "language_country.encoding", but other forms (i.e. - "de" or "german") are also common. + URxvt.keysym.Home: \033[1~ + URxvt.keysym.End: \033[4~ + URxvt.keysym.C-apostrophe: \033 + URxvt.keysym.C-slash: \033 + URxvt.keysym.C-semicolon: \033 + URxvt.keysym.C-grave: \033 + URxvt.keysym.C-comma: \033 + URxvt.keysym.C-period: \033 + URxvt.keysym.C-0x60: \033 + URxvt.keysym.C-Tab: \033 + URxvt.keysym.C-Return: \033 + URxvt.keysym.S-Return: \033 + URxvt.keysym.S-space: \033 + URxvt.keysym.M-Up: \033 + URxvt.keysym.M-Down: \033 + URxvt.keysym.M-Left: \033 + URxvt.keysym.M-Right: \033 + URxvt.keysym.M-C-0: list \033 + URxvt.keysym.M-C-a: list \033 + URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 - Rxvt-unicode ignores all other locale categories, and except for the - encoding, ignores country or language-specific settings, i.e. - "de_DE.UTF-8" and "ja_JP.UTF-8" are the normally same to rxvt-unicode. + See some more examples in the documentation for the keysym resource. - If you want to use a specific encoding you have to make sure you start - rxvt-unicode with the correct "LC_CTYPE" category. + I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map + KP_Insert == Insert + F22 == Print + F27 == Home + F29 == Prior + F33 == End + F35 == Next - Can I switch locales at runtime? - Yes, using an escape sequence. Try something like this, which sets - rxvt-unicode's idea of "LC_CTYPE". + Rather than have rxvt-unicode try to accommodate all the various + possible keyboard mappings, it is better to use `xmodmap' to remap the + keys as required for your particular machine. - printf '\e]701;%s\007' ja_JP.SJIS + Terminal Configuration + Why doesn't rxvt-unicode read my resources? + Well, why, indeed? It does, in a way very similar to other X + applications. Most importantly, this means that if you or your OS loads + resources into the X display (the right way to do it), rxvt-unicode will + ignore any resource files in your home directory. It will only read + $HOME/.Xdefaults when no resources are attached to the display. - See also the previous answer. + If you have or use an $HOME/.Xresources file, chances are that resources + are loaded into your X-server. In this case, you have to re-login after + every change (or run xrdb -merge $HOME/.Xresources). - Sometimes this capability is rather handy when you want to work in one - locale (e.g. "de_DE.UTF-8") but some programs don't support it (e.g. - UTF-8). For example, I use this script to start "xjdic", which first - switches to a locale supported by xjdic and back later: + Also consider the form resources have to use: - printf '\e]701;%s\007' ja_JP.SJIS - xjdic -js - printf '\e]701;%s\007' de_DE.UTF-8 + URxvt.resource: value - You can also use xterm's "luit" program, which usually works fine, - except for some locales where character width differs between program- - and rxvt-unicode-locales. + If you want to use another form (there are lots of different ways of + specifying resources), make sure you understand wether and why it works. + If unsure, use the form above. - Can I switch the fonts at runtime? - Yes, using an escape sequence. Try something like this, which has the - same effect as using the "-fn" switch, and takes effect immediately: + When I log-in to another system it tells me about missing terminfo data? + The terminal description used by rxvt-unicode is not as widely available + as that for xterm, or even rxvt (for which the same problem often + arises). - printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" + The correct solution for this problem is to install the terminfo, this + can be done like this (with ncurses' infocmp): - This is useful if you e.g. work primarily with japanese (and prefer a - japanese font), but you have to switch to chinese temporarily, where - japanese fonts would only be in your way. + REMOTE=remotesystem.domain + infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" - You can think of this as a kind of manual ISO-2022 switching. + ... or by installing rxvt-unicode normally on the remote system, - Why do italic characters look as if clipped? - Many fonts have difficulties with italic characters and hinting. For - example, the otherwise very nicely hinted font "xft:Bitstream Vera Sans - Mono" completely fails in it's italic face. A workaround might be to - enable freetype autohinting, i.e. like this: + If you cannot or do not want to do this, then you can simply set + "TERM=rxvt" or even "TERM=xterm", and live with the small number of + problems arising, which includes wrong keymapping, less and different + colours and some refresh errors in fullscreen applications. It's a nice + quick-and-dirty workaround for rare cases, though. - URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true - URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true + If you always want to do this (and are fine with the consequences) you + can either recompile rxvt-unicode with the desired TERM value or use a + resource to set it: - My input method wants but I want UTF-8, what can I do? - You can specify separate locales for the input method and the rest of - the terminal, using the resource "imlocale": + URxvt.termName: rxvt - URxvt.imlocale: ja_JP.EUC-JP + If you don't plan to use rxvt (quite common...) you could also replace + the rxvt terminfo file with the rxvt-unicode one. - Now you can start your terminal with "LC_CTYPE=ja_JP.UTF-8" and still - use your input method. Please note, however, that you will not be able - to input characters outside "EUC-JP" in a normal way then, as your input - method limits you. + "tic" outputs some error when compiling the terminfo entry. + Most likely it's the empty definition for "enacs=". Just replace it by + "enacs=\E[0@" and try again. - Rxvt-unicode crashes when the X Input Method changes or exits. - Unfortunately, this is unavoidable, as the XIM protocol is racy by - design. Applications can avoid some crashes at the expense of memory - leaks, and Input Methods can avoid some crashes by careful ordering at - exit time. kinput2 (and derived input methods) generally succeeds, while - SCIM (or similar input methods) fails. In the end, however, crashes - cannot be completely avoided even if both sides cooperate. + "bash"'s readline does not work correctly under rxvt. + See next entry. - So the only workaround is not to kill your Input Method Servers. + I need a termcap file entry. + One reason you might want this is that some distributions or operating + systems still compile some programs using the long-obsoleted termcap + library (Fedora Core's bash is one example) and rely on a termcap entry + for "rxvt-unicode". - Rxvt-unicode uses gobs of memory, how can I reduce that? - Rxvt-unicode tries to obey the rule of not charging you for something - you don't use. One thing you should try is to configure out all settings - that you don't need, for example, Xft support is a resource hog by - design, when used. Compiling it out ensures that no Xft font will be - loaded accidentally when rxvt-unicode tries to find a font for your - characters. + You could use rxvt's termcap entry with resonable results in many cases. + You can also create a termcap entry by using terminfo's infocmp program + like this: - Also, many people (me included) like large windows and even larger - scrollback buffers: Without "--enable-unicode3", rxvt-unicode will use 6 - bytes per screen cell. For a 160x?? window this amounts to almost a - kilobyte per line. A scrollback buffer of 10000 lines will then (if - full) use 10 Megabytes of memory. With "--enable-unicode3" it gets - worse, as rxvt-unicode then uses 8 bytes per screen cell. + infocmp -C rxvt-unicode - Can I speed up Xft rendering somehow? - Yes, the most obvious way to speed it up is to avoid Xft entirely, as it - is simply slow. If you still want Xft fonts you might try to disable - antialiasing (by appending ":antialias=false"), which saves lots of - memory and also speeds up rendering considerably. + Or you could use this termcap entry, generated by the command above: - Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? - Rxvt-unicode will use whatever you specify as a font. If it needs to - fall back to it's default font search list it will prefer X11 core - fonts, because they are small and fast, and then use Xft fonts. It has - antialiasing disabled for most of them, because the author thinks they - look best that way. + rxvt-unicode|rxvt-unicode terminal (X Window System):\ + :am:bw:eo:km:mi:ms:xn:xo:\ + :co#80:it#8:li#24:lm#0:\ + :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ + :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ + :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ + :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ + :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ + :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ + :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ + :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ + :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ + :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ + :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ + :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ + :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ + :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ + :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ + :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ + :vs=\E[?25h: - If you want antialiasing, you have to specify the fonts manually. + Why does "ls" no longer have coloured output? + The "ls" in the GNU coreutils unfortunately doesn't use terminfo to + decide wether a terminal has colour, but uses it's own configuration + file. Needless to say, "rxvt-unicode" is not in it's default file (among + with most other terminals supporting colour). Either add: - Mouse cut/paste suddenly no longer works. - Make sure that mouse reporting is actually turned off since killing some - editors prematurely may leave the mouse in mouse report mode. I've heard - that tcsh may use mouse reporting unless it otherwise specified. A quick - check is to see if cut/paste works when the Alt or Shift keys are - depressed. + TERM rxvt-unicode - What's with this bold/blink stuff? - If no bold colour is set via "colorBD:", bold will invert text using the - standard foreground colour. + to "/etc/DIR_COLORS" or simply add: - For the standard background colour, blinking will actually make the text - blink when compiled with "--enable-blinking". with standard colours. - Without "--enable-blinking", the blink attribute will be ignored. + alias ls='ls --color=auto' - On ANSI colours, bold/blink attributes are used to set high-intensity - foreground/background colors. + to your ".profile" or ".bashrc". - color0-7 are the low-intensity colors. + Why doesn't vim/emacs etc. use the 88 colour mode? + See next entry. - color8-15 are the corresponding high-intensity colors. + Why doesn't vim/emacs etc. make use of italic? + See next entry. - I don't like the screen colors. How do I change them? - You can change the screen colors at run-time using ~/.Xdefaults - resources (or as long-options). + Why are the secondary screen-related options not working properly? + Make sure you are using "TERM=rxvt-unicode". Some pre-packaged + distributions (most notably Debian GNU/Linux) break rxvt-unicode by + setting "TERM" to "rxvt", which doesn't have these extra features. + Unfortunately, some of these (most notably, again, Debian GNU/Linux) + furthermore fail to even install the "rxvt-unicode" terminfo file, so + you will need to install it on your own (See the question When I log-in + to another system it tells me about missing terminfo data? on how to do + this). - Here are values that are supposed to resemble a VGA screen, including - the murky brown that passes for low-intensity yellow: + Encoding / Locale / Input Method Issues + Rxvt-unicode does not seem to understand the selected encoding? + See next entry. - URxvt.color0: #000000 - URxvt.color1: #A80000 - URxvt.color2: #00A800 - URxvt.color3: #A8A800 - URxvt.color4: #0000A8 - URxvt.color5: #A800A8 - URxvt.color6: #00A8A8 - URxvt.color7: #A8A8A8 + Unicode does not seem to work? + If you encounter strange problems like typing an accented character but + getting two unrelated other characters or similar, or if program output + is subtly garbled, then you should check your locale settings. - URxvt.color8: #000054 - URxvt.color9: #FF0054 - URxvt.color10: #00FF54 - URxvt.color11: #FFFF54 - URxvt.color12: #0000FF - URxvt.color13: #FF00FF - URxvt.color14: #00FFFF - URxvt.color15: #FFFFFF + Rxvt-unicode must be started with the same "LC_CTYPE" setting as the + programs. Often rxvt-unicode is started in the "C" locale, while the + login script running within the rxvt-unicode window changes the locale + to something else, e.g. "en_GB.UTF-8". Needless to say, this is not + going to work. - And here is a more complete set of non-standard colors described (not by - me) as "pretty girly". + The best thing is to fix your startup environment, as you will likely + run into other problems. If nothing works you can try this in your + .profile. - URxvt.cursorColor: #dc74d1 - URxvt.pointerColor: #dc74d1 - URxvt.background: #0e0e0e - URxvt.foreground: #4ad5e1 - URxvt.color0: #000000 - URxvt.color8: #8b8f93 - URxvt.color1: #dc74d1 - URxvt.color9: #dc74d1 - URxvt.color2: #0eb8c7 - URxvt.color10: #0eb8c7 - URxvt.color3: #dfe37e - URxvt.color11: #dfe37e - URxvt.color5: #9e88f0 - URxvt.color13: #9e88f0 - URxvt.color6: #73f7ff - URxvt.color14: #73f7ff - URxvt.color7: #e1dddd - URxvt.color15: #e1dddd + printf '\e]701;%s\007' "$LC_CTYPE" - How can I start rxvtd in a race-free way? - Try "rxvtd -f -o", which tells rxvtd to open the display, create the - listening socket and then fork. + If this doesn't work, then maybe you use a "LC_CTYPE" specification not + supported on your systems. Some systems have a "locale" command which + displays this (also, "perl -e0" can be used to check locale settings, as + it will complain loudly if it cannot set the locale). If it displays + something like: - What's with the strange Backspace/Delete key behaviour? - Assuming that the physical Backspace key corresponds to the BackSpace - keysym (not likely for Linux ... see the following question) there are - two standard values that can be used for Backspace: "^H" and "^?". + locale: Cannot set LC_CTYPE to default locale: ... - Historically, either value is correct, but rxvt-unicode adopts the - debian policy of using "^?" when unsure, because it's the one only only - correct choice :). + Then the locale you specified is not supported on your system. - Rxvt-unicode tries to inherit the current stty settings and uses the - value of `erase' to guess the value for backspace. If rxvt-unicode - wasn't started from a terminal (say, from a menu or by remote shell), - then the system value of `erase', which corresponds to CERASE in - , will be used (which may not be the same as your stty - setting). + If nothing works and you are sure that everything is set correctly then + you will need to remember a little known fact: Some programs just don't + support locales :( - For starting a new rxvt-unicode: + How does rxvt-unicode determine the encoding to use? + See next entry. - # use Backspace = ^H - $ stty erase ^H - $ rxvt + Is there an option to switch encodings? + Unlike some other terminals, rxvt-unicode has no encoding switch, and no + specific "utf-8" mode, such as xterm. In fact, it doesn't even know + about UTF-8 or any other encodings with respect to terminal I/O. - # use Backspace = ^? - $ stty erase ^? - $ rxvt + The reasons is that there exists a perfectly fine mechanism for + selecting the encoding, doing I/O and (most important) communicating + this to all applications so everybody agrees on character properties + such as width and code number. This mechanism is the *locale*. + Applications not using that info will have problems (for example, + "xterm" gets the width of characters wrong as it uses it's own, + locale-independent table under all locales). - Toggle with "ESC [ 36 h" / "ESC [ 36 l". + Rxvt-unicode uses the "LC_CTYPE" locale category to select encoding. All + programs doing the same (that is, most) will automatically agree in the + interpretation of characters. - For an existing rxvt-unicode: + Unfortunately, there is no system-independent way to select locales, nor + is there a standard on how locale specifiers will look like. - # use Backspace = ^H - $ stty erase ^H - $ echo -n "^[[36h" + On most systems, the content of the "LC_CTYPE" environment variable + contains an arbitrary string which corresponds to an already-installed + locale. Common names for locales are "en_US.UTF-8", "de_DE.ISO-8859-15", + "ja_JP.EUC-JP", i.e. "language_country.encoding", but other forms (i.e. + "de" or "german") are also common. - # use Backspace = ^? - $ stty erase ^? - $ echo -n "^[[36l" + Rxvt-unicode ignores all other locale categories, and except for the + encoding, ignores country or language-specific settings, i.e. + "de_DE.UTF-8" and "ja_JP.UTF-8" are the normally same to rxvt-unicode. - This helps satisfy some of the Backspace discrepancies that occur, but - if you use Backspace = "^H", make sure that the termcap/terminfo value - properly reflects that. + If you want to use a specific encoding you have to make sure you start + rxvt-unicode with the correct "LC_CTYPE" category. - The Delete key is a another casualty of the ill-defined Backspace - problem. To avoid confusion between the Backspace and Delete keys, the - Delete key has been assigned an escape sequence to match the vt100 for - Execute ("ESC [ 3 ~") and is in the supplied termcap/terminfo. + Can I switch locales at runtime? + Yes, using an escape sequence. Try something like this, which sets + rxvt-unicode's idea of "LC_CTYPE". - Some other Backspace problems: + printf '\e]701;%s\007' ja_JP.SJIS - some editors use termcap/terminfo, some editors (vim I'm told) expect - Backspace = ^H, GNU Emacs (and Emacs-like editors) use ^H for help. + See also the previous answer. - Perhaps someday this will all be resolved in a consistent manner. + Sometimes this capability is rather handy when you want to work in one + locale (e.g. "de_DE.UTF-8") but some programs don't support it (e.g. + UTF-8). For example, I use this script to start "xjdic", which first + switches to a locale supported by xjdic and back later: - I don't like the key-bindings. How do I change them? - There are some compile-time selections available via configure. Unless - you have run "configure" with the "--disable-resources" option you can - use the `keysym' resource to alter the keystrings associated with - keysyms. + printf '\e]701;%s\007' ja_JP.SJIS + xjdic -js + printf '\e]701;%s\007' de_DE.UTF-8 - Here's an example for a URxvt session started using "rxvt -name URxvt" + You can also use xterm's "luit" program, which usually works fine, + except for some locales where character width differs between program- + and rxvt-unicode-locales. - URxvt.keysym.Home: \033[1~ - URxvt.keysym.End: \033[4~ - URxvt.keysym.C-apostrophe: \033 - URxvt.keysym.C-slash: \033 - URxvt.keysym.C-semicolon: \033 - URxvt.keysym.C-grave: \033 - URxvt.keysym.C-comma: \033 - URxvt.keysym.C-period: \033 - URxvt.keysym.C-0x60: \033 - URxvt.keysym.C-Tab: \033 - URxvt.keysym.C-Return: \033 - URxvt.keysym.S-Return: \033 - URxvt.keysym.S-space: \033 - URxvt.keysym.M-Up: \033 - URxvt.keysym.M-Down: \033 - URxvt.keysym.M-Left: \033 - URxvt.keysym.M-Right: \033 - URxvt.keysym.M-C-0: list \033 - URxvt.keysym.M-C-a: list \033 - URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 + My input method wants but I want UTF-8, what can I do? + You can specify separate locales for the input method and the rest of + the terminal, using the resource "imlocale": - See some more examples in the documentation for the keysym resource. + URxvt.imlocale: ja_JP.EUC-JP - I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following mappings that rxvt-unicode doesn't recognize. - KP_Insert == Insert - F22 == Print - F27 == Home - F29 == Prior - F33 == End - F35 == Next + Now you can start your terminal with "LC_CTYPE=ja_JP.UTF-8" and still + use your input method. Please note, however, that you will not be able + to input characters outside "EUC-JP" in a normal way then, as your input + method limits you. - Rather than have rxvt-unicode try to accommodate all the various - possible keyboard mappings, it is better to use `xmodmap' to remap the - keys as required for your particular machine. + Rxvt-unicode crashes when the X Input Method changes or exits. + Unfortunately, this is unavoidable, as the XIM protocol is racy by + design. Applications can avoid some crashes at the expense of memory + leaks, and Input Methods can avoid some crashes by careful ordering at + exit time. kinput2 (and derived input methods) generally succeeds, while + SCIM (or similar input methods) fails. In the end, however, crashes + cannot be completely avoided even if both sides cooperate. - How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc. - rxvt and rxvt-unicode always export the variable "COLORTERM", so you can - check and see if that is set. Note that several programs, JED, slrn, - Midnight Commander automatically check this variable to decide whether - or not to use color. + So the only workaround is not to kill your Input Method Servers. - How do I set the correct, full IP address for the DISPLAY variable? - If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled - insecure mode then it is possible to use the following shell script - snippets to correctly set the display. If your version of rxvt-unicode - wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) - then the COLORTERM variable can be used to distinguish rxvt-unicode from - a regular xterm. + Operating Systems / Package Maintaining + I am using Debian GNU/Linux and have a problem... + The Debian GNU/Linux package of rxvt-unicode in sarge contains large + patches that considerably change the behaviour of rxvt-unicode (but + unfortunately this notice has been removed). Before reporting a bug to + the original rxvt-unicode author please download and install the genuine + version () and try to reproduce + the problem. If you cannot, chances are that the problems are specific + to Debian GNU/Linux, in which case it should be reported via the Debian + Bug Tracking System (use "reportbug" to report the bug). - Courtesy of Chuck Blake with the following shell script - snippets: + For other problems that also affect the Debian package, you can and + probably should use the Debian BTS, too, because, after all, it's also a + bug in the Debian version and it serves as a reminder for other users + that might encounter the same issue. - # Bourne/Korn/POSIX family of shells: - [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know - if [ ${TERM:-foo} = xterm ]; then - stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not - echo -n '^[Z' - read term_id - stty icanon echo - if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then - echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string - read DISPLAY # set it in our local shell - fi - fi + I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation? + You should build one binary with the default options. configure now + enables most useful options, and the trend goes to making them + runtime-switchable, too, so there is usually no drawback to enbaling + them, except higher disk and possibly memory usage. The perl interpreter + should be enabled, as important functionality (menus, selection, likely + more in the future) depends on it. - How do I compile the manual pages for myself? - You need to have a recent version of perl installed as /usr/bin/perl, - one that comes with pod2man, pod2text and pod2html. Then go to the doc - subdirectory and enter "make alldoc". + You should not overwrite the "perl-ext-common" snd "perl-ext" resources + system-wide (except maybe with "defaults"). This will result in useful + behaviour. If your distribution aims at low memory, add an empty + "perl-ext-common" resource to the app-defaults file. This will keep the + perl interpreter disabled until the user enables it. - My question isn't answered here, can I ask a human? - Before sending me mail, you could go to IRC: "irc.freenode.net", channel - "#rxvt-unicode" has some rxvt-unicode enthusiasts that might be - interested in learning about new and exciting problems (but not FAQs :). + If you can/want build more binaries, I recommend building a minimal one + with "--disable-everything" (very useful) and a maximal one with + "--enable-everything" (less useful, it will be very big due to a lot of + encodings built-in that increase download times and are rarely used). + + I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe? + It should be, starting with release 7.1. You are encouraged to properly + install urxvt with privileges necessary for your OS now. + + When rxvt-unicode detects that it runs setuid or setgid, it will fork + into a helper process for privileged operations (pty handling on some + systems, utmp/wtmp/lastlog handling on others) and drop privileges + immediately. This is much safer than most other terminals that keep + privileges while running (but is more relevant to urxvt, as it contains + things as perl interpreters, which might be "helpful" to attackers). + + This forking is done as the very first within main(), which is very + early and reduces possible bugs to initialisation code run before + main(), or things like the dynamic loader of your system, which should + result in very little risk. + + On Solaris 9, many line-drawing characters are too wide. + Seems to be a known bug, read + . Some people use the + following ugly workaround to get non-double-wide-characters working: + + #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) + + I am on FreeBSD and rxvt-unicode does not seem to work at all. + Rxvt-unicode requires the symbol "__STDC_ISO_10646__" to be defined in + your compile environment, or an implementation that implements it, + wether it defines the symbol or not. "__STDC_ISO_10646__" requires that + wchar_t is represented as unicode. + + As you might have guessed, FreeBSD does neither define this symobl nor + does it support it. Instead, it uses it's own internal representation of + wchar_t. This is, of course, completely fine with respect to standards. + + However, that means rxvt-unicode only works in "POSIX", "ISO-8859-1" and + "UTF-8" locales under FreeBSD (which all use Unicode as wchar_t. + + "__STDC_ISO_10646__" is the only sane way to support multi-language apps + in an OS, as using a locale-dependent (and non-standardized) + representation of wchar_t makes it impossible to convert between wchar_t + (as used by X11 and your applications) and any other encoding without + implementing OS-specific-wrappers for each and every locale. There + simply are no APIs to convert wchar_t into anything except the current + locale encoding. + + Some applications (such as the formidable mlterm) work around this by + carrying their own replacement functions for character set handling with + them, and either implementing OS-dependent hacks or doing multiple + conversions (which is slow and unreliable in case the OS implements + encodings slightly different than the terminal emulator). + + The rxvt-unicode author insists that the right way to fix this is in the + system libraries once and for all, instead of forcing every app to carry + complete replacements for them :) + + I use Solaris 9 and it doesn't compile/work/etc. + Try the diff in doc/solaris9.patch as a base. It fixes the worst + problems with "wcwidth" and a compile problem. + + How can I use rxvt-unicode under cygwin? + rxvt-unicode should compile and run out of the box on cygwin, using the + X11 libraries that come with cygwin. libW11 emulation is no longer + supported (and makes no sense, either, as it only supported a single + font). I recommend starting the X-server in "-multiwindow" or + "-rootless" mode instead, which will result in similar look&feel as the + old libW11 emulation. + + At the time of this writing, cygwin didn't seem to support any + multi-byte encodings (you might try "LC_CTYPE=C-UTF-8"), so you are + likely limited to 8-bit encodings. diff --git a/doc/rxvt.7.html b/doc/rxvt.7.html index de788312..e254a15e 100644 --- a/doc/rxvt.7.html +++ b/doc/rxvt.7.html @@ -18,67 +18,92 @@
  • FREQUENTLY ASKED QUESTIONS
  • RXVT TECHNICAL REFERENCE
    • @@ -129,115 +154,94 @@ all escape sequences, and other background information.

    FREQUENTLY ASKED QUESTIONS

    -

    The new selection selects pieces that are too big, how can I select -single words?

    -

    If you want to select e.g. alphanumeric words, you can use the following -setting:

    -
    -   URxvt.selection.pattern-0: ([[:word:]]+)
    -

    If you click more than twice, the selection will be extended -more and more.

    -

    To get a selection that is very similar to the old code, try this pattern:

    -
    -   URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+)
    -

    Please also note that the LeftClick Shift-LeftClik combination also -selects words like the old code.

    +

    Meta, Features & Commandline Issues

    -

    I don't like the new selection/popups/hotkeys/perl, how do I -change/disable it?

    -

    You can disable the perl extension completely by setting the -perl-ext-common resource to the empty string, which also keeps -rxvt-unicode from initialising perl, saving memory.

    -

    If you only want to disable specific features, you first have to -identify which perl extension is responsible. For this, read the section -PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For -example, to disable the selection-popup and option-popup, specify -this perl-ext-common resource:

    +

    My question isn't answered here, can I ask a human?

    +

    Before sending me mail, you could go to IRC: irc.freenode.net, +channel #rxvt-unicode has some rxvt-unicode enthusiasts that might be +interested in learning about new and exciting problems (but not FAQs :).

    +

    +

    +

    Does it support tabs, can I have a tabbed rxvt-unicode?

    +

    Beginning with version 7.3, there is a perl extension that implements a +simple tabbed terminal. It is installed by default, so any of these should +give you tabs:

    -   URxvt.perl-ext-common: default,-selection-popup,-option-popup
    -

    This will keep the default extensions, but disable the two popup -extensions. Some extensions can also be configured, for example, -scrollback search mode is triggered by M-s. You can move it to any -other combination either by setting the searchable-scrollback resource:

    + rxvt -pe tabbed 
    -   URxvt.searchable-scrollback: CM-s
    + URxvt.perl-ext-common: default,tabbed +

    It will also work fine with tabbing functionality of many window managers +or similar tabbing programs, and its embedding-features allow it to be +embedded into other programs, as witnessed by doc/rxvt-tabbed or +the upcoming Gtk2::URxvt perl module, which features a tabbed urxvt +(murxvt) terminal as an example embedding application.

    -

    The cursor moves when selecting text in the current input line, how -do I switch this off?

    -

    See next entry.

    +

    How do I know which rxvt-unicode version I'm using?

    +

    The version number is displayed with the usage (-h). Also the escape +sequence ESC [ 8 n sets the window title to the version number. When +using the rxvtc client, the version displayed is that of the +daemon.

    -

    During rlogin/ssh/telnet/etc. sessions, clicking near the cursor -outputs strange escape sequences, how do I fix this?

    -

    These are caused by the readline perl extension. Under normal -circumstances, it will move your cursor around when you click into the -line that contains it. It tries hard not to do this at the wrong moment, -but when running a program that doesn't parse cursor movements or in some -cases during rlogin sessions, it fails to detect this properly.

    -

    You can permamently switch this feature off by disabling the readline -extension:

    -
    -   URxvt.perl-ext-common: default,-readline
    +

    Rxvt-unicode uses gobs of memory, how can I reduce that?

    +

    Rxvt-unicode tries to obey the rule of not charging you for something you +don't use. One thing you should try is to configure out all settings that +you don't need, for example, Xft support is a resource hog by design, +when used. Compiling it out ensures that no Xft font will be loaded +accidentally when rxvt-unicode tries to find a font for your characters.

    +

    Also, many people (me included) like large windows and even larger +scrollback buffers: Without --enable-unicode3, rxvt-unicode will use +6 bytes per screen cell. For a 160x?? window this amounts to almost a +kilobyte per line. A scrollback buffer of 10000 lines will then (if full) +use 10 Megabytes of memory. With --enable-unicode3 it gets worse, as +rxvt-unicode then uses 8 bytes per screen cell.

    -

    Why doesn't rxvt-unicode read my resources?

    -

    Well, why, indeed? It does, in a way very similar to other X -applications. Most importantly, this means that if you or your OS loads -resources into the X display (the right way to do it), rxvt-unicode will -ignore any resource files in your home directory. It will only read -$HOME/.Xdefaults when no resources are attached to the display.

    -

    If you have or use an $HOME/.Xresources file, chances are that -resources are loaded into your X-server. In this case, you have to -re-login after every change (or run xrdb -merge $HOME/.Xresources).

    -

    Also consider the form resources have to use:

    -
    -  URxvt.resource: value
    -

    If you want to use another form (there are lots of different ways of -specifying resources), make sure you understand wether and why it -works. If unsure, use the form above.

    +

    How can I start rxvtd in a race-free way?

    +

    Try rxvtd -f -o, which tells rxvtd to open the +display, create the listening socket and then fork.

    -

    I can't get transparency working, what am I doing wrong?

    -

    First of all, transparency isn't officially supported in rxvt-unicode, so -you are mostly on your own. Do not bug the author about it (but you may -bug everybody else). Also, if you can't get it working consider it a rite -of passage: ... and you failed.

    -

    Here are four ways to get transparency. Do read the manpage and option -descriptions for the programs mentioned and rxvt-unicode. Really, do it!

    -

    1. Use inheritPixmap:

    -
    -   Esetroot wallpaper.jpg
-   rxvt -ip -tint red -sh 40
    -

    That works. If you think it doesn't, you lack transparency and tinting -support, or you are unable to read.

    -

    2. Use a simple pixmap and emulate pseudo-transparency. This enables you -to use effects other than tinting and shading: Just shade/tint/whatever -your picture with gimp:

    -
    -   convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm
-   rxvt -pixmap background.xpm -pe automove-background
    -

    That works. If you think it doesn't, you lack XPM and Perl support, or you -are unable to read.

    -

    3. Use an ARGB visual:

    -
    -   rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc
    -

    This requires XFT support, and the support of your X-server. If that -doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't -there yet, no matter what they claim. Rxvt-Unicode contains the neccessary -bugfixes and workarounds for Xft and Xlib to make it work, but that -doesn't mean that your WM has the required kludges in place.

    -

    4. Use xcompmgr and let it do the job:

    +

    How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc.

    +

    rxvt and rxvt-unicode always export the variable ``COLORTERM'', so you can +check and see if that is set. Note that several programs, JED, slrn, +Midnight Commander automatically check this variable to decide whether or +not to use color.

    +

    +

    +

    How do I set the correct, full IP address for the DISPLAY variable?

    +

    If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled +insecure mode then it is possible to use the following shell script +snippets to correctly set the display. If your version of rxvt-unicode +wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) then +the COLORTERM variable can be used to distinguish rxvt-unicode from a +regular xterm.

    +

    Courtesy of Chuck Blake <cblake@BBN.COM> with the following shell script +snippets:

    -  xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \
-        -set _NET_WM_WINDOW_OPACITY 0xc0000000
    -

    Then click on a window you want to make transparent. Replace 0xc0000000 -by other values to change the degree of opacity. If it doesn't work and -your server crashes, you got to keep the pieces.

    + # Bourne/Korn/POSIX family of shells: + [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know + if [ ${TERM:-foo} = xterm ]; then + stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not + echo -n '^[Z' + read term_id + stty icanon echo + if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then + echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string + read DISPLAY # set it in our local shell + fi + fi

    -

    Isn't rxvt supposed to be small? Don't all those features bloat?

    +

    How do I compile the manual pages on my own?

    +

    You need to have a recent version of perl installed as /usr/bin/perl, +one that comes with pod2man, pod2text and pod2html. Then go to +the doc subdirectory and enter make alldoc.

    +

    +

    +

    Isn't rxvt-unicode supposed to be small? Don't all those features bloat?

    I often get asked about this, and I think, no, they didn't cause extra bloat. If you compare a minimal rxvt and a minimal urxvt, you can see that the urxvt binary is larger (due to some encoding tables always being @@ -275,7 +279,7 @@ startup time, including the hundreds of warnings it spits out), it fares extremely well *g*.

    -

    Why C++, isn't that unportable/bloated/uncool?

    +

    Why C++, isn't that unportable/bloated/uncool?

    Is this a question? :) It comes up very often. The simple answer is: I had to write it, and C++ allowed me to write and maintain it in a fraction of the time and effort (which is a scarce resource for me). Put even @@ -305,263 +309,69 @@ system with a minimal config:

    except maybe libX11 :)

    -

    Does it support tabs, can I have a tabbed rxvt-unicode?

    -

    Beginning with version 7.3, there is a perl extension that implements a -simple tabbed terminal. It is installed by default, so any of these should -give you tabs:

    -
    -   rxvt -pe tabbed
    -
    -   URxvt.perl-ext-common: default,tabbed
    -

    It will also work fine with tabbing functionality of many window managers -or similar tabbing programs, and its embedding-features allow it to be -embedded into other programs, as witnessed by doc/rxvt-tabbed or -the upcoming Gtk2::URxvt perl module, which features a tabbed urxvt -(murxvt) terminal as an example embedding application.

    +

    Rendering, Font & Look and Feel Issues

    -

    How do I know which rxvt-unicode version I'm using?

    -

    The version number is displayed with the usage (-h). Also the escape -sequence ESC [ 8 n sets the window title to the version number. When -using the rxvtc client, the version displayed is that of the -daemon.

    +

    I can't get transparency working, what am I doing wrong?

    +

    First of all, transparency isn't officially supported in rxvt-unicode, so +you are mostly on your own. Do not bug the author about it (but you may +bug everybody else). Also, if you can't get it working consider it a rite +of passage: ... and you failed.

    +

    Here are four ways to get transparency. Do read the manpage and option +descriptions for the programs mentioned and rxvt-unicode. Really, do it!

    +

    1. Use inheritPixmap:

    +
    +   Esetroot wallpaper.jpg
+   rxvt -ip -tint red -sh 40
    +

    That works. If you think it doesn't, you lack transparency and tinting +support, or you are unable to read.

    +

    2. Use a simple pixmap and emulate pseudo-transparency. This enables you +to use effects other than tinting and shading: Just shade/tint/whatever +your picture with gimp or any other tool:

    +
    +   convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm
+   rxvt -pixmap background.xpm -pe automove-background
    +

    That works. If you think it doesn't, you lack XPM and Perl support, or you +are unable to read.

    +

    3. Use an ARGB visual:

    +
    +   rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc
    +

    This requires XFT support, and the support of your X-server. If that +doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't +there yet, no matter what they claim. Rxvt-Unicode contains the neccessary +bugfixes and workarounds for Xft and Xlib to make it work, but that +doesn't mean that your WM has the required kludges in place.

    +

    4. Use xcompmgr and let it do the job:

    +
    +  xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \
+        -set _NET_WM_WINDOW_OPACITY 0xc0000000
    +

    Then click on a window you want to make transparent. Replace 0xc0000000 +by other values to change the degree of opacity. If it doesn't work and +your server crashes, you got to keep the pieces.

    -

    I am using Debian GNU/Linux and have a problem...

    -

    The Debian GNU/Linux package of rxvt-unicode in sarge contains large -patches that considerably change the behaviour of rxvt-unicode (but -unfortunately this notice has been removed). Before reporting a bug to -the original rxvt-unicode author please download and install the genuine -version (http://software.schmorp.de#rxvt-unicode) and try to reproduce -the problem. If you cannot, chances are that the problems are specific to -Debian GNU/Linux, in which case it should be reported via the Debian Bug -Tracking System (use reportbug to report the bug).

    -

    For other problems that also affect the Debian package, you can and -probably should use the Debian BTS, too, because, after all, it's also a -bug in the Debian version and it serves as a reminder for other users that -might encounter the same issue.

    +

    Why do some chinese characters look so different than others?

    +

    This is because there is a difference between script and language -- +rxvt-unicode does not know which language the text that is output is, +as it only knows the unicode character codes. If rxvt-unicode first +sees a japanese/chinese character, it might choose a japanese font for +display. Subsequent japanese characters will use that font. Now, many +chinese characters aren't represented in japanese fonts, so when the first +non-japanese character comes up, rxvt-unicode will look for a chinese font +-- unfortunately at this point, it will still use the japanese font for +chinese characters that are also in the japanese font.

    +

    The workaround is easy: just tag a chinese font at the end of your font +list (see the previous question). The key is to view the font list as +a preference list: If you expect more japanese, list a japanese font +first. If you expect more chinese, put a chinese font first.

    +

    In the future it might be possible to switch language preferences at +runtime (the internal data structure has no problem with using different +fonts for the same character at the same time, but no interface for this +has been designed yet).

    +

    Until then, you might get away with switching fonts at runtime (see Can I switch the fonts at runtime? later in this document).

    -

    I am maintaining rxvt-unicode for distribution/OS XXX, any -recommendation?

    -

    You should build one binary with the default options. configure -now enables most useful options, and the trend goes to making them -runtime-switchable, too, so there is usually no drawback to enbaling them, -except higher disk and possibly memory usage. The perl interpreter should -be enabled, as important functionality (menus, selection, likely more in -the future) depends on it.

    -

    You should not overwrite the perl-ext-common snd perl-ext resources -system-wide (except maybe with defaults). This will result in useful -behaviour. If your distribution aims at low memory, add an empty -perl-ext-common resource to the app-defaults file. This will keep the -perl interpreter disabled until the user enables it.

    -

    If you can/want build more binaries, I recommend building a minimal -one with --disable-everything (very useful) and a maximal one with ---enable-everything (less useful, it will be very big due to a lot of -encodings built-in that increase download times and are rarely used).

    -

    -

    -

    I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe?

    -

    It should be, starting with release 7.1. You are encouraged to properly -install urxvt with privileges necessary for your OS now.

    -

    When rxvt-unicode detects that it runs setuid or setgid, it will fork -into a helper process for privileged operations (pty handling on some -systems, utmp/wtmp/lastlog handling on others) and drop privileges -immediately. This is much safer than most other terminals that keep -privileges while running (but is more relevant to urxvt, as it contains -things as perl interpreters, which might be ``helpful'' to attackers).

    -

    This forking is done as the very first within main(), which is very early -and reduces possible bugs to initialisation code run before main(), or -things like the dynamic loader of your system, which should result in very -little risk.

    -

    -

    -

    When I log-in to another system it tells me about missing terminfo data?

    -

    The terminal description used by rxvt-unicode is not as widely available -as that for xterm, or even rxvt (for which the same problem often arises).

    -

    The correct solution for this problem is to install the terminfo, this can -be done like this (with ncurses' infocmp):

    -
    -   REMOTE=remotesystem.domain
-   infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti"
    -

    ... or by installing rxvt-unicode normally on the remote system,

    -

    If you cannot or do not want to do this, then you can simply set -TERM=rxvt or even TERM=xterm, and live with the small number of -problems arising, which includes wrong keymapping, less and different -colours and some refresh errors in fullscreen applications. It's a nice -quick-and-dirty workaround for rare cases, though.

    -

    If you always want to do this (and are fine with the consequences) you -can either recompile rxvt-unicode with the desired TERM value or use a -resource to set it:

    -
    -   URxvt.termName: rxvt
    -

    If you don't plan to use rxvt (quite common...) you could also replace -the rxvt terminfo file with the rxvt-unicode one.

    -

    -

    -

    tic outputs some error when compiling the terminfo entry.

    -

    Most likely it's the empty definition for enacs=. Just replace it by -enacs=\E[0@ and try again.

    -

    -

    -

    bash's readline does not work correctly under rxvt.

    -

    See next entry.

    -

    -

    -

    I need a termcap file entry.

    -

    One reason you might want this is that some distributions or operating -systems still compile some programs using the long-obsoleted termcap -library (Fedora Core's bash is one example) and rely on a termcap entry -for rxvt-unicode.

    -

    You could use rxvt's termcap entry with resonable results in many cases. -You can also create a termcap entry by using terminfo's infocmp program -like this:

    -
    -   infocmp -C rxvt-unicode
    -

    Or you could use this termcap entry, generated by the command above:

    -
    -   rxvt-unicode|rxvt-unicode terminal (X Window System):\
-           :am:bw:eo:km:mi:ms:xn:xo:\
-           :co#80:it#8:li#24:lm#0:\
-           :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\
-           :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\
-           :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\
-           :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\
-           :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\
-           :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\
-           :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\
-           :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\
-           :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\
-           :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\
-           :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\
-           :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\
-           :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\
-           :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\
-           :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\
-           :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\
-           :vs=\E[?25h:
    -

    -

    -

    Why does ls no longer have coloured output?

    -

    The ls in the GNU coreutils unfortunately doesn't use terminfo to -decide wether a terminal has colour, but uses it's own configuration -file. Needless to say, rxvt-unicode is not in it's default file (among -with most other terminals supporting colour). Either add:

    -
    -   TERM rxvt-unicode
    -

    to /etc/DIR_COLORS or simply add:

    -
    -   alias ls='ls --color=auto'
    -

    to your .profile or .bashrc.

    -

    -

    -

    Why doesn't vim/emacs etc. use the 88 colour mode?

    -

    See next entry.

    -

    -

    -

    Why doesn't vim/emacs etc. make use of italic?

    -

    See next entry.

    -

    -

    -

    Why are the secondary screen-related options not working properly?

    -

    Make sure you are using TERM=rxvt-unicode. Some pre-packaged -distributions (most notably Debian GNU/Linux) break rxvt-unicode -by setting TERM to rxvt, which doesn't have these extra -features. Unfortunately, some of these (most notably, again, Debian -GNU/Linux) furthermore fail to even install the rxvt-unicode terminfo -file, so you will need to install it on your own (See the question When -I log-in to another system it tells me about missing terminfo data? on -how to do this).

    -

    -

    -

    My numerical keypad acts weird and generates differing output?

    -

    Some Debian GNUL/Linux users seem to have this problem, although no -specific details were reported so far. It is possible that this is caused -by the wrong TERM setting, although the details of wether and how -this can happen are unknown, as TERM=rxvt should offer a compatible -keymap. See the answer to the previous question, and please report if that -helped.

    -

    -

    -

    Rxvt-unicode does not seem to understand the selected encoding?

    -

    See next entry.

    -

    -

    -

    Unicode does not seem to work?

    -

    If you encounter strange problems like typing an accented character but -getting two unrelated other characters or similar, or if program output is -subtly garbled, then you should check your locale settings.

    -

    Rxvt-unicode must be started with the same LC_CTYPE setting as the -programs. Often rxvt-unicode is started in the C locale, while the -login script running within the rxvt-unicode window changes the locale to -something else, e.g. en_GB.UTF-8. Needless to say, this is not going to work.

    -

    The best thing is to fix your startup environment, as you will likely run -into other problems. If nothing works you can try this in your .profile.

    -
    -  printf '\e]701;%s\007' "$LC_CTYPE"
    -

    If this doesn't work, then maybe you use a LC_CTYPE specification not -supported on your systems. Some systems have a locale command which -displays this (also, perl -e0 can be used to check locale settings, as -it will complain loudly if it cannot set the locale). If it displays something -like:

    -
    -  locale: Cannot set LC_CTYPE to default locale: ...
    -

    Then the locale you specified is not supported on your system.

    -

    If nothing works and you are sure that everything is set correctly then -you will need to remember a little known fact: Some programs just don't -support locales :(

    -

    -

    -

    Why do some characters look so much different than others?

    -

    See next entry.

    -

    -

    -

    How does rxvt-unicode choose fonts?

    -

    Most fonts do not contain the full range of Unicode, which is -fine. Chances are that the font you (or the admin/package maintainer of -your system/os) have specified does not cover all the characters you want -to display.

    -

    rxvt-unicode makes a best-effort try at finding a replacement -font. Often the result is fine, but sometimes the chosen font looks -bad/ugly/wrong. Some fonts have totally strange characters that don't -resemble the correct glyph at all, and rxvt-unicode lacks the artificial -intelligence to detect that a specific glyph is wrong: it has to believe -the font that the characters it claims to contain indeed look correct.

    -

    In that case, select a font of your taste and add it to the font list, -e.g.:

    -
    -   rxvt -fn basefont,font2,font3...
    -

    When rxvt-unicode sees a character, it will first look at the base -font. If the base font does not contain the character, it will go to the -next font, and so on. Specifying your own fonts will also speed up this -search and use less resources within rxvt-unicode and the X-server.

    -

    The only limitation is that none of the fonts may be larger than the base -font, as the base font defines the terminal character cell size, which -must be the same due to the way terminals work.

    -

    -

    -

    Why do some chinese characters look so different than others?

    -

    This is because there is a difference between script and language -- -rxvt-unicode does not know which language the text that is output is, -as it only knows the unicode character codes. If rxvt-unicode first -sees a japanese/chinese character, it might choose a japanese font for -display. Subsequent japanese characters will use that font. Now, many -chinese characters aren't represented in japanese fonts, so when the first -non-japanese character comes up, rxvt-unicode will look for a chinese font --- unfortunately at this point, it will still use the japanese font for -chinese characters that are also in the japanese font.

    -

    The workaround is easy: just tag a chinese font at the end of your font -list (see the previous question). The key is to view the font list as -a preference list: If you expect more japanese, list a japanese font -first. If you expect more chinese, put a chinese font first.

    -

    In the future it might be possible to switch language preferences at -runtime (the internal data structure has no problem with using different -fonts for the same character at the same time, but no interface for this -has been designed yet).

    -

    Until then, you might get away with switching fonts at runtime (see Can I switch the fonts at runtime? later in this document).

    -

    -

    -

    Why does rxvt-unicode sometimes leave pixel droppings?

    +

    Why does rxvt-unicode sometimes leave pixel droppings?

    Most fonts were not designed for terminal use, which means that character size varies a lot. A font that is otherwise fine for terminal use might contain some characters that are simply too wide. Rxvt-unicode will avoid @@ -580,34 +390,7 @@ might be forced to use a different font.

    box data is correct.

    -

    On Solaris 9, many line-drawing characters are too wide.

    -

    Seems to be a known bug, read -http://nixdoc.net/files/forum/about34198.html. Some people use the -following ugly workaround to get non-double-wide-characters working:

    -
    -   #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x)
    -

    -

    -

    My Compose (Multi_key) key is no longer working.

    -

    The most common causes for this are that either your locale is not set -correctly, or you specified a preeditStyle that is not supported by -your input method. For example, if you specified OverTheSpot and -your input method (e.g. the default input method handling Compose keys) -does not support this (for instance because it is not visual), then -rxvt-unicode will continue without an input method.

    -

    In this case either do not specify a preeditStyle or specify more than -one pre-edit style, such as OverTheSpot,Root,None.

    -

    -

    -

    I cannot type Ctrl-Shift-2 to get an ASCII NUL character due to ISO 14755

    -

    Either try Ctrl-2 alone (it often is mapped to ASCII NUL even on -international keyboards) or simply use ISO 14755 support to your -advantage, typing <Ctrl-Shift-0> to get a ASCII NUL. This works for other -codes, too, such as Ctrl-Shift-1-d to type the default telnet escape -character and so on.

    -

    -

    -

    How can I keep rxvt-unicode from using reverse video so much?

    +

    How can I keep rxvt-unicode from using reverse video so much?

    First of all, make sure you are running with the right terminal settings (TERM=rxvt-unicode), which will get rid of most of these effects. Then make sure you have specified colours for italic and bold, as otherwise @@ -617,7 +400,7 @@ rxvt-unicode might use reverse video to simulate the effect:

    URxvt.colorIT: green

    -

    Some programs assume totally weird colours (red instead of blue), how can I fix that?

    +

    Some programs assume totally weird colours (red instead of blue), how can I fix that?

    For some unexplainable reason, some rare programs assume a very weird colour palette when confronted with a terminal with more than the standard 8 colours (rxvt-unicode supports 88). The right fix is, of course, to fix @@ -627,166 +410,35 @@ definition to only claim 8 colour support or use TERM=rxvt, which w fix colours but keep you from using other rxvt-unicode features.

    -

    I am on FreeBSD and rxvt-unicode does not seem to work at all.

    -

    Rxvt-unicode requires the symbol __STDC_ISO_10646__ to be defined -in your compile environment, or an implementation that implements it, -wether it defines the symbol or not. __STDC_ISO_10646__ requires that -wchar_t is represented as unicode.

    -

    As you might have guessed, FreeBSD does neither define this symobl nor -does it support it. Instead, it uses it's own internal representation of -wchar_t. This is, of course, completely fine with respect to standards.

    -

    However, that means rxvt-unicode only works in POSIX, ISO-8859-1 and -UTF-8 locales under FreeBSD (which all use Unicode as wchar_t.

    -

    __STDC_ISO_10646__ is the only sane way to support multi-language -apps in an OS, as using a locale-dependent (and non-standardized) -representation of wchar_t makes it impossible to convert between -wchar_t (as used by X11 and your applications) and any other encoding -without implementing OS-specific-wrappers for each and every locale. There -simply are no APIs to convert wchar_t into anything except the current -locale encoding.

    -

    Some applications (such as the formidable mlterm) work around this -by carrying their own replacement functions for character set handling -with them, and either implementing OS-dependent hacks or doing multiple -conversions (which is slow and unreliable in case the OS implements -encodings slightly different than the terminal emulator).

    -

    The rxvt-unicode author insists that the right way to fix this is in the -system libraries once and for all, instead of forcing every app to carry -complete replacements for them :)

    +

    Can I switch the fonts at runtime?

    +

    Yes, using an escape sequence. Try something like this, which has the same +effect as using the -fn switch, and takes effect immediately:

    +
    +   printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic"
    +

    This is useful if you e.g. work primarily with japanese (and prefer a +japanese font), but you have to switch to chinese temporarily, where +japanese fonts would only be in your way.

    +

    You can think of this as a kind of manual ISO-2022 switching.

    -

    I use Solaris 9 and it doesn't compile/work/etc.

    -

    Try the diff in doc/solaris9.patch as a base. It fixes the worst -problems with wcwidth and a compile problem.

    +

    Why do italic characters look as if clipped?

    +

    Many fonts have difficulties with italic characters and hinting. For +example, the otherwise very nicely hinted font xft:Bitstream Vera Sans +Mono completely fails in it's italic face. A workaround might be to +enable freetype autohinting, i.e. like this:

    +
    +   URxvt.italicFont:        xft:Bitstream Vera Sans Mono:italic:autohint=true
+   URxvt.boldItalicFont:    xft:Bitstream Vera Sans Mono:bold:italic:autohint=true

    -

    How can I use rxvt-unicode under cygwin?

    -

    rxvt-unicode should compile and run out of the box on cygwin, using -the X11 libraries that come with cygwin. libW11 emulation is no -longer supported (and makes no sense, either, as it only supported a -single font). I recommend starting the X-server in -multiwindow or --rootless mode instead, which will result in similar look&feel as the -old libW11 emulation.

    -

    At the time of this writing, cygwin didn't seem to support any multi-byte -encodings (you might try LC_CTYPE=C-UTF-8), so you are likely limited -to 8-bit encodings.

    -

    -

    -

    How does rxvt-unicode determine the encoding to use?

    -

    See next entry.

    -

    -

    -

    Is there an option to switch encodings?

    -

    Unlike some other terminals, rxvt-unicode has no encoding switch, and no -specific ``utf-8'' mode, such as xterm. In fact, it doesn't even know about -UTF-8 or any other encodings with respect to terminal I/O.

    -

    The reasons is that there exists a perfectly fine mechanism for selecting -the encoding, doing I/O and (most important) communicating this to all -applications so everybody agrees on character properties such as width -and code number. This mechanism is the locale. Applications not using -that info will have problems (for example, xterm gets the width of -characters wrong as it uses it's own, locale-independent table under all -locales).

    -

    Rxvt-unicode uses the LC_CTYPE locale category to select encoding. All -programs doing the same (that is, most) will automatically agree in the -interpretation of characters.

    -

    Unfortunately, there is no system-independent way to select locales, nor -is there a standard on how locale specifiers will look like.

    -

    On most systems, the content of the LC_CTYPE environment variable -contains an arbitrary string which corresponds to an already-installed -locale. Common names for locales are en_US.UTF-8, de_DE.ISO-8859-15, -ja_JP.EUC-JP, i.e. language_country.encoding, but other forms -(i.e. de or german) are also common.

    -

    Rxvt-unicode ignores all other locale categories, and except for -the encoding, ignores country or language-specific settings, -i.e. de_DE.UTF-8 and ja_JP.UTF-8 are the normally same to -rxvt-unicode.

    -

    If you want to use a specific encoding you have to make sure you start -rxvt-unicode with the correct LC_CTYPE category.

    -

    -

    -

    Can I switch locales at runtime?

    -

    Yes, using an escape sequence. Try something like this, which sets -rxvt-unicode's idea of LC_CTYPE.

    -
    -  printf '\e]701;%s\007' ja_JP.SJIS
    -

    See also the previous answer.

    -

    Sometimes this capability is rather handy when you want to work in -one locale (e.g. de_DE.UTF-8) but some programs don't support it -(e.g. UTF-8). For example, I use this script to start xjdic, which -first switches to a locale supported by xjdic and back later:

    -
    -   printf '\e]701;%s\007' ja_JP.SJIS
-   xjdic -js
-   printf '\e]701;%s\007' de_DE.UTF-8
    -

    You can also use xterm's luit program, which usually works fine, except -for some locales where character width differs between program- and -rxvt-unicode-locales.

    -

    -

    -

    Can I switch the fonts at runtime?

    -

    Yes, using an escape sequence. Try something like this, which has the same -effect as using the -fn switch, and takes effect immediately:

    -
    -   printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic"
    -

    This is useful if you e.g. work primarily with japanese (and prefer a -japanese font), but you have to switch to chinese temporarily, where -japanese fonts would only be in your way.

    -

    You can think of this as a kind of manual ISO-2022 switching.

    -

    -

    -

    Why do italic characters look as if clipped?

    -

    Many fonts have difficulties with italic characters and hinting. For -example, the otherwise very nicely hinted font xft:Bitstream Vera Sans -Mono completely fails in it's italic face. A workaround might be to -enable freetype autohinting, i.e. like this:

    -
    -   URxvt.italicFont:        xft:Bitstream Vera Sans Mono:italic:autohint=true
-   URxvt.boldItalicFont:    xft:Bitstream Vera Sans Mono:bold:italic:autohint=true
    -

    -

    -

    My input method wants <some encoding> but I want UTF-8, what can I do?

    -

    You can specify separate locales for the input method and the rest of the -terminal, using the resource imlocale:

    -
    -   URxvt.imlocale: ja_JP.EUC-JP
    -

    Now you can start your terminal with LC_CTYPE=ja_JP.UTF-8 and still -use your input method. Please note, however, that you will not be able to -input characters outside EUC-JP in a normal way then, as your input -method limits you.

    -

    -

    -

    Rxvt-unicode crashes when the X Input Method changes or exits.

    -

    Unfortunately, this is unavoidable, as the XIM protocol is racy by -design. Applications can avoid some crashes at the expense of memory -leaks, and Input Methods can avoid some crashes by careful ordering at -exit time. kinput2 (and derived input methods) generally succeeds, -while SCIM (or similar input methods) fails. In the end, however, -crashes cannot be completely avoided even if both sides cooperate.

    -

    So the only workaround is not to kill your Input Method Servers.

    -

    -

    -

    Rxvt-unicode uses gobs of memory, how can I reduce that?

    -

    Rxvt-unicode tries to obey the rule of not charging you for something you -don't use. One thing you should try is to configure out all settings that -you don't need, for example, Xft support is a resource hog by design, -when used. Compiling it out ensures that no Xft font will be loaded -accidentally when rxvt-unicode tries to find a font for your characters.

    -

    Also, many people (me included) like large windows and even larger -scrollback buffers: Without --enable-unicode3, rxvt-unicode will use -6 bytes per screen cell. For a 160x?? window this amounts to almost a -kilobyte per line. A scrollback buffer of 10000 lines will then (if full) -use 10 Megabytes of memory. With --enable-unicode3 it gets worse, as -rxvt-unicode then uses 8 bytes per screen cell.

    -

    -

    -

    Can I speed up Xft rendering somehow?

    +

    Can I speed up Xft rendering somehow?

    Yes, the most obvious way to speed it up is to avoid Xft entirely, as it is simply slow. If you still want Xft fonts you might try to disable antialiasing (by appending :antialias=false), which saves lots of memory and also speeds up rendering considerably.

    -

    Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?

    +

    Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?

    Rxvt-unicode will use whatever you specify as a font. If it needs to fall back to it's default font search list it will prefer X11 core fonts, because they are small and fast, and then use Xft fonts. It has @@ -795,15 +447,7 @@ look best that way.

    If you want antialiasing, you have to specify the fonts manually.

    -

    Mouse cut/paste suddenly no longer works.

    -

    Make sure that mouse reporting is actually turned off since killing -some editors prematurely may leave the mouse in mouse report mode. I've -heard that tcsh may use mouse reporting unless it otherwise specified. A -quick check is to see if cut/paste works when the Alt or Shift keys are -depressed.

    -

    -

    -

    What's with this bold/blink stuff?

    +

    What's with this bold/blink stuff?

    If no bold colour is set via colorBD:, bold will invert text using the standard foreground colour.

    For the standard background colour, blinking will actually make the @@ -816,7 +460,7 @@ foreground/background colors.

    color8-15 are the corresponding high-intensity colors.

    -

    I don't like the screen colors. How do I change them?

    +

    I don't like the screen colors. How do I change them?

    You can change the screen colors at run-time using ~/.Xdefaults resources (or as long-options).

    Here are values that are supposed to resemble a VGA screen, @@ -862,143 +506,547 @@ me) as ``pretty girly''.

    URxvt.color15: #e1dddd

    -

    How can I start rxvtd in a race-free way?

    -

    Try rxvtd -f -o, which tells rxvtd to open the -display, create the listening socket and then fork.

    +

    Why do some characters look so much different than others?

    +
    +
+See next entry.
    +
    +
+=head3 How does rxvt-unicode choose fonts?
    +
    +
+Most fonts do not contain the full range of Unicode, which is
+fine. Chances are that the font you (or the admin/package maintainer of
+your system/os) have specified does not cover all the characters you want
+to display.
    +
    +
+B<rxvt-unicode> makes a best-effort try at finding a replacement
+font. Often the result is fine, but sometimes the chosen font looks
+bad/ugly/wrong. Some fonts have totally strange characters that don't
+resemble the correct glyph at all, and rxvt-unicode lacks the artificial
+intelligence to detect that a specific glyph is wrong: it has to believe
+the font that the characters it claims to contain indeed look correct.
    +
    +
+In that case, select a font of your taste and add it to the font list,
+e.g.:
    +
    +
+   rxvt -fn basefont,font2,font3...
    +
    +
+When rxvt-unicode sees a character, it will first look at the base
+font. If the base font does not contain the character, it will go to the
+next font, and so on. Specifying your own fonts will also speed up this
+search and use less resources within rxvt-unicode and the X-server.
    +
    +
+The only limitation is that none of the fonts may be larger than the base
+font, as the base font defines the terminal character cell size, which
+must be the same due to the way terminals work.
    +

    +

    +

    Keyboard, Mouse & User Interaction

    +

    +

    +

    The new selection selects pieces that are too big, how can I select single words?

    +

    If you want to select e.g. alphanumeric words, you can use the following +setting:

    +
    +   URxvt.selection.pattern-0: ([[:word:]]+)
    +

    If you click more than twice, the selection will be extended +more and more.

    +

    To get a selection that is very similar to the old code, try this pattern:

    +
    +   URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+)
    +

    Please also note that the LeftClick Shift-LeftClik combination also +selects words like the old code.

    +

    +

    +

    I don't like the new selection/popups/hotkeys/perl, how do I change/disable it?

    +

    You can disable the perl extension completely by setting the +perl-ext-common resource to the empty string, which also keeps +rxvt-unicode from initialising perl, saving memory.

    +

    If you only want to disable specific features, you first have to +identify which perl extension is responsible. For this, read the section +PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For +example, to disable the selection-popup and option-popup, specify +this perl-ext-common resource:

    +
    +   URxvt.perl-ext-common: default,-selection-popup,-option-popup
    +

    This will keep the default extensions, but disable the two popup +extensions. Some extensions can also be configured, for example, +scrollback search mode is triggered by M-s. You can move it to any +other combination either by setting the searchable-scrollback resource:

    +
    +   URxvt.searchable-scrollback: CM-s
    +

    +

    +

    The cursor moves when selecting text in the current input line, how do I switch this off?

    +

    See next entry.

    +

    +

    +

    During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this?

    +

    These are caused by the readline perl extension. Under normal +circumstances, it will move your cursor around when you click into the +line that contains it. It tries hard not to do this at the wrong moment, +but when running a program that doesn't parse cursor movements or in some +cases during rlogin sessions, it fails to detect this properly.

    +

    You can permamently switch this feature off by disabling the readline +extension:

    +
    +   URxvt.perl-ext-common: default,-readline
    +

    +

    +

    My numerical keypad acts weird and generates differing output?

    +

    Some Debian GNUL/Linux users seem to have this problem, although no +specific details were reported so far. It is possible that this is caused +by the wrong TERM setting, although the details of wether and how +this can happen are unknown, as TERM=rxvt should offer a compatible +keymap. See the answer to the previous question, and please report if that +helped.

    +

    +

    +

    My Compose (Multi_key) key is no longer working.

    +

    The most common causes for this are that either your locale is not set +correctly, or you specified a preeditStyle that is not supported by +your input method. For example, if you specified OverTheSpot and +your input method (e.g. the default input method handling Compose keys) +does not support this (for instance because it is not visual), then +rxvt-unicode will continue without an input method.

    +

    In this case either do not specify a preeditStyle or specify more than +one pre-edit style, such as OverTheSpot,Root,None.

    +

    +

    +

    I cannot type Ctrl-Shift-2 to get an ASCII NUL character due to ISO 14755

    +

    Either try Ctrl-2 alone (it often is mapped to ASCII NUL even on +international keyboards) or simply use ISO 14755 support to your +advantage, typing <Ctrl-Shift-0> to get a ASCII NUL. This works for other +codes, too, such as Ctrl-Shift-1-d to type the default telnet escape +character and so on.

    +

    +

    +

    Mouse cut/paste suddenly no longer works.

    +

    Make sure that mouse reporting is actually turned off since killing +some editors prematurely may leave the mouse in mouse report mode. I've +heard that tcsh may use mouse reporting unless it otherwise specified. A +quick check is to see if cut/paste works when the Alt or Shift keys are +depressed.

    +

    +

    +

    What's with the strange Backspace/Delete key behaviour?

    +

    Assuming that the physical Backspace key corresponds to the +BackSpace keysym (not likely for Linux ... see the following +question) there are two standard values that can be used for +Backspace: ^H and ^?.

    +

    Historically, either value is correct, but rxvt-unicode adopts the debian +policy of using ^? when unsure, because it's the one only only correct +choice :).

    +

    Rxvt-unicode tries to inherit the current stty settings and uses the value +of `erase' to guess the value for backspace. If rxvt-unicode wasn't +started from a terminal (say, from a menu or by remote shell), then the +system value of `erase', which corresponds to CERASE in <termios.h>, will +be used (which may not be the same as your stty setting).

    +

    For starting a new rxvt-unicode:

    +
    +   # use Backspace = ^H
+   $ stty erase ^H
+   $ rxvt
    +
    +   # use Backspace = ^?
+   $ stty erase ^?
+   $ rxvt
    +

    Toggle with ESC [ 36 h / ESC [ 36 l.

    +

    For an existing rxvt-unicode:

    +
    +   # use Backspace = ^H
+   $ stty erase ^H
+   $ echo -n "^[[36h"
    +
    +   # use Backspace = ^?
+   $ stty erase ^?
+   $ echo -n "^[[36l"
    +

    This helps satisfy some of the Backspace discrepancies that occur, but +if you use Backspace = ^H, make sure that the termcap/terminfo value +properly reflects that.

    +

    The Delete key is a another casualty of the ill-defined Backspace problem. +To avoid confusion between the Backspace and Delete keys, the Delete +key has been assigned an escape sequence to match the vt100 for Execute +(ESC [ 3 ~) and is in the supplied termcap/terminfo.

    +

    Some other Backspace problems:

    +

    some editors use termcap/terminfo, +some editors (vim I'm told) expect Backspace = ^H, +GNU Emacs (and Emacs-like editors) use ^H for help.

    +

    Perhaps someday this will all be resolved in a consistent manner.

    +

    +

    +

    I don't like the key-bindings. How do I change them?

    +

    There are some compile-time selections available via configure. Unless +you have run ``configure'' with the --disable-resources option you can +use the `keysym' resource to alter the keystrings associated with keysyms.

    +

    Here's an example for a URxvt session started using rxvt -name URxvt

    +
    +   URxvt.keysym.Home:          \033[1~
+   URxvt.keysym.End:           \033[4~
+   URxvt.keysym.C-apostrophe:  \033<C-'>
+   URxvt.keysym.C-slash:       \033<C-/>
+   URxvt.keysym.C-semicolon:   \033<C-;>
+   URxvt.keysym.C-grave:       \033<C-`>
+   URxvt.keysym.C-comma:       \033<C-,>
+   URxvt.keysym.C-period:      \033<C-.>
+   URxvt.keysym.C-0x60:        \033<C-`>
+   URxvt.keysym.C-Tab:         \033<C-Tab>
+   URxvt.keysym.C-Return:      \033<C-Return>
+   URxvt.keysym.S-Return:      \033<S-Return>
+   URxvt.keysym.S-space:       \033<S-Space>
+   URxvt.keysym.M-Up:          \033<M-Up>
+   URxvt.keysym.M-Down:        \033<M-Down>
+   URxvt.keysym.M-Left:        \033<M-Left>
+   URxvt.keysym.M-Right:       \033<M-Right>
+   URxvt.keysym.M-C-0:         list \033<M-C- 0123456789 >
+   URxvt.keysym.M-C-a:         list \033<M-C- abcdefghijklmnopqrstuvwxyz >
+   URxvt.keysym.F12:           command:\033]701;zh_CN.GBK\007
    +

    See some more examples in the documentation for the keysym resource.

    +

    +

    +

    I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map

    +
    +   KP_Insert == Insert
+   F22 == Print
+   F27 == Home
+   F29 == Prior
+   F33 == End
+   F35 == Next
    +

    Rather than have rxvt-unicode try to accommodate all the various possible +keyboard mappings, it is better to use `xmodmap' to remap the keys as +required for your particular machine.

    +

    +

    +

    Terminal Configuration

    +

    +

    +

    Why doesn't rxvt-unicode read my resources?

    +

    Well, why, indeed? It does, in a way very similar to other X +applications. Most importantly, this means that if you or your OS loads +resources into the X display (the right way to do it), rxvt-unicode will +ignore any resource files in your home directory. It will only read +$HOME/.Xdefaults when no resources are attached to the display.

    +

    If you have or use an $HOME/.Xresources file, chances are that +resources are loaded into your X-server. In this case, you have to +re-login after every change (or run xrdb -merge $HOME/.Xresources).

    +

    Also consider the form resources have to use:

    +
    +  URxvt.resource: value
    +

    If you want to use another form (there are lots of different ways of +specifying resources), make sure you understand wether and why it +works. If unsure, use the form above.

    +

    +

    +

    When I log-in to another system it tells me about missing terminfo data?

    +

    The terminal description used by rxvt-unicode is not as widely available +as that for xterm, or even rxvt (for which the same problem often arises).

    +

    The correct solution for this problem is to install the terminfo, this can +be done like this (with ncurses' infocmp):

    +
    +   REMOTE=remotesystem.domain
+   infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti"
    +

    ... or by installing rxvt-unicode normally on the remote system,

    +

    If you cannot or do not want to do this, then you can simply set +TERM=rxvt or even TERM=xterm, and live with the small number of +problems arising, which includes wrong keymapping, less and different +colours and some refresh errors in fullscreen applications. It's a nice +quick-and-dirty workaround for rare cases, though.

    +

    If you always want to do this (and are fine with the consequences) you +can either recompile rxvt-unicode with the desired TERM value or use a +resource to set it:

    +
    +   URxvt.termName: rxvt
    +

    If you don't plan to use rxvt (quite common...) you could also replace +the rxvt terminfo file with the rxvt-unicode one.

    +

    +

    +

    tic outputs some error when compiling the terminfo entry.

    +

    Most likely it's the empty definition for enacs=. Just replace it by +enacs=\E[0@ and try again.

    +

    +

    +

    bash's readline does not work correctly under rxvt.

    +

    See next entry.

    +

    +

    +

    I need a termcap file entry.

    +

    One reason you might want this is that some distributions or operating +systems still compile some programs using the long-obsoleted termcap +library (Fedora Core's bash is one example) and rely on a termcap entry +for rxvt-unicode.

    +

    You could use rxvt's termcap entry with resonable results in many cases. +You can also create a termcap entry by using terminfo's infocmp program +like this:

    +
    +   infocmp -C rxvt-unicode
    +

    Or you could use this termcap entry, generated by the command above:

    +
    +   rxvt-unicode|rxvt-unicode terminal (X Window System):\
+           :am:bw:eo:km:mi:ms:xn:xo:\
+           :co#80:it#8:li#24:lm#0:\
+           :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\
+           :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\
+           :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\
+           :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\
+           :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\
+           :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\
+           :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\
+           :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\
+           :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\
+           :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\
+           :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\
+           :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\
+           :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\
+           :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\
+           :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\
+           :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\
+           :vs=\E[?25h:
    +

    +

    +

    Why does ls no longer have coloured output?

    +

    The ls in the GNU coreutils unfortunately doesn't use terminfo to +decide wether a terminal has colour, but uses it's own configuration +file. Needless to say, rxvt-unicode is not in it's default file (among +with most other terminals supporting colour). Either add:

    +
    +   TERM rxvt-unicode
    +

    to /etc/DIR_COLORS or simply add:

    +
    +   alias ls='ls --color=auto'
    +

    to your .profile or .bashrc.

    +

    +

    +

    Why doesn't vim/emacs etc. use the 88 colour mode?

    +

    See next entry.

    +

    +

    +

    Why doesn't vim/emacs etc. make use of italic?

    +

    See next entry.

    +

    +

    +

    Why are the secondary screen-related options not working properly?

    +

    Make sure you are using TERM=rxvt-unicode. Some pre-packaged +distributions (most notably Debian GNU/Linux) break rxvt-unicode +by setting TERM to rxvt, which doesn't have these extra +features. Unfortunately, some of these (most notably, again, Debian +GNU/Linux) furthermore fail to even install the rxvt-unicode terminfo +file, so you will need to install it on your own (See the question When +I log-in to another system it tells me about missing terminfo data? on +how to do this).

    +

    +

    +

    Encoding / Locale / Input Method Issues

    +

    +

    +

    Rxvt-unicode does not seem to understand the selected encoding?

    +

    See next entry.

    +

    +

    +

    Unicode does not seem to work?

    +

    If you encounter strange problems like typing an accented character but +getting two unrelated other characters or similar, or if program output is +subtly garbled, then you should check your locale settings.

    +

    Rxvt-unicode must be started with the same LC_CTYPE setting as the +programs. Often rxvt-unicode is started in the C locale, while the +login script running within the rxvt-unicode window changes the locale to +something else, e.g. en_GB.UTF-8. Needless to say, this is not going to work.

    +

    The best thing is to fix your startup environment, as you will likely run +into other problems. If nothing works you can try this in your .profile.

    +
    +  printf '\e]701;%s\007' "$LC_CTYPE"
    +

    If this doesn't work, then maybe you use a LC_CTYPE specification not +supported on your systems. Some systems have a locale command which +displays this (also, perl -e0 can be used to check locale settings, as +it will complain loudly if it cannot set the locale). If it displays something +like:

    +
    +  locale: Cannot set LC_CTYPE to default locale: ...
    +

    Then the locale you specified is not supported on your system.

    +

    If nothing works and you are sure that everything is set correctly then +you will need to remember a little known fact: Some programs just don't +support locales :(

    +

    +

    +

    How does rxvt-unicode determine the encoding to use?

    +

    See next entry.

    +

    +

    +

    Is there an option to switch encodings?

    +

    Unlike some other terminals, rxvt-unicode has no encoding switch, and no +specific ``utf-8'' mode, such as xterm. In fact, it doesn't even know about +UTF-8 or any other encodings with respect to terminal I/O.

    +

    The reasons is that there exists a perfectly fine mechanism for selecting +the encoding, doing I/O and (most important) communicating this to all +applications so everybody agrees on character properties such as width +and code number. This mechanism is the locale. Applications not using +that info will have problems (for example, xterm gets the width of +characters wrong as it uses it's own, locale-independent table under all +locales).

    +

    Rxvt-unicode uses the LC_CTYPE locale category to select encoding. All +programs doing the same (that is, most) will automatically agree in the +interpretation of characters.

    +

    Unfortunately, there is no system-independent way to select locales, nor +is there a standard on how locale specifiers will look like.

    +

    On most systems, the content of the LC_CTYPE environment variable +contains an arbitrary string which corresponds to an already-installed +locale. Common names for locales are en_US.UTF-8, de_DE.ISO-8859-15, +ja_JP.EUC-JP, i.e. language_country.encoding, but other forms +(i.e. de or german) are also common.

    +

    Rxvt-unicode ignores all other locale categories, and except for +the encoding, ignores country or language-specific settings, +i.e. de_DE.UTF-8 and ja_JP.UTF-8 are the normally same to +rxvt-unicode.

    +

    If you want to use a specific encoding you have to make sure you start +rxvt-unicode with the correct LC_CTYPE category.

    -

    What's with the strange Backspace/Delete key behaviour?

    -

    Assuming that the physical Backspace key corresponds to the -BackSpace keysym (not likely for Linux ... see the following -question) there are two standard values that can be used for -Backspace: ^H and ^?.

    -

    Historically, either value is correct, but rxvt-unicode adopts the debian -policy of using ^? when unsure, because it's the one only only correct -choice :).

    -

    Rxvt-unicode tries to inherit the current stty settings and uses the value -of `erase' to guess the value for backspace. If rxvt-unicode wasn't -started from a terminal (say, from a menu or by remote shell), then the -system value of `erase', which corresponds to CERASE in <termios.h>, will -be used (which may not be the same as your stty setting).

    -

    For starting a new rxvt-unicode:

    -
    -   # use Backspace = ^H
-   $ stty erase ^H
-   $ rxvt
    -
    -   # use Backspace = ^?
-   $ stty erase ^?
-   $ rxvt
    -

    Toggle with ESC [ 36 h / ESC [ 36 l.

    -

    For an existing rxvt-unicode:

    +

    Can I switch locales at runtime?

    +

    Yes, using an escape sequence. Try something like this, which sets +rxvt-unicode's idea of LC_CTYPE.

    -   # use Backspace = ^H
-   $ stty erase ^H
-   $ echo -n "^[[36h"
    + printf '\e]701;%s\007' ja_JP.SJIS +

    See also the previous answer.

    +

    Sometimes this capability is rather handy when you want to work in +one locale (e.g. de_DE.UTF-8) but some programs don't support it +(e.g. UTF-8). For example, I use this script to start xjdic, which +first switches to a locale supported by xjdic and back later:

    -   # use Backspace = ^?
-   $ stty erase ^?
-   $ echo -n "^[[36l"
    -

    This helps satisfy some of the Backspace discrepancies that occur, but -if you use Backspace = ^H, make sure that the termcap/terminfo value -properly reflects that.

    -

    The Delete key is a another casualty of the ill-defined Backspace problem. -To avoid confusion between the Backspace and Delete keys, the Delete -key has been assigned an escape sequence to match the vt100 for Execute -(ESC [ 3 ~) and is in the supplied termcap/terminfo.

    -

    Some other Backspace problems:

    -

    some editors use termcap/terminfo, -some editors (vim I'm told) expect Backspace = ^H, -GNU Emacs (and Emacs-like editors) use ^H for help.

    -

    Perhaps someday this will all be resolved in a consistent manner.

    + printf '\e]701;%s\007' ja_JP.SJIS + xjdic -js + printf '\e]701;%s\007' de_DE.UTF-8 +

    You can also use xterm's luit program, which usually works fine, except +for some locales where character width differs between program- and +rxvt-unicode-locales.

    -

    I don't like the key-bindings. How do I change them?

    -

    There are some compile-time selections available via configure. Unless -you have run ``configure'' with the --disable-resources option you can -use the `keysym' resource to alter the keystrings associated with keysyms.

    -

    Here's an example for a URxvt session started using rxvt -name URxvt

    +

    My input method wants <some encoding> but I want UTF-8, what can I do?

    +

    You can specify separate locales for the input method and the rest of the +terminal, using the resource imlocale:

    -   URxvt.keysym.Home:          \033[1~
-   URxvt.keysym.End:           \033[4~
-   URxvt.keysym.C-apostrophe:  \033<C-'>
-   URxvt.keysym.C-slash:       \033<C-/>
-   URxvt.keysym.C-semicolon:   \033<C-;>
-   URxvt.keysym.C-grave:       \033<C-`>
-   URxvt.keysym.C-comma:       \033<C-,>
-   URxvt.keysym.C-period:      \033<C-.>
-   URxvt.keysym.C-0x60:        \033<C-`>
-   URxvt.keysym.C-Tab:         \033<C-Tab>
-   URxvt.keysym.C-Return:      \033<C-Return>
-   URxvt.keysym.S-Return:      \033<S-Return>
-   URxvt.keysym.S-space:       \033<S-Space>
-   URxvt.keysym.M-Up:          \033<M-Up>
-   URxvt.keysym.M-Down:        \033<M-Down>
-   URxvt.keysym.M-Left:        \033<M-Left>
-   URxvt.keysym.M-Right:       \033<M-Right>
-   URxvt.keysym.M-C-0:         list \033<M-C- 0123456789 >
-   URxvt.keysym.M-C-a:         list \033<M-C- abcdefghijklmnopqrstuvwxyz >
-   URxvt.keysym.F12:           command:\033]701;zh_CN.GBK\007
    -

    See some more examples in the documentation for the keysym resource.

    + URxvt.imlocale: ja_JP.EUC-JP +

    Now you can start your terminal with LC_CTYPE=ja_JP.UTF-8 and still +use your input method. Please note, however, that you will not be able to +input characters outside EUC-JP in a normal way then, as your input +method limits you.

    -

    I'm using keyboard model XXX that has extra Prior/Next/Insert keys. -How do I make use of them? For example, the Sun Keyboard type 4 -has the following mappings that rxvt-unicode doesn't recognize.

    -
    -   KP_Insert == Insert
-   F22 == Print
-   F27 == Home
-   F29 == Prior
-   F33 == End
-   F35 == Next
    -

    Rather than have rxvt-unicode try to accommodate all the various possible -keyboard mappings, it is better to use `xmodmap' to remap the keys as -required for your particular machine.

    +

    Rxvt-unicode crashes when the X Input Method changes or exits.

    +

    Unfortunately, this is unavoidable, as the XIM protocol is racy by +design. Applications can avoid some crashes at the expense of memory +leaks, and Input Methods can avoid some crashes by careful ordering at +exit time. kinput2 (and derived input methods) generally succeeds, +while SCIM (or similar input methods) fails. In the end, however, +crashes cannot be completely avoided even if both sides cooperate.

    +

    So the only workaround is not to kill your Input Method Servers.

    -

    How do I distinguish wether I'm running rxvt-unicode or a regular xterm? -I need this to decide about setting colors etc.

    -

    rxvt and rxvt-unicode always export the variable ``COLORTERM'', so you can -check and see if that is set. Note that several programs, JED, slrn, -Midnight Commander automatically check this variable to decide whether or -not to use color.

    +

    Operating Systems / Package Maintaining

    -

    How do I set the correct, full IP address for the DISPLAY variable?

    -

    If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled -insecure mode then it is possible to use the following shell script -snippets to correctly set the display. If your version of rxvt-unicode -wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) then -the COLORTERM variable can be used to distinguish rxvt-unicode from a -regular xterm.

    -

    Courtesy of Chuck Blake <cblake@BBN.COM> with the following shell script -snippets:

    +

    I am using Debian GNU/Linux and have a problem...

    +

    The Debian GNU/Linux package of rxvt-unicode in sarge contains large +patches that considerably change the behaviour of rxvt-unicode (but +unfortunately this notice has been removed). Before reporting a bug to +the original rxvt-unicode author please download and install the genuine +version (http://software.schmorp.de#rxvt-unicode) and try to reproduce +the problem. If you cannot, chances are that the problems are specific to +Debian GNU/Linux, in which case it should be reported via the Debian Bug +Tracking System (use reportbug to report the bug).

    +

    For other problems that also affect the Debian package, you can and +probably should use the Debian BTS, too, because, after all, it's also a +bug in the Debian version and it serves as a reminder for other users that +might encounter the same issue.

    +

    +

    +

    I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation?

    +

    You should build one binary with the default options. configure +now enables most useful options, and the trend goes to making them +runtime-switchable, too, so there is usually no drawback to enbaling them, +except higher disk and possibly memory usage. The perl interpreter should +be enabled, as important functionality (menus, selection, likely more in +the future) depends on it.

    +

    You should not overwrite the perl-ext-common snd perl-ext resources +system-wide (except maybe with defaults). This will result in useful +behaviour. If your distribution aims at low memory, add an empty +perl-ext-common resource to the app-defaults file. This will keep the +perl interpreter disabled until the user enables it.

    +

    If you can/want build more binaries, I recommend building a minimal +one with --disable-everything (very useful) and a maximal one with +--enable-everything (less useful, it will be very big due to a lot of +encodings built-in that increase download times and are rarely used).

    +

    +

    +

    I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe?

    +

    It should be, starting with release 7.1. You are encouraged to properly +install urxvt with privileges necessary for your OS now.

    +

    When rxvt-unicode detects that it runs setuid or setgid, it will fork +into a helper process for privileged operations (pty handling on some +systems, utmp/wtmp/lastlog handling on others) and drop privileges +immediately. This is much safer than most other terminals that keep +privileges while running (but is more relevant to urxvt, as it contains +things as perl interpreters, which might be ``helpful'' to attackers).

    +

    This forking is done as the very first within main(), which is very early +and reduces possible bugs to initialisation code run before main(), or +things like the dynamic loader of your system, which should result in very +little risk.

    +

    +

    +

    On Solaris 9, many line-drawing characters are too wide.

    +

    Seems to be a known bug, read +http://nixdoc.net/files/forum/about34198.html. Some people use the +following ugly workaround to get non-double-wide-characters working:

    -   # Bourne/Korn/POSIX family of shells:
-   [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know
-   if [ ${TERM:-foo} = xterm ]; then
-      stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not
-      echo -n '^[Z'
-      read term_id
-      stty icanon echo
-      if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then
-         echo -n '^[[7n'        # query the rxvt we are in for the DISPLAY string
-         read DISPLAY           # set it in our local shell
-      fi
-   fi
    + #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x)

    -

    How do I compile the manual pages for myself?

    -

    You need to have a recent version of perl installed as /usr/bin/perl, -one that comes with pod2man, pod2text and pod2html. Then go to -the doc subdirectory and enter make alldoc.

    +

    I am on FreeBSD and rxvt-unicode does not seem to work at all.

    +

    Rxvt-unicode requires the symbol __STDC_ISO_10646__ to be defined +in your compile environment, or an implementation that implements it, +wether it defines the symbol or not. __STDC_ISO_10646__ requires that +wchar_t is represented as unicode.

    +

    As you might have guessed, FreeBSD does neither define this symobl nor +does it support it. Instead, it uses it's own internal representation of +wchar_t. This is, of course, completely fine with respect to standards.

    +

    However, that means rxvt-unicode only works in POSIX, ISO-8859-1 and +UTF-8 locales under FreeBSD (which all use Unicode as wchar_t.

    +

    __STDC_ISO_10646__ is the only sane way to support multi-language +apps in an OS, as using a locale-dependent (and non-standardized) +representation of wchar_t makes it impossible to convert between +wchar_t (as used by X11 and your applications) and any other encoding +without implementing OS-specific-wrappers for each and every locale. There +simply are no APIs to convert wchar_t into anything except the current +locale encoding.

    +

    Some applications (such as the formidable mlterm) work around this +by carrying their own replacement functions for character set handling +with them, and either implementing OS-dependent hacks or doing multiple +conversions (which is slow and unreliable in case the OS implements +encodings slightly different than the terminal emulator).

    +

    The rxvt-unicode author insists that the right way to fix this is in the +system libraries once and for all, instead of forcing every app to carry +complete replacements for them :)

    -

    My question isn't answered here, can I ask a human?

    -

    Before sending me mail, you could go to IRC: irc.freenode.net, -channel #rxvt-unicode has some rxvt-unicode enthusiasts that might be -interested in learning about new and exciting problems (but not FAQs :).

    +

    I use Solaris 9 and it doesn't compile/work/etc.

    +

    Try the diff in doc/solaris9.patch as a base. It fixes the worst +problems with wcwidth and a compile problem.

    +

    +

    +

    How can I use rxvt-unicode under cygwin?

    +

    rxvt-unicode should compile and run out of the box on cygwin, using +the X11 libraries that come with cygwin. libW11 emulation is no +longer supported (and makes no sense, either, as it only supported a +single font). I recommend starting the X-server in -multiwindow or +-rootless mode instead, which will result in similar look&feel as the +old libW11 emulation.

    +

    At the time of this writing, cygwin didn't seem to support any multi-byte +encodings (you might try LC_CTYPE=C-UTF-8), so you are likely limited +to 8-bit encodings.

    @@ -2380,18 +2428,24 @@ Use the X Window System (pretty much default, eh?).
    Look for the XPM includes in DIR. + +

    --with-xpm-library=DIR
    Look for the XPM library in DIR. + +

    --with-xpm
    Not needed - define via --enable-xpm-background. + +

    @@ -2401,7 +2455,9 @@ Not needed - define via --enable-xpm-background.

    Marc Lehmann <rxvt@schmorp.de> converted this document to pod and reworked it from the original Rxvt documentation, which was done by Geoff Wing <gcw@pobox.com>, who in turn used the XTerm documentation and other -sources.

    +sources. + +

    diff --git a/doc/rxvt.7.man.in b/doc/rxvt.7.man.in index af701a25..7a3b4999 100644 --- a/doc/rxvt.7.man.in +++ b/doc/rxvt.7.man.in @@ -157,145 +157,112 @@ The newest version of this document is also available on the World Wide Web at . .SH "FREQUENTLY ASKED QUESTIONS" .IX Header "FREQUENTLY ASKED QUESTIONS" -.Sh "The new selection selects pieces that are too big, how can I select single words?" -.IX Subsection "The new selection selects pieces that are too big, how can I select single words?" -Yes. For example, if you want to select alphanumeric words, you can use -the following resource: +.Sh "Meta, Features & Commandline Issues" +.IX Subsection "Meta, Features & Commandline Issues" +\fIMy question isn't answered here, can I ask a human?\fR +.IX Subsection "My question isn't answered here, can I ask a human?" .PP -.Vb 1 -\& URxvt.selection.pattern-0: ([[:word:]]+) -.Ve +Before sending me mail, you could go to \s-1IRC:\s0 \f(CW\*(C`irc.freenode.net\*(C'\fR, +channel \f(CW\*(C`#rxvt\-unicode\*(C'\fR has some rxvt-unicode enthusiasts that might be +interested in learning about new and exciting problems (but not FAQs :). .PP -If you click more than twice, the selection will be extended -more and more. +\fIDoes it support tabs, can I have a tabbed rxvt\-unicode?\fR +.IX Subsection "Does it support tabs, can I have a tabbed rxvt-unicode?" .PP -To get a selection that is very similar to the old code, try this pattern: +Beginning with version 7.3, there is a perl extension that implements a +simple tabbed terminal. It is installed by default, so any of these should +give you tabs: .PP .Vb 1 -\& URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\e\e\e\e]^`{|})]+) +\& @@RXVT_NAME@@ -pe tabbed .Ve .PP -Please also note that the \fILeftClick Shift-LeftClik\fR combination also -selects words like the old code. -.IP "I don't like the new selection/popups/hotkeys/perl, how do I change/disable it?" 4 -.IX Item "I don't like the new selection/popups/hotkeys/perl, how do I change/disable it?" -You can disable the perl extension completely by setting the -\&\fBperl-ext-common\fR resource to the empty string, which also keeps -rxvt-unicode from initialising perl, saving memory. -.Sp -If you only want to disable specific features, you first have to -identify which perl extension is responsible. For this, read the section -\&\fB\s-1PREPACKAGED\s0 \s-1EXTENSIONS\s0\fR in the @@RXVT_NAME@@\fIperl\fR\|(3) manpage. For -example, to disable the \fBselection-popup\fR and \fBoption-popup\fR, specify -this \fBperl-ext-common\fR resource: -.Sp -.Vb 1 -\& URxvt.perl-ext-common: default,-selection-popup,-option-popup -.Ve -.Sp -This will keep the default extensions, but disable the two popup -extensions. Some extensions can also be configured, for example, -scrollback search mode is triggered by \fBM\-s\fR. You can move it to any -other combination either by setting the \fBsearchable-scrollback\fR resource: -.Sp -.Vb 1 -\& URxvt.searchable-scrollback: CM-s -.Ve -.IP "The cursor moves when selecting text in the current input line, how do I switch this off?" 4 -.IX Item "The cursor moves when selecting text in the current input line, how do I switch this off?" -.PD 0 -.IP "During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this?" 4 -.IX Item "During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this?" -.PD -These are caused by the \f(CW\*(C`readline\*(C'\fR perl extension. Under normal -circumstances, it will move your cursor around when you click into the -line that contains it. It tries hard not to do this at the wrong moment, -but when running a program that doesn't parse cursor movements or in some -cases during rlogin sessions, it fails to detect this properly. -.Sp -You can permamently switch this feature off by disabling the \f(CW\*(C`readline\*(C'\fR -extension: -.Sp -.Vb 1 -\& URxvt.perl-ext-common: default,-readline -.Ve -.IP "Why doesn't rxvt-unicode read my resources?" 4 -.IX Item "Why doesn't rxvt-unicode read my resources?" -Well, why, indeed? It does, in a way very similar to other X -applications. Most importantly, this means that if you or your \s-1OS\s0 loads -resources into the X display (the right way to do it), rxvt-unicode will -ignore any resource files in your home directory. It will only read -\&\fI$HOME/.Xdefaults\fR when no resources are attached to the display. -.Sp -If you have or use an \fI$HOME/.Xresources\fR file, chances are that -resources are loaded into your X\-server. In this case, you have to -re-login after every change (or run \fIxrdb \-merge \f(CI$HOME\fI/.Xresources\fR). -.Sp -Also consider the form resources have to use: -.Sp -.Vb 1 -\& URxvt.resource: value -.Ve -.Sp -If you want to use another form (there are lots of different ways of -specifying resources), make sure you understand wether and why it -works. If unsure, use the form above. -.IP "I can't get transparency working, what am I doing wrong?" 4 -.IX Item "I can't get transparency working, what am I doing wrong?" -First of all, transparency isn't officially supported in rxvt\-unicode, so -you are mostly on your own. Do not bug the author about it (but you may -bug everybody else). Also, if you can't get it working consider it a rite -of passage: ... and you failed. -.Sp -Here are four ways to get transparency. \fBDo\fR read the manpage and option -descriptions for the programs mentioned and rxvt\-unicode. Really, do it! -.Sp -1. Use inheritPixmap: -.Sp -.Vb 2 -\& Esetroot wallpaper.jpg -\& @@RXVT_NAME@@ -ip -tint red -sh 40 -.Ve -.Sp -That works. If you think it doesn't, you lack transparency and tinting -support, or you are unable to read. -.Sp -2. Use a simple pixmap and emulate pseudo\-transparency. This enables you -to use effects other than tinting and shading: Just shade/tint/whatever -your picture with gimp: -.Sp -.Vb 2 -\& convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm -\& @@RXVT_NAME@@ -pixmap background.xpm -pe automove-background -.Ve -.Sp -That works. If you think it doesn't, you lack \s-1XPM\s0 and Perl support, or you -are unable to read. -.Sp -3. Use an \s-1ARGB\s0 visual: -.Sp .Vb 1 -\& @@RXVT_NAME@@ -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc +\& URxvt.perl-ext-common: default,tabbed .Ve -.Sp -This requires \s-1XFT\s0 support, and the support of your X\-server. If that -doesn't work for you, blame Xorg and Keith Packard. \s-1ARGB\s0 visuals aren't -there yet, no matter what they claim. Rxvt-Unicode contains the neccessary -bugfixes and workarounds for Xft and Xlib to make it work, but that -doesn't mean that your \s-1WM\s0 has the required kludges in place. -.Sp -4. Use xcompmgr and let it do the job: -.Sp -.Vb 2 -\& xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \e -\& -set _NET_WM_WINDOW_OPACITY 0xc0000000 +.PP +It will also work fine with tabbing functionality of many window managers +or similar tabbing programs, and its embedding-features allow it to be +embedded into other programs, as witnessed by \fIdoc/rxvt\-tabbed\fR or +the upcoming \f(CW\*(C`Gtk2::URxvt\*(C'\fR perl module, which features a tabbed urxvt +(murxvt) terminal as an example embedding application. +.PP +\fIHow do I know which rxvt-unicode version I'm using?\fR +.IX Subsection "How do I know which rxvt-unicode version I'm using?" +.PP +The version number is displayed with the usage (\-h). Also the escape +sequence \f(CW\*(C`ESC [ 8 n\*(C'\fR sets the window title to the version number. When +using the @@RXVT_NAME@@c client, the version displayed is that of the +daemon. +.PP +\fIRxvt-unicode uses gobs of memory, how can I reduce that?\fR +.IX Subsection "Rxvt-unicode uses gobs of memory, how can I reduce that?" +.PP +Rxvt-unicode tries to obey the rule of not charging you for something you +don't use. One thing you should try is to configure out all settings that +you don't need, for example, Xft support is a resource hog by design, +when used. Compiling it out ensures that no Xft font will be loaded +accidentally when rxvt-unicode tries to find a font for your characters. +.PP +Also, many people (me included) like large windows and even larger +scrollback buffers: Without \f(CW\*(C`\-\-enable\-unicode3\*(C'\fR, rxvt-unicode will use +6 bytes per screen cell. For a 160x?? window this amounts to almost a +kilobyte per line. A scrollback buffer of 10000 lines will then (if full) +use 10 Megabytes of memory. With \f(CW\*(C`\-\-enable\-unicode3\*(C'\fR it gets worse, as +rxvt-unicode then uses 8 bytes per screen cell. +.PP +\fIHow can I start @@RXVT_NAME@@d in a race-free way?\fR +.IX Subsection "How can I start @@RXVT_NAME@@d in a race-free way?" +.PP +Try \f(CW\*(C`@@RXVT_NAME@@d \-f \-o\*(C'\fR, which tells @@RXVT_NAME@@d to open the +display, create the listening socket and then fork. +.PP +\fIHow do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc.\fR +.IX Subsection "How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc." +.PP +rxvt and rxvt-unicode always export the variable \*(L"\s-1COLORTERM\s0\*(R", so you can +check and see if that is set. Note that several programs, \s-1JED\s0, slrn, +Midnight Commander automatically check this variable to decide whether or +not to use color. +.PP +\fIHow do I set the correct, full \s-1IP\s0 address for the \s-1DISPLAY\s0 variable?\fR +.IX Subsection "How do I set the correct, full IP address for the DISPLAY variable?" +.PP +If you've compiled rxvt-unicode with \s-1DISPLAY_IS_IP\s0 and have enabled +insecure mode then it is possible to use the following shell script +snippets to correctly set the display. If your version of rxvt-unicode +wasn't also compiled with \s-1ESCZ_ANSWER\s0 (as assumed in these snippets) then +the \s-1COLORTERM\s0 variable can be used to distinguish rxvt-unicode from a +regular xterm. +.PP +Courtesy of Chuck Blake with the following shell script +snippets: +.PP +.Vb 12 +\& # Bourne/Korn/POSIX family of shells: +\& [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know +\& if [ ${TERM:-foo} = xterm ]; then +\& stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not +\& echo -n '^[Z' +\& read term_id +\& stty icanon echo +\& if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then +\& echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string +\& read DISPLAY # set it in our local shell +\& fi +\& fi .Ve -.Sp -Then click on a window you want to make transparent. Replace \f(CW0xc0000000\fR -by other values to change the degree of opacity. If it doesn't work and -your server crashes, you got to keep the pieces. -.IP "Isn't rxvt supposed to be small? Don't all those features bloat?" 4 -.IX Item "Isn't rxvt supposed to be small? Don't all those features bloat?" +.PP +\fIHow do I compile the manual pages on my own?\fR +.IX Subsection "How do I compile the manual pages on my own?" +.PP +You need to have a recent version of perl installed as \fI/usr/bin/perl\fR, +one that comes with \fIpod2man\fR, \fIpod2text\fR and \fIpod2html\fR. Then go to +the doc subdirectory and enter \f(CW\*(C`make alldoc\*(C'\fR. +.PP +\fIIsn't rxvt-unicode supposed to be small? Don't all those features bloat?\fR +.IX Subsection "Isn't rxvt-unicode supposed to be small? Don't all those features bloat?" +.PP I often get asked about this, and I think, no, they didn't cause extra bloat. If you compare a minimal rxvt and a minimal urxvt, you can see that the urxvt binary is larger (due to some encoding tables always being @@ -303,23 +270,23 @@ compiled in), but it actually uses less memory (\s-1RSS\s0) after startup. Even with \f(CW\*(C`\-\-disable\-everything\*(C'\fR, this comparison is a bit unfair, as many features unique to urxvt (locale, encoding conversion, iso14755 etc.) are already in use in this mode. -.Sp +.PP .Vb 3 \& text data bss drs rss filename \& 98398 1664 24 15695 1824 rxvt --disable-everything \& 188985 9048 66616 18222 1788 urxvt --disable-everything .Ve -.Sp +.PP When you \f(CW\*(C`\-\-enable\-everything\*(C'\fR (which _is_ unfair, as this involves xft and full locale/XIM support which are quite bloaty inside libX11 and my libc), the two diverge, but not unreasnobaly so. -.Sp +.PP .Vb 3 \& text data bss drs rss filename \& 163431 2152 24 20123 2060 rxvt --enable-everything \& 1035683 49680 66648 29096 3680 urxvt --enable-everything .Ve -.Sp +.PP The very large size of the text section is explained by the east-asian encoding tables, which, if unused, take up disk space but nothing else and can be compiled out unless you rely on X11 core fonts that use those @@ -328,44 +295,46 @@ compiler allocates (but of course doesn't use unless you are out of memory). Also, using an xft font instead of a core font immediately adds a few megabytes of \s-1RSS\s0. Xft indeed is responsible for a lot of \s-1RSS\s0 even when not used. -.Sp +.PP Of course, due to every character using two or four bytes instead of one, a large scrollback buffer will ultimately make rxvt-unicode use more memory. -.Sp +.PP Compared to e.g. Eterm (5112k), aterm (3132k) and xterm (4680k), this still fares rather well. And compared to some monsters like gnome-terminal (21152k + extra 4204k in separate processes) or konsole (22200k + extra 43180k in daemons that stay around after exit, plus half a minute of startup time, including the hundreds of warnings it spits out), it fares extremely well *g*. -.IP "Why \*(C+, isn't that unportable/bloated/uncool?" 4 -.IX Item "Why , isn't that unportable/bloated/uncool?" +.PP +\fIWhy \*(C+, isn't that unportable/bloated/uncool?\fR +.IX Subsection "Why , isn't that unportable/bloated/uncool?" +.PP Is this a question? :) It comes up very often. The simple answer is: I had to write it, and \*(C+ allowed me to write and maintain it in a fraction of the time and effort (which is a scarce resource for me). Put even shorter: It simply wouldn't exist without \*(C+. -.Sp +.PP My personal stance on this is that \*(C+ is less portable than C, but in the case of rxvt-unicode this hardly matters, as its portability limits are defined by things like X11, pseudo terminals, locale support and unix domain sockets, which are all less portable than \*(C+ itself. -.Sp +.PP Regarding the bloat, see the above question: It's easy to write programs in C that use gobs of memory, an certainly possible to write programs in \&\*(C+ that don't. \*(C+ also often comes with large libraries, but this is not necessarily the case with \s-1GCC\s0. Here is what rxvt links against on my system with a minimal config: -.Sp +.PP .Vb 4 \& libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) \& libc.so.6 => /lib/libc.so.6 (0x00002aaaaadde000) \& libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab01d000) \& /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) .Ve -.Sp +.PP And here is rxvt\-unicode: -.Sp +.PP .Vb 5 \& libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) \& libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x00002aaaaada2000) @@ -373,279 +342,70 @@ And here is rxvt\-unicode: \& libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab0ee000) \& /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) .Ve -.Sp +.PP No large bloated libraries (of course, none were linked in statically), except maybe libX11 :) -.IP "Does it support tabs, can I have a tabbed rxvt\-unicode?" 4 -.IX Item "Does it support tabs, can I have a tabbed rxvt-unicode?" -Beginning with version 7.3, there is a perl extension that implements a -simple tabbed terminal. It is installed by default, so any of these should -give you tabs: -.Sp -.Vb 1 -\& @@RXVT_NAME@@ -pe tabbed +.Sh "Rendering, Font & Look and Feel Issues" +.IX Subsection "Rendering, Font & Look and Feel Issues" +\fII can't get transparency working, what am I doing wrong?\fR +.IX Subsection "I can't get transparency working, what am I doing wrong?" +.PP +First of all, transparency isn't officially supported in rxvt\-unicode, so +you are mostly on your own. Do not bug the author about it (but you may +bug everybody else). Also, if you can't get it working consider it a rite +of passage: ... and you failed. +.PP +Here are four ways to get transparency. \fBDo\fR read the manpage and option +descriptions for the programs mentioned and rxvt\-unicode. Really, do it! +.PP +1. Use inheritPixmap: +.PP +.Vb 2 +\& Esetroot wallpaper.jpg +\& @@RXVT_NAME@@ -ip -tint red -sh 40 .Ve -.Sp +.PP +That works. If you think it doesn't, you lack transparency and tinting +support, or you are unable to read. +.PP +2. Use a simple pixmap and emulate pseudo\-transparency. This enables you +to use effects other than tinting and shading: Just shade/tint/whatever +your picture with gimp or any other tool: +.PP +.Vb 2 +\& convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm +\& @@RXVT_NAME@@ -pixmap background.xpm -pe automove-background +.Ve +.PP +That works. If you think it doesn't, you lack \s-1XPM\s0 and Perl support, or you +are unable to read. +.PP +3. Use an \s-1ARGB\s0 visual: +.PP .Vb 1 -\& URxvt.perl-ext-common: default,tabbed +\& @@RXVT_NAME@@ -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc .Ve -.Sp -It will also work fine with tabbing functionality of many window managers -or similar tabbing programs, and its embedding-features allow it to be -embedded into other programs, as witnessed by \fIdoc/rxvt\-tabbed\fR or -the upcoming \f(CW\*(C`Gtk2::URxvt\*(C'\fR perl module, which features a tabbed urxvt -(murxvt) terminal as an example embedding application. -.IP "How do I know which rxvt-unicode version I'm using?" 4 -.IX Item "How do I know which rxvt-unicode version I'm using?" -The version number is displayed with the usage (\-h). Also the escape -sequence \f(CW\*(C`ESC [ 8 n\*(C'\fR sets the window title to the version number. When -using the @@RXVT_NAME@@c client, the version displayed is that of the -daemon. -.IP "I am using Debian GNU/Linux and have a problem..." 4 -.IX Item "I am using Debian GNU/Linux and have a problem..." -The Debian GNU/Linux package of rxvt-unicode in sarge contains large -patches that considerably change the behaviour of rxvt-unicode (but -unfortunately this notice has been removed). Before reporting a bug to -the original rxvt-unicode author please download and install the genuine -version () and try to reproduce -the problem. If you cannot, chances are that the problems are specific to -Debian GNU/Linux, in which case it should be reported via the Debian Bug -Tracking System (use \f(CW\*(C`reportbug\*(C'\fR to report the bug). -.Sp -For other problems that also affect the Debian package, you can and -probably should use the Debian \s-1BTS\s0, too, because, after all, it's also a -bug in the Debian version and it serves as a reminder for other users that -might encounter the same issue. -.IP "I am maintaining rxvt-unicode for distribution/OS \s-1XXX\s0, any recommendation?" 4 -.IX Item "I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation?" -You should build one binary with the default options. \fIconfigure\fR -now enables most useful options, and the trend goes to making them -runtime\-switchable, too, so there is usually no drawback to enbaling them, -except higher disk and possibly memory usage. The perl interpreter should -be enabled, as important functionality (menus, selection, likely more in -the future) depends on it. -.Sp -You should not overwrite the \f(CW\*(C`perl\-ext\-common\*(C'\fR snd \f(CW\*(C`perl\-ext\*(C'\fR resources -system-wide (except maybe with \f(CW\*(C`defaults\*(C'\fR). This will result in useful -behaviour. If your distribution aims at low memory, add an empty -\&\f(CW\*(C`perl\-ext\-common\*(C'\fR resource to the app-defaults file. This will keep the -perl interpreter disabled until the user enables it. -.Sp -If you can/want build more binaries, I recommend building a minimal -one with \f(CW\*(C`\-\-disable\-everything\*(C'\fR (very useful) and a maximal one with -\&\f(CW\*(C`\-\-enable\-everything\*(C'\fR (less useful, it will be very big due to a lot of -encodings built-in that increase download times and are rarely used). -.IP "I need to make it setuid/setgid to support utmp/ptys on my \s-1OS\s0, is this safe?" 4 -.IX Item "I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe?" -It should be, starting with release 7.1. You are encouraged to properly -install urxvt with privileges necessary for your \s-1OS\s0 now. -.Sp -When rxvt-unicode detects that it runs setuid or setgid, it will fork -into a helper process for privileged operations (pty handling on some -systems, utmp/wtmp/lastlog handling on others) and drop privileges -immediately. This is much safer than most other terminals that keep -privileges while running (but is more relevant to urxvt, as it contains -things as perl interpreters, which might be \*(L"helpful\*(R" to attackers). -.Sp -This forking is done as the very first within \fImain()\fR, which is very early -and reduces possible bugs to initialisation code run before \fImain()\fR, or -things like the dynamic loader of your system, which should result in very -little risk. -.IP "When I log-in to another system it tells me about missing terminfo data?" 4 -.IX Item "When I log-in to another system it tells me about missing terminfo data?" -The terminal description used by rxvt-unicode is not as widely available -as that for xterm, or even rxvt (for which the same problem often arises). -.Sp -The correct solution for this problem is to install the terminfo, this can -be done like this (with ncurses' infocmp): -.Sp +.PP +This requires \s-1XFT\s0 support, and the support of your X\-server. If that +doesn't work for you, blame Xorg and Keith Packard. \s-1ARGB\s0 visuals aren't +there yet, no matter what they claim. Rxvt-Unicode contains the neccessary +bugfixes and workarounds for Xft and Xlib to make it work, but that +doesn't mean that your \s-1WM\s0 has the required kludges in place. +.PP +4. Use xcompmgr and let it do the job: +.PP .Vb 2 -\& REMOTE=remotesystem.domain -\& infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" -.Ve -.Sp -\&... or by installing rxvt-unicode normally on the remote system, -.Sp -If you cannot or do not want to do this, then you can simply set -\&\f(CW\*(C`TERM=rxvt\*(C'\fR or even \f(CW\*(C`TERM=xterm\*(C'\fR, and live with the small number of -problems arising, which includes wrong keymapping, less and different -colours and some refresh errors in fullscreen applications. It's a nice -quick-and-dirty workaround for rare cases, though. -.Sp -If you always want to do this (and are fine with the consequences) you -can either recompile rxvt-unicode with the desired \s-1TERM\s0 value or use a -resource to set it: -.Sp -.Vb 1 -\& URxvt.termName: rxvt -.Ve -.Sp -If you don't plan to use \fBrxvt\fR (quite common...) you could also replace -the rxvt terminfo file with the rxvt-unicode one. -.ie n .IP """tic"" outputs some error when compiling the terminfo entry." 4 -.el .IP "\f(CWtic\fR outputs some error when compiling the terminfo entry." 4 -.IX Item "tic outputs some error when compiling the terminfo entry." -Most likely it's the empty definition for \f(CW\*(C`enacs=\*(C'\fR. Just replace it by -\&\f(CW\*(C`enacs=\eE[0@\*(C'\fR and try again. -.ie n .IP """bash""'s readline does not work correctly under @@RXVT_NAME@@." 4 -.el .IP "\f(CWbash\fR's readline does not work correctly under @@RXVT_NAME@@." 4 -.IX Item "bash's readline does not work correctly under @@RXVT_NAME@@." -.PD 0 -.IP "I need a termcap file entry." 4 -.IX Item "I need a termcap file entry." -.PD -One reason you might want this is that some distributions or operating -systems still compile some programs using the long-obsoleted termcap -library (Fedora Core's bash is one example) and rely on a termcap entry -for \f(CW\*(C`rxvt\-unicode\*(C'\fR. -.Sp -You could use rxvt's termcap entry with resonable results in many cases. -You can also create a termcap entry by using terminfo's infocmp program -like this: -.Sp -.Vb 1 -\& infocmp -C rxvt-unicode -.Ve -.Sp -Or you could use this termcap entry, generated by the command above: -.Sp -.Vb 20 -\& rxvt-unicode|rxvt-unicode terminal (X Window System):\e -\& :am:bw:eo:km:mi:ms:xn:xo:\e -\& :co#80:it#8:li#24:lm#0:\e -\& :AL=\eE[%dL:DC=\eE[%dP:DL=\eE[%dM:DO=\eE[%dB:IC=\eE[%d@:\e -\& :K1=\eEOw:K2=\eEOu:K3=\eEOy:K4=\eEOq:K5=\eEOs:LE=\eE[%dD:\e -\& :RI=\eE[%dC:SF=\eE[%dS:SR=\eE[%dT:UP=\eE[%dA:ae=\eE(B:al=\eE[L:\e -\& :as=\eE(0:bl=^G:cd=\eE[J:ce=\eE[K:cl=\eE[H\eE[2J:\e -\& :cm=\eE[%i%d;%dH:cr=^M:cs=\eE[%i%d;%dr:ct=\eE[3g:dc=\eE[P:\e -\& :dl=\eE[M:do=^J:ec=\eE[%dX:ei=\eE[4l:ho=\eE[H:\e -\& :i1=\eE[?47l\eE=\eE[?1l:ic=\eE[@:im=\eE[4h:\e -\& :is=\eE[r\eE[m\eE[2J\eE[H\eE[?7h\eE[?1;3;4;6l\eE[4l:\e -\& :k1=\eE[11~:k2=\eE[12~:k3=\eE[13~:k4=\eE[14~:k5=\eE[15~:\e -\& :k6=\eE[17~:k7=\eE[18~:k8=\eE[19~:k9=\eE[20~:kD=\eE[3~:\e -\& :kI=\eE[2~:kN=\eE[6~:kP=\eE[5~:kb=\e177:kd=\eEOB:ke=\eE[?1l\eE>:\e -\& :kh=\eE[7~:kl=\eEOD:kr=\eEOC:ks=\eE[?1h\eE=:ku=\eEOA:le=^H:\e -\& :mb=\eE[5m:md=\eE[1m:me=\eE[m\e017:mr=\eE[7m:nd=\eE[C:rc=\eE8:\e -\& :sc=\eE7:se=\eE[27m:sf=^J:so=\eE[7m:sr=\eEM:st=\eEH:ta=^I:\e -\& :te=\eE[r\eE[?1049l:ti=\eE[?1049h:ue=\eE[24m:up=\eE[A:\e -\& :us=\eE[4m:vb=\eE[?5h\eE[?5l:ve=\eE[?25h:vi=\eE[?25l:\e -\& :vs=\eE[?25h: -.Ve -.ie n .IP "Why does ""ls"" no longer have coloured output?" 4 -.el .IP "Why does \f(CWls\fR no longer have coloured output?" 4 -.IX Item "Why does ls no longer have coloured output?" -The \f(CW\*(C`ls\*(C'\fR in the \s-1GNU\s0 coreutils unfortunately doesn't use terminfo to -decide wether a terminal has colour, but uses it's own configuration -file. Needless to say, \f(CW\*(C`rxvt\-unicode\*(C'\fR is not in it's default file (among -with most other terminals supporting colour). Either add: -.Sp -.Vb 1 -\& TERM rxvt-unicode -.Ve -.Sp -to \f(CW\*(C`/etc/DIR_COLORS\*(C'\fR or simply add: -.Sp -.Vb 1 -\& alias ls='ls --color=auto' -.Ve -.Sp -to your \f(CW\*(C`.profile\*(C'\fR or \f(CW\*(C`.bashrc\*(C'\fR. -.IP "Why doesn't vim/emacs etc. use the 88 colour mode?" 4 -.IX Item "Why doesn't vim/emacs etc. use the 88 colour mode?" -.PD 0 -.IP "Why doesn't vim/emacs etc. make use of italic?" 4 -.IX Item "Why doesn't vim/emacs etc. make use of italic?" -.IP "Why are the secondary screen-related options not working properly?" 4 -.IX Item "Why are the secondary screen-related options not working properly?" -.PD -Make sure you are using \f(CW\*(C`TERM=rxvt\-unicode\*(C'\fR. Some pre-packaged -distributions (most notably Debian GNU/Linux) break rxvt-unicode -by setting \f(CW\*(C`TERM\*(C'\fR to \f(CW\*(C`rxvt\*(C'\fR, which doesn't have these extra -features. Unfortunately, some of these (most notably, again, Debian -GNU/Linux) furthermore fail to even install the \f(CW\*(C`rxvt\-unicode\*(C'\fR terminfo -file, so you will need to install it on your own (See the question \fBWhen -I log-in to another system it tells me about missing terminfo data?\fR on -how to do this). -.IP "My numerical keypad acts weird and generates differing output?" 4 -.IX Item "My numerical keypad acts weird and generates differing output?" -Some Debian GNUL/Linux users seem to have this problem, although no -specific details were reported so far. It is possible that this is caused -by the wrong \f(CW\*(C`TERM\*(C'\fR setting, although the details of wether and how -this can happen are unknown, as \f(CW\*(C`TERM=rxvt\*(C'\fR should offer a compatible -keymap. See the answer to the previous question, and please report if that -helped. -.IP "Rxvt-unicode does not seem to understand the selected encoding?" 4 -.IX Item "Rxvt-unicode does not seem to understand the selected encoding?" -.PD 0 -.IP "Unicode does not seem to work?" 4 -.IX Item "Unicode does not seem to work?" -.PD -If you encounter strange problems like typing an accented character but -getting two unrelated other characters or similar, or if program output is -subtly garbled, then you should check your locale settings. -.Sp -Rxvt-unicode must be started with the same \f(CW\*(C`LC_CTYPE\*(C'\fR setting as the -programs. Often rxvt-unicode is started in the \f(CW\*(C`C\*(C'\fR locale, while the -login script running within the rxvt-unicode window changes the locale to -something else, e.g. \f(CW\*(C`en_GB.UTF\-8\*(C'\fR. Needless to say, this is not going to work. -.Sp -The best thing is to fix your startup environment, as you will likely run -into other problems. If nothing works you can try this in your .profile. -.Sp -.Vb 1 -\& printf '\ee]701;%s\e007' "$LC_CTYPE" -.Ve -.Sp -If this doesn't work, then maybe you use a \f(CW\*(C`LC_CTYPE\*(C'\fR specification not -supported on your systems. Some systems have a \f(CW\*(C`locale\*(C'\fR command which -displays this (also, \f(CW\*(C`perl \-e0\*(C'\fR can be used to check locale settings, as -it will complain loudly if it cannot set the locale). If it displays something -like: -.Sp -.Vb 1 -\& locale: Cannot set LC_CTYPE to default locale: ... -.Ve -.Sp -Then the locale you specified is not supported on your system. -.Sp -If nothing works and you are sure that everything is set correctly then -you will need to remember a little known fact: Some programs just don't -support locales :( -.IP "Why do some characters look so much different than others?" 4 -.IX Item "Why do some characters look so much different than others?" -.PD 0 -.IP "How does rxvt-unicode choose fonts?" 4 -.IX Item "How does rxvt-unicode choose fonts?" -.PD -Most fonts do not contain the full range of Unicode, which is -fine. Chances are that the font you (or the admin/package maintainer of -your system/os) have specified does not cover all the characters you want -to display. -.Sp -\&\fBrxvt-unicode\fR makes a best-effort try at finding a replacement -font. Often the result is fine, but sometimes the chosen font looks -bad/ugly/wrong. Some fonts have totally strange characters that don't -resemble the correct glyph at all, and rxvt-unicode lacks the artificial -intelligence to detect that a specific glyph is wrong: it has to believe -the font that the characters it claims to contain indeed look correct. -.Sp -In that case, select a font of your taste and add it to the font list, -e.g.: -.Sp -.Vb 1 -\& @@RXVT_NAME@@ -fn basefont,font2,font3... +\& xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \e +\& -set _NET_WM_WINDOW_OPACITY 0xc0000000 .Ve -.Sp -When rxvt-unicode sees a character, it will first look at the base -font. If the base font does not contain the character, it will go to the -next font, and so on. Specifying your own fonts will also speed up this -search and use less resources within rxvt-unicode and the X\-server. -.Sp -The only limitation is that none of the fonts may be larger than the base -font, as the base font defines the terminal character cell size, which -must be the same due to the way terminals work. -.IP "Why do some chinese characters look so different than others?" 4 -.IX Item "Why do some chinese characters look so different than others?" +.PP +Then click on a window you want to make transparent. Replace \f(CW0xc0000000\fR +by other values to change the degree of opacity. If it doesn't work and +your server crashes, you got to keep the pieces. +.PP +\fIWhy do some chinese characters look so different than others?\fR +.IX Subsection "Why do some chinese characters look so different than others?" +.PP This is because there is a difference between script and language \*(-- rxvt-unicode does not know which language the text that is output is, as it only knows the unicode character codes. If rxvt-unicode first @@ -655,306 +415,142 @@ chinese characters aren't represented in japanese fonts, so when the first non-japanese character comes up, rxvt-unicode will look for a chinese font \&\*(-- unfortunately at this point, it will still use the japanese font for chinese characters that are also in the japanese font. -.Sp +.PP The workaround is easy: just tag a chinese font at the end of your font list (see the previous question). The key is to view the font list as a preference list: If you expect more japanese, list a japanese font first. If you expect more chinese, put a chinese font first. -.Sp +.PP In the future it might be possible to switch language preferences at runtime (the internal data structure has no problem with using different fonts for the same character at the same time, but no interface for this has been designed yet). -.Sp +.PP Until then, you might get away with switching fonts at runtime (see \*(L"Can I switch the fonts at runtime?\*(R" later in this document). -.IP "Why does rxvt-unicode sometimes leave pixel droppings?" 4 -.IX Item "Why does rxvt-unicode sometimes leave pixel droppings?" +.PP +\fIWhy does rxvt-unicode sometimes leave pixel droppings?\fR +.IX Subsection "Why does rxvt-unicode sometimes leave pixel droppings?" +.PP Most fonts were not designed for terminal use, which means that character size varies a lot. A font that is otherwise fine for terminal use might contain some characters that are simply too wide. Rxvt-unicode will avoid these characters. For characters that are just \*(L"a bit\*(R" too wide a special \&\*(L"careful\*(R" rendering mode is used that redraws adjacent characters. -.Sp +.PP All of this requires that fonts do not lie about character sizes, however: Xft fonts often draw glyphs larger than their acclaimed bounding box, and rxvt-unicode has no way of detecting this (the correct way is to ask for the character bounding box, which unfortunately is wrong in these cases). -.Sp +.PP It's not clear (to me at least), wether this is a bug in Xft, freetype, or the respective font. If you encounter this problem you might try using the \f(CW\*(C`\-lsp\*(C'\fR option to give the font more height. If that doesn't work, you might be forced to use a different font. -.Sp +.PP All of this is not a problem when using X11 core fonts, as their bounding box data is correct. -.IP "On Solaris 9, many line-drawing characters are too wide." 4 -.IX Item "On Solaris 9, many line-drawing characters are too wide." -Seems to be a known bug, read -. Some people use the -following ugly workaround to get non-double-wide-characters working: -.Sp -.Vb 1 -\& #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) -.Ve -.IP "My Compose (Multi_key) key is no longer working." 4 -.IX Item "My Compose (Multi_key) key is no longer working." -The most common causes for this are that either your locale is not set -correctly, or you specified a \fBpreeditStyle\fR that is not supported by -your input method. For example, if you specified \fBOverTheSpot\fR and -your input method (e.g. the default input method handling Compose keys) -does not support this (for instance because it is not visual), then -rxvt-unicode will continue without an input method. -.Sp -In this case either do not specify a \fBpreeditStyle\fR or specify more than -one pre-edit style, such as \fBOverTheSpot,Root,None\fR. -.ie n .IP "I cannot type ""Ctrl\-Shift\-2"" to get an \s-1ASCII\s0 \s-1NUL\s0 character due to \s-1ISO\s0 14755" 4 -.el .IP "I cannot type \f(CWCtrl\-Shift\-2\fR to get an \s-1ASCII\s0 \s-1NUL\s0 character due to \s-1ISO\s0 14755" 4 -.IX Item "I cannot type Ctrl-Shift-2 to get an ASCII NUL character due to ISO 14755" -Either try \f(CW\*(C`Ctrl\-2\*(C'\fR alone (it often is mapped to \s-1ASCII\s0 \s-1NUL\s0 even on -international keyboards) or simply use \s-1ISO\s0 14755 support to your -advantage, typing to get a \s-1ASCII\s0 \s-1NUL\s0. This works for other -codes, too, such as \f(CW\*(C`Ctrl\-Shift\-1\-d\*(C'\fR to type the default telnet escape -character and so on. -.IP "How can I keep rxvt-unicode from using reverse video so much?" 4 -.IX Item "How can I keep rxvt-unicode from using reverse video so much?" +.PP +\fIHow can I keep rxvt-unicode from using reverse video so much?\fR +.IX Subsection "How can I keep rxvt-unicode from using reverse video so much?" +.PP First of all, make sure you are running with the right terminal settings (\f(CW\*(C`TERM=rxvt\-unicode\*(C'\fR), which will get rid of most of these effects. Then make sure you have specified colours for italic and bold, as otherwise rxvt-unicode might use reverse video to simulate the effect: -.Sp +.PP .Vb 2 \& URxvt.colorBD: white \& URxvt.colorIT: green .Ve -.IP "Some programs assume totally weird colours (red instead of blue), how can I fix that?" 4 -.IX Item "Some programs assume totally weird colours (red instead of blue), how can I fix that?" +.PP +\fISome programs assume totally weird colours (red instead of blue), how can I fix that?\fR +.IX Subsection "Some programs assume totally weird colours (red instead of blue), how can I fix that?" +.PP For some unexplainable reason, some rare programs assume a very weird colour palette when confronted with a terminal with more than the standard 8 colours (rxvt\-unicode supports 88). The right fix is, of course, to fix these programs not to assume non-ISO colours without very good reasons. -.Sp +.PP In the meantime, you can either edit your \f(CW\*(C`rxvt\-unicode\*(C'\fR terminfo definition to only claim 8 colour support or use \f(CW\*(C`TERM=rxvt\*(C'\fR, which will fix colours but keep you from using other rxvt-unicode features. -.IP "I am on FreeBSD and rxvt-unicode does not seem to work at all." 4 -.IX Item "I am on FreeBSD and rxvt-unicode does not seem to work at all." -Rxvt-unicode requires the symbol \f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR to be defined -in your compile environment, or an implementation that implements it, -wether it defines the symbol or not. \f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR requires that -\&\fBwchar_t\fR is represented as unicode. -.Sp -As you might have guessed, FreeBSD does neither define this symobl nor -does it support it. Instead, it uses it's own internal representation of -\&\fBwchar_t\fR. This is, of course, completely fine with respect to standards. -.Sp -However, that means rxvt-unicode only works in \f(CW\*(C`POSIX\*(C'\fR, \f(CW\*(C`ISO\-8859\-1\*(C'\fR and -\&\f(CW\*(C`UTF\-8\*(C'\fR locales under FreeBSD (which all use Unicode as \fBwchar_t\fR. -.Sp -\&\f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR is the only sane way to support multi-language -apps in an \s-1OS\s0, as using a locale-dependent (and non\-standardized) -representation of \fBwchar_t\fR makes it impossible to convert between -\&\fBwchar_t\fR (as used by X11 and your applications) and any other encoding -without implementing OS-specific-wrappers for each and every locale. There -simply are no APIs to convert \fBwchar_t\fR into anything except the current -locale encoding. -.Sp -Some applications (such as the formidable \fBmlterm\fR) work around this -by carrying their own replacement functions for character set handling -with them, and either implementing OS-dependent hacks or doing multiple -conversions (which is slow and unreliable in case the \s-1OS\s0 implements -encodings slightly different than the terminal emulator). -.Sp -The rxvt-unicode author insists that the right way to fix this is in the -system libraries once and for all, instead of forcing every app to carry -complete replacements for them :) -.IP "I use Solaris 9 and it doesn't compile/work/etc." 4 -.IX Item "I use Solaris 9 and it doesn't compile/work/etc." -Try the diff in \fIdoc/solaris9.patch\fR as a base. It fixes the worst -problems with \f(CW\*(C`wcwidth\*(C'\fR and a compile problem. -.IP "How can I use rxvt-unicode under cygwin?" 4 -.IX Item "How can I use rxvt-unicode under cygwin?" -rxvt-unicode should compile and run out of the box on cygwin, using -the X11 libraries that come with cygwin. libW11 emulation is no -longer supported (and makes no sense, either, as it only supported a -single font). I recommend starting the X\-server in \f(CW\*(C`\-multiwindow\*(C'\fR or -\&\f(CW\*(C`\-rootless\*(C'\fR mode instead, which will result in similar look&feel as the -old libW11 emulation. -.Sp -At the time of this writing, cygwin didn't seem to support any multi-byte -encodings (you might try \f(CW\*(C`LC_CTYPE=C\-UTF\-8\*(C'\fR), so you are likely limited -to 8\-bit encodings. -.IP "How does rxvt-unicode determine the encoding to use?" 4 -.IX Item "How does rxvt-unicode determine the encoding to use?" -.PD 0 -.IP "Is there an option to switch encodings?" 4 -.IX Item "Is there an option to switch encodings?" -.PD -Unlike some other terminals, rxvt-unicode has no encoding switch, and no -specific \*(L"utf\-8\*(R" mode, such as xterm. In fact, it doesn't even know about -\&\s-1UTF\-8\s0 or any other encodings with respect to terminal I/O. -.Sp -The reasons is that there exists a perfectly fine mechanism for selecting -the encoding, doing I/O and (most important) communicating this to all -applications so everybody agrees on character properties such as width -and code number. This mechanism is the \fIlocale\fR. Applications not using -that info will have problems (for example, \f(CW\*(C`xterm\*(C'\fR gets the width of -characters wrong as it uses it's own, locale-independent table under all -locales). -.Sp -Rxvt-unicode uses the \f(CW\*(C`LC_CTYPE\*(C'\fR locale category to select encoding. All -programs doing the same (that is, most) will automatically agree in the -interpretation of characters. -.Sp -Unfortunately, there is no system-independent way to select locales, nor -is there a standard on how locale specifiers will look like. -.Sp -On most systems, the content of the \f(CW\*(C`LC_CTYPE\*(C'\fR environment variable -contains an arbitrary string which corresponds to an already-installed -locale. Common names for locales are \f(CW\*(C`en_US.UTF\-8\*(C'\fR, \f(CW\*(C`de_DE.ISO\-8859\-15\*(C'\fR, -\&\f(CW\*(C`ja_JP.EUC\-JP\*(C'\fR, i.e. \f(CW\*(C`language_country.encoding\*(C'\fR, but other forms -(i.e. \f(CW\*(C`de\*(C'\fR or \f(CW\*(C`german\*(C'\fR) are also common. -.Sp -Rxvt-unicode ignores all other locale categories, and except for -the encoding, ignores country or language-specific settings, -i.e. \f(CW\*(C`de_DE.UTF\-8\*(C'\fR and \f(CW\*(C`ja_JP.UTF\-8\*(C'\fR are the normally same to -rxvt\-unicode. -.Sp -If you want to use a specific encoding you have to make sure you start -rxvt-unicode with the correct \f(CW\*(C`LC_CTYPE\*(C'\fR category. -.IP "Can I switch locales at runtime?" 4 -.IX Item "Can I switch locales at runtime?" -Yes, using an escape sequence. Try something like this, which sets -rxvt\-unicode's idea of \f(CW\*(C`LC_CTYPE\*(C'\fR. -.Sp -.Vb 1 -\& printf '\ee]701;%s\e007' ja_JP.SJIS -.Ve -.Sp -See also the previous answer. -.Sp -Sometimes this capability is rather handy when you want to work in -one locale (e.g. \f(CW\*(C`de_DE.UTF\-8\*(C'\fR) but some programs don't support it -(e.g. \s-1UTF\-8\s0). For example, I use this script to start \f(CW\*(C`xjdic\*(C'\fR, which -first switches to a locale supported by xjdic and back later: -.Sp -.Vb 3 -\& printf '\ee]701;%s\e007' ja_JP.SJIS -\& xjdic -js -\& printf '\ee]701;%s\e007' de_DE.UTF-8 -.Ve -.Sp -You can also use xterm's \f(CW\*(C`luit\*(C'\fR program, which usually works fine, except -for some locales where character width differs between program\- and -rxvt\-unicode\-locales. -.IP "Can I switch the fonts at runtime?" 4 -.IX Item "Can I switch the fonts at runtime?" +.PP +\fICan I switch the fonts at runtime?\fR +.IX Subsection "Can I switch the fonts at runtime?" +.PP Yes, using an escape sequence. Try something like this, which has the same effect as using the \f(CW\*(C`\-fn\*(C'\fR switch, and takes effect immediately: -.Sp +.PP .Vb 1 \& printf '\ee]50;%s\e007' "9x15bold,xft:Kochi Gothic" .Ve -.Sp +.PP This is useful if you e.g. work primarily with japanese (and prefer a japanese font), but you have to switch to chinese temporarily, where japanese fonts would only be in your way. -.Sp +.PP You can think of this as a kind of manual \s-1ISO\-2022\s0 switching. -.IP "Why do italic characters look as if clipped?" 4 -.IX Item "Why do italic characters look as if clipped?" +.PP +\fIWhy do italic characters look as if clipped?\fR +.IX Subsection "Why do italic characters look as if clipped?" +.PP Many fonts have difficulties with italic characters and hinting. For example, the otherwise very nicely hinted font \f(CW\*(C`xft:Bitstream Vera Sans Mono\*(C'\fR completely fails in it's italic face. A workaround might be to enable freetype autohinting, i.e. like this: -.Sp +.PP .Vb 2 \& URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true \& URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true .Ve -.IP "My input method wants but I want \s-1UTF\-8\s0, what can I do?" 4 -.IX Item "My input method wants but I want UTF-8, what can I do?" -You can specify separate locales for the input method and the rest of the -terminal, using the resource \f(CW\*(C`imlocale\*(C'\fR: -.Sp -.Vb 1 -\& URxvt.imlocale: ja_JP.EUC-JP -.Ve -.Sp -Now you can start your terminal with \f(CW\*(C`LC_CTYPE=ja_JP.UTF\-8\*(C'\fR and still -use your input method. Please note, however, that you will not be able to -input characters outside \f(CW\*(C`EUC\-JP\*(C'\fR in a normal way then, as your input -method limits you. -.IP "Rxvt-unicode crashes when the X Input Method changes or exits." 4 -.IX Item "Rxvt-unicode crashes when the X Input Method changes or exits." -Unfortunately, this is unavoidable, as the \s-1XIM\s0 protocol is racy by -design. Applications can avoid some crashes at the expense of memory -leaks, and Input Methods can avoid some crashes by careful ordering at -exit time. \fBkinput2\fR (and derived input methods) generally succeeds, -while \fB\s-1SCIM\s0\fR (or similar input methods) fails. In the end, however, -crashes cannot be completely avoided even if both sides cooperate. -.Sp -So the only workaround is not to kill your Input Method Servers. -.IP "Rxvt-unicode uses gobs of memory, how can I reduce that?" 4 -.IX Item "Rxvt-unicode uses gobs of memory, how can I reduce that?" -Rxvt-unicode tries to obey the rule of not charging you for something you -don't use. One thing you should try is to configure out all settings that -you don't need, for example, Xft support is a resource hog by design, -when used. Compiling it out ensures that no Xft font will be loaded -accidentally when rxvt-unicode tries to find a font for your characters. -.Sp -Also, many people (me included) like large windows and even larger -scrollback buffers: Without \f(CW\*(C`\-\-enable\-unicode3\*(C'\fR, rxvt-unicode will use -6 bytes per screen cell. For a 160x?? window this amounts to almost a -kilobyte per line. A scrollback buffer of 10000 lines will then (if full) -use 10 Megabytes of memory. With \f(CW\*(C`\-\-enable\-unicode3\*(C'\fR it gets worse, as -rxvt-unicode then uses 8 bytes per screen cell. -.IP "Can I speed up Xft rendering somehow?" 4 -.IX Item "Can I speed up Xft rendering somehow?" +.PP +\fICan I speed up Xft rendering somehow?\fR +.IX Subsection "Can I speed up Xft rendering somehow?" +.PP Yes, the most obvious way to speed it up is to avoid Xft entirely, as it is simply slow. If you still want Xft fonts you might try to disable antialiasing (by appending \f(CW\*(C`:antialias=false\*(C'\fR), which saves lots of memory and also speeds up rendering considerably. -.IP "Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?" 4 -.IX Item "Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?" +.PP +\fIRxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?\fR +.IX Subsection "Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong?" +.PP Rxvt-unicode will use whatever you specify as a font. If it needs to fall back to it's default font search list it will prefer X11 core fonts, because they are small and fast, and then use Xft fonts. It has antialiasing disabled for most of them, because the author thinks they look best that way. -.Sp +.PP If you want antialiasing, you have to specify the fonts manually. -.IP "Mouse cut/paste suddenly no longer works." 4 -.IX Item "Mouse cut/paste suddenly no longer works." -Make sure that mouse reporting is actually turned off since killing -some editors prematurely may leave the mouse in mouse report mode. I've -heard that tcsh may use mouse reporting unless it otherwise specified. A -quick check is to see if cut/paste works when the Alt or Shift keys are -depressed. -.IP "What's with this bold/blink stuff?" 4 -.IX Item "What's with this bold/blink stuff?" +.PP +\fIWhat's with this bold/blink stuff?\fR +.IX Subsection "What's with this bold/blink stuff?" +.PP If no bold colour is set via \f(CW\*(C`colorBD:\*(C'\fR, bold will invert text using the standard foreground colour. -.Sp +.PP For the standard background colour, blinking will actually make the text blink when compiled with \f(CW\*(C`\-\-enable\-blinking\*(C'\fR. with standard colours. Without \f(CW\*(C`\-\-enable\-blinking\*(C'\fR, the blink attribute will be ignored. -.Sp +.PP On \s-1ANSI\s0 colours, bold/blink attributes are used to set high-intensity foreground/background colors. -.Sp +.PP color0\-7 are the low-intensity colors. -.Sp +.PP color8\-15 are the corresponding high-intensity colors. -.IP "I don't like the screen colors. How do I change them?" 4 -.IX Item "I don't like the screen colors. How do I change them?" +.PP +\fII don't like the screen colors. How do I change them?\fR +.IX Subsection "I don't like the screen colors. How do I change them?" +.PP You can change the screen colors at run-time using \fI~/.Xdefaults\fR resources (or as long\-options). -.Sp +.PP Here are values that are supposed to resemble a \s-1VGA\s0 screen, including the murky brown that passes for low-intensity yellow: -.Sp +.PP .Vb 8 \& URxvt.color0: #000000 \& URxvt.color1: #A80000 @@ -965,7 +561,7 @@ including the murky brown that passes for low-intensity yellow: \& URxvt.color6: #00A8A8 \& URxvt.color7: #A8A8A8 .Ve -.Sp +.PP .Vb 8 \& URxvt.color8: #000054 \& URxvt.color9: #FF0054 @@ -976,10 +572,10 @@ including the murky brown that passes for low-intensity yellow: \& URxvt.color14: #00FFFF \& URxvt.color15: #FFFFFF .Ve -.Sp +.PP And here is a more complete set of non-standard colors described (not by me) as \*(L"pretty girly\*(R". -.Sp +.PP .Vb 18 \& URxvt.cursorColor: #dc74d1 \& URxvt.pointerColor: #dc74d1 @@ -1000,81 +596,228 @@ me) as \*(L"pretty girly\*(R". \& URxvt.color7: #e1dddd \& URxvt.color15: #e1dddd .Ve -.IP "How can I start @@RXVT_NAME@@d in a race-free way?" 4 -.IX Item "How can I start @@RXVT_NAME@@d in a race-free way?" -Try \f(CW\*(C`@@RXVT_NAME@@d \-f \-o\*(C'\fR, which tells @@RXVT_NAME@@d to open the -display, create the listening socket and then fork. -.IP "What's with the strange Backspace/Delete key behaviour?" 4 -.IX Item "What's with the strange Backspace/Delete key behaviour?" -Assuming that the physical Backspace key corresponds to the -BackSpace keysym (not likely for Linux ... see the following -question) there are two standard values that can be used for -Backspace: \f(CW\*(C`^H\*(C'\fR and \f(CW\*(C`^?\*(C'\fR. -.Sp -Historically, either value is correct, but rxvt-unicode adopts the debian -policy of using \f(CW\*(C`^?\*(C'\fR when unsure, because it's the one only only correct -choice :). -.Sp -Rxvt-unicode tries to inherit the current stty settings and uses the value -of `erase' to guess the value for backspace. If rxvt-unicode wasn't -started from a terminal (say, from a menu or by remote shell), then the -system value of `erase', which corresponds to \s-1CERASE\s0 in , will -be used (which may not be the same as your stty setting). -.Sp -For starting a new rxvt\-unicode: -.Sp -.Vb 3 -\& # use Backspace = ^H -\& $ stty erase ^H -\& $ @@RXVT_NAME@@ +.PP +\fIWhy do some characters look so much different than others?\fR +.IX Subsection "Why do some characters look so much different than others?" +.PP +See next entry. +.PP +\fIHow does rxvt-unicode choose fonts?\fR +.IX Subsection "How does rxvt-unicode choose fonts?" +.PP +Most fonts do not contain the full range of Unicode, which is +fine. Chances are that the font you (or the admin/package maintainer of +your system/os) have specified does not cover all the characters you want +to display. +.PP +\&\fBrxvt-unicode\fR makes a best-effort try at finding a replacement +font. Often the result is fine, but sometimes the chosen font looks +bad/ugly/wrong. Some fonts have totally strange characters that don't +resemble the correct glyph at all, and rxvt-unicode lacks the artificial +intelligence to detect that a specific glyph is wrong: it has to believe +the font that the characters it claims to contain indeed look correct. +.PP +In that case, select a font of your taste and add it to the font list, +e.g.: +.PP +.Vb 1 +\& @@RXVT_NAME@@ -fn basefont,font2,font3... .Ve -.Sp -.Vb 3 -\& # use Backspace = ^? -\& $ stty erase ^? -\& $ @@RXVT_NAME@@ +.PP +When rxvt-unicode sees a character, it will first look at the base +font. If the base font does not contain the character, it will go to the +next font, and so on. Specifying your own fonts will also speed up this +search and use less resources within rxvt-unicode and the X\-server. +.PP +The only limitation is that none of the fonts may be larger than the base +font, as the base font defines the terminal character cell size, which +must be the same due to the way terminals work. +.Sh "Keyboard, Mouse & User Interaction" +.IX Subsection "Keyboard, Mouse & User Interaction" +\fIThe new selection selects pieces that are too big, how can I select single words?\fR +.IX Subsection "The new selection selects pieces that are too big, how can I select single words?" +.PP +If you want to select e.g. alphanumeric words, you can use the following +setting: +.PP +.Vb 1 +\& URxvt.selection.pattern-0: ([[:word:]]+) .Ve -.Sp +.PP +If you click more than twice, the selection will be extended +more and more. +.PP +To get a selection that is very similar to the old code, try this pattern: +.PP +.Vb 1 +\& URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\e\e\e\e]^`{|})]+) +.Ve +.PP +Please also note that the \fILeftClick Shift-LeftClik\fR combination also +selects words like the old code. +.PP +\fII don't like the new selection/popups/hotkeys/perl, how do I change/disable it?\fR +.IX Subsection "I don't like the new selection/popups/hotkeys/perl, how do I change/disable it?" +.PP +You can disable the perl extension completely by setting the +\&\fBperl-ext-common\fR resource to the empty string, which also keeps +rxvt-unicode from initialising perl, saving memory. +.PP +If you only want to disable specific features, you first have to +identify which perl extension is responsible. For this, read the section +\&\fB\s-1PREPACKAGED\s0 \s-1EXTENSIONS\s0\fR in the @@RXVT_NAME@@\fIperl\fR\|(3) manpage. For +example, to disable the \fBselection-popup\fR and \fBoption-popup\fR, specify +this \fBperl-ext-common\fR resource: +.PP +.Vb 1 +\& URxvt.perl-ext-common: default,-selection-popup,-option-popup +.Ve +.PP +This will keep the default extensions, but disable the two popup +extensions. Some extensions can also be configured, for example, +scrollback search mode is triggered by \fBM\-s\fR. You can move it to any +other combination either by setting the \fBsearchable-scrollback\fR resource: +.PP +.Vb 1 +\& URxvt.searchable-scrollback: CM-s +.Ve +.PP +\fIThe cursor moves when selecting text in the current input line, how do I switch this off?\fR +.IX Subsection "The cursor moves when selecting text in the current input line, how do I switch this off?" +.PP +See next entry. +.PP +\fIDuring rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this?\fR +.IX Subsection "During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this?" +.PP +These are caused by the \f(CW\*(C`readline\*(C'\fR perl extension. Under normal +circumstances, it will move your cursor around when you click into the +line that contains it. It tries hard not to do this at the wrong moment, +but when running a program that doesn't parse cursor movements or in some +cases during rlogin sessions, it fails to detect this properly. +.PP +You can permamently switch this feature off by disabling the \f(CW\*(C`readline\*(C'\fR +extension: +.PP +.Vb 1 +\& URxvt.perl-ext-common: default,-readline +.Ve +.PP +\fIMy numerical keypad acts weird and generates differing output?\fR +.IX Subsection "My numerical keypad acts weird and generates differing output?" +.PP +Some Debian GNUL/Linux users seem to have this problem, although no +specific details were reported so far. It is possible that this is caused +by the wrong \f(CW\*(C`TERM\*(C'\fR setting, although the details of wether and how +this can happen are unknown, as \f(CW\*(C`TERM=rxvt\*(C'\fR should offer a compatible +keymap. See the answer to the previous question, and please report if that +helped. +.PP +\fIMy Compose (Multi_key) key is no longer working.\fR +.IX Subsection "My Compose (Multi_key) key is no longer working." +.PP +The most common causes for this are that either your locale is not set +correctly, or you specified a \fBpreeditStyle\fR that is not supported by +your input method. For example, if you specified \fBOverTheSpot\fR and +your input method (e.g. the default input method handling Compose keys) +does not support this (for instance because it is not visual), then +rxvt-unicode will continue without an input method. +.PP +In this case either do not specify a \fBpreeditStyle\fR or specify more than +one pre-edit style, such as \fBOverTheSpot,Root,None\fR. +.PP +\fII cannot type \f(CI\*(C`Ctrl\-Shift\-2\*(C'\fI to get an \s-1ASCII\s0 \s-1NUL\s0 character due to \s-1ISO\s0 14755\fR +.IX Subsection "I cannot type Ctrl-Shift-2 to get an ASCII NUL character due to ISO 14755" +.PP +Either try \f(CW\*(C`Ctrl\-2\*(C'\fR alone (it often is mapped to \s-1ASCII\s0 \s-1NUL\s0 even on +international keyboards) or simply use \s-1ISO\s0 14755 support to your +advantage, typing to get a \s-1ASCII\s0 \s-1NUL\s0. This works for other +codes, too, such as \f(CW\*(C`Ctrl\-Shift\-1\-d\*(C'\fR to type the default telnet escape +character and so on. +.PP +\fIMouse cut/paste suddenly no longer works.\fR +.IX Subsection "Mouse cut/paste suddenly no longer works." +.PP +Make sure that mouse reporting is actually turned off since killing +some editors prematurely may leave the mouse in mouse report mode. I've +heard that tcsh may use mouse reporting unless it otherwise specified. A +quick check is to see if cut/paste works when the Alt or Shift keys are +depressed. +.PP +\fIWhat's with the strange Backspace/Delete key behaviour?\fR +.IX Subsection "What's with the strange Backspace/Delete key behaviour?" +.PP +Assuming that the physical Backspace key corresponds to the +BackSpace keysym (not likely for Linux ... see the following +question) there are two standard values that can be used for +Backspace: \f(CW\*(C`^H\*(C'\fR and \f(CW\*(C`^?\*(C'\fR. +.PP +Historically, either value is correct, but rxvt-unicode adopts the debian +policy of using \f(CW\*(C`^?\*(C'\fR when unsure, because it's the one only only correct +choice :). +.PP +Rxvt-unicode tries to inherit the current stty settings and uses the value +of `erase' to guess the value for backspace. If rxvt-unicode wasn't +started from a terminal (say, from a menu or by remote shell), then the +system value of `erase', which corresponds to \s-1CERASE\s0 in , will +be used (which may not be the same as your stty setting). +.PP +For starting a new rxvt\-unicode: +.PP +.Vb 3 +\& # use Backspace = ^H +\& $ stty erase ^H +\& $ @@RXVT_NAME@@ +.Ve +.PP +.Vb 3 +\& # use Backspace = ^? +\& $ stty erase ^? +\& $ @@RXVT_NAME@@ +.Ve +.PP Toggle with \f(CW\*(C`ESC [ 36 h\*(C'\fR / \f(CW\*(C`ESC [ 36 l\*(C'\fR. -.Sp +.PP For an existing rxvt\-unicode: -.Sp +.PP .Vb 3 \& # use Backspace = ^H \& $ stty erase ^H \& $ echo -n "^[[36h" .Ve -.Sp +.PP .Vb 3 \& # use Backspace = ^? \& $ stty erase ^? \& $ echo -n "^[[36l" .Ve -.Sp +.PP This helps satisfy some of the Backspace discrepancies that occur, but if you use Backspace = \f(CW\*(C`^H\*(C'\fR, make sure that the termcap/terminfo value properly reflects that. -.Sp +.PP The Delete key is a another casualty of the ill-defined Backspace problem. To avoid confusion between the Backspace and Delete keys, the Delete key has been assigned an escape sequence to match the vt100 for Execute (\f(CW\*(C`ESC [ 3 ~\*(C'\fR) and is in the supplied termcap/terminfo. -.Sp +.PP Some other Backspace problems: -.Sp -some editors use termcap/terminfo, +.PP +some editors use termcap/terminfo, some editors (vim I'm told) expect Backspace = ^H, \&\s-1GNU\s0 Emacs (and Emacs-like editors) use ^H for help. -.Sp +.PP Perhaps someday this will all be resolved in a consistent manner. -.IP "I don't like the key\-bindings. How do I change them?" 4 -.IX Item "I don't like the key-bindings. How do I change them?" +.PP +\fII don't like the key\-bindings. How do I change them?\fR +.IX Subsection "I don't like the key-bindings. How do I change them?" +.PP There are some compile-time selections available via configure. Unless you have run \*(L"configure\*(R" with the \f(CW\*(C`\-\-disable\-resources\*(C'\fR option you can use the `keysym' resource to alter the keystrings associated with keysyms. -.Sp +.PP Here's an example for a URxvt session started using \f(CW\*(C`@@RXVT_NAME@@ \-name URxvt\*(C'\fR -.Sp +.PP .Vb 20 \& URxvt.keysym.Home: \e033[1~ \& URxvt.keysym.End: \e033[4~ @@ -1097,10 +840,12 @@ Here's an example for a URxvt session started using \f(CW\*(C`@@RXVT_NAME@@ \-na \& URxvt.keysym.M-C-a: list \e033 \& URxvt.keysym.F12: command:\e033]701;zh_CN.GBK\e007 .Ve -.Sp +.PP See some more examples in the documentation for the \fBkeysym\fR resource. -.IP "I'm using keyboard model \s-1XXX\s0 that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following mappings that rxvt-unicode doesn't recognize." 4 -.IX Item "I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following mappings that rxvt-unicode doesn't recognize." +.PP +\fII'm using keyboard model \s-1XXX\s0 that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map\fR +.IX Subsection "I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map" +.PP .Vb 6 \& KP_Insert == Insert \& F22 == Print @@ -1109,52 +854,418 @@ See some more examples in the documentation for the \fBkeysym\fR resource. \& F33 == End \& F35 == Next .Ve -.Sp +.PP Rather than have rxvt-unicode try to accommodate all the various possible keyboard mappings, it is better to use `xmodmap' to remap the keys as required for your particular machine. -.IP "How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc." 4 -.IX Item "How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc." -rxvt and rxvt-unicode always export the variable \*(L"\s-1COLORTERM\s0\*(R", so you can -check and see if that is set. Note that several programs, \s-1JED\s0, slrn, -Midnight Commander automatically check this variable to decide whether or -not to use color. -.IP "How do I set the correct, full \s-1IP\s0 address for the \s-1DISPLAY\s0 variable?" 4 -.IX Item "How do I set the correct, full IP address for the DISPLAY variable?" -If you've compiled rxvt-unicode with \s-1DISPLAY_IS_IP\s0 and have enabled -insecure mode then it is possible to use the following shell script -snippets to correctly set the display. If your version of rxvt-unicode -wasn't also compiled with \s-1ESCZ_ANSWER\s0 (as assumed in these snippets) then -the \s-1COLORTERM\s0 variable can be used to distinguish rxvt-unicode from a -regular xterm. -.Sp -Courtesy of Chuck Blake with the following shell script -snippets: -.Sp -.Vb 12 -\& # Bourne/Korn/POSIX family of shells: -\& [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know -\& if [ ${TERM:-foo} = xterm ]; then -\& stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not -\& echo -n '^[Z' -\& read term_id -\& stty icanon echo -\& if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then -\& echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string -\& read DISPLAY # set it in our local shell -\& fi -\& fi +.Sh "Terminal Configuration" +.IX Subsection "Terminal Configuration" +\fIWhy doesn't rxvt-unicode read my resources?\fR +.IX Subsection "Why doesn't rxvt-unicode read my resources?" +.PP +Well, why, indeed? It does, in a way very similar to other X +applications. Most importantly, this means that if you or your \s-1OS\s0 loads +resources into the X display (the right way to do it), rxvt-unicode will +ignore any resource files in your home directory. It will only read +\&\fI$HOME/.Xdefaults\fR when no resources are attached to the display. +.PP +If you have or use an \fI$HOME/.Xresources\fR file, chances are that +resources are loaded into your X\-server. In this case, you have to +re-login after every change (or run \fIxrdb \-merge \f(CI$HOME\fI/.Xresources\fR). +.PP +Also consider the form resources have to use: +.PP +.Vb 1 +\& URxvt.resource: value .Ve -.IP "How do I compile the manual pages for myself?" 4 -.IX Item "How do I compile the manual pages for myself?" -You need to have a recent version of perl installed as \fI/usr/bin/perl\fR, -one that comes with \fIpod2man\fR, \fIpod2text\fR and \fIpod2html\fR. Then go to -the doc subdirectory and enter \f(CW\*(C`make alldoc\*(C'\fR. -.IP "My question isn't answered here, can I ask a human?" 4 -.IX Item "My question isn't answered here, can I ask a human?" -Before sending me mail, you could go to \s-1IRC:\s0 \f(CW\*(C`irc.freenode.net\*(C'\fR, -channel \f(CW\*(C`#rxvt\-unicode\*(C'\fR has some rxvt-unicode enthusiasts that might be -interested in learning about new and exciting problems (but not FAQs :). +.PP +If you want to use another form (there are lots of different ways of +specifying resources), make sure you understand wether and why it +works. If unsure, use the form above. +.PP +\fIWhen I log-in to another system it tells me about missing terminfo data?\fR +.IX Subsection "When I log-in to another system it tells me about missing terminfo data?" +.PP +The terminal description used by rxvt-unicode is not as widely available +as that for xterm, or even rxvt (for which the same problem often arises). +.PP +The correct solution for this problem is to install the terminfo, this can +be done like this (with ncurses' infocmp): +.PP +.Vb 2 +\& REMOTE=remotesystem.domain +\& infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" +.Ve +.PP +\&... or by installing rxvt-unicode normally on the remote system, +.PP +If you cannot or do not want to do this, then you can simply set +\&\f(CW\*(C`TERM=rxvt\*(C'\fR or even \f(CW\*(C`TERM=xterm\*(C'\fR, and live with the small number of +problems arising, which includes wrong keymapping, less and different +colours and some refresh errors in fullscreen applications. It's a nice +quick-and-dirty workaround for rare cases, though. +.PP +If you always want to do this (and are fine with the consequences) you +can either recompile rxvt-unicode with the desired \s-1TERM\s0 value or use a +resource to set it: +.PP +.Vb 1 +\& URxvt.termName: rxvt +.Ve +.PP +If you don't plan to use \fBrxvt\fR (quite common...) you could also replace +the rxvt terminfo file with the rxvt-unicode one. +.PP +\fI\f(CI\*(C`tic\*(C'\fI outputs some error when compiling the terminfo entry.\fR +.IX Subsection "tic outputs some error when compiling the terminfo entry." +.PP +Most likely it's the empty definition for \f(CW\*(C`enacs=\*(C'\fR. Just replace it by +\&\f(CW\*(C`enacs=\eE[0@\*(C'\fR and try again. +.PP +\fI\f(CI\*(C`bash\*(C'\fI's readline does not work correctly under @@RXVT_NAME@@.\fR +.IX Subsection "bash's readline does not work correctly under @@RXVT_NAME@@." +.PP +See next entry. +.PP +\fII need a termcap file entry.\fR +.IX Subsection "I need a termcap file entry." +.PP +One reason you might want this is that some distributions or operating +systems still compile some programs using the long-obsoleted termcap +library (Fedora Core's bash is one example) and rely on a termcap entry +for \f(CW\*(C`rxvt\-unicode\*(C'\fR. +.PP +You could use rxvt's termcap entry with resonable results in many cases. +You can also create a termcap entry by using terminfo's infocmp program +like this: +.PP +.Vb 1 +\& infocmp -C rxvt-unicode +.Ve +.PP +Or you could use this termcap entry, generated by the command above: +.PP +.Vb 20 +\& rxvt-unicode|rxvt-unicode terminal (X Window System):\e +\& :am:bw:eo:km:mi:ms:xn:xo:\e +\& :co#80:it#8:li#24:lm#0:\e +\& :AL=\eE[%dL:DC=\eE[%dP:DL=\eE[%dM:DO=\eE[%dB:IC=\eE[%d@:\e +\& :K1=\eEOw:K2=\eEOu:K3=\eEOy:K4=\eEOq:K5=\eEOs:LE=\eE[%dD:\e +\& :RI=\eE[%dC:SF=\eE[%dS:SR=\eE[%dT:UP=\eE[%dA:ae=\eE(B:al=\eE[L:\e +\& :as=\eE(0:bl=^G:cd=\eE[J:ce=\eE[K:cl=\eE[H\eE[2J:\e +\& :cm=\eE[%i%d;%dH:cr=^M:cs=\eE[%i%d;%dr:ct=\eE[3g:dc=\eE[P:\e +\& :dl=\eE[M:do=^J:ec=\eE[%dX:ei=\eE[4l:ho=\eE[H:\e +\& :i1=\eE[?47l\eE=\eE[?1l:ic=\eE[@:im=\eE[4h:\e +\& :is=\eE[r\eE[m\eE[2J\eE[H\eE[?7h\eE[?1;3;4;6l\eE[4l:\e +\& :k1=\eE[11~:k2=\eE[12~:k3=\eE[13~:k4=\eE[14~:k5=\eE[15~:\e +\& :k6=\eE[17~:k7=\eE[18~:k8=\eE[19~:k9=\eE[20~:kD=\eE[3~:\e +\& :kI=\eE[2~:kN=\eE[6~:kP=\eE[5~:kb=\e177:kd=\eEOB:ke=\eE[?1l\eE>:\e +\& :kh=\eE[7~:kl=\eEOD:kr=\eEOC:ks=\eE[?1h\eE=:ku=\eEOA:le=^H:\e +\& :mb=\eE[5m:md=\eE[1m:me=\eE[m\e017:mr=\eE[7m:nd=\eE[C:rc=\eE8:\e +\& :sc=\eE7:se=\eE[27m:sf=^J:so=\eE[7m:sr=\eEM:st=\eEH:ta=^I:\e +\& :te=\eE[r\eE[?1049l:ti=\eE[?1049h:ue=\eE[24m:up=\eE[A:\e +\& :us=\eE[4m:vb=\eE[?5h\eE[?5l:ve=\eE[?25h:vi=\eE[?25l:\e +\& :vs=\eE[?25h: +.Ve +.PP +\fIWhy does \f(CI\*(C`ls\*(C'\fI no longer have coloured output?\fR +.IX Subsection "Why does ls no longer have coloured output?" +.PP +The \f(CW\*(C`ls\*(C'\fR in the \s-1GNU\s0 coreutils unfortunately doesn't use terminfo to +decide wether a terminal has colour, but uses it's own configuration +file. Needless to say, \f(CW\*(C`rxvt\-unicode\*(C'\fR is not in it's default file (among +with most other terminals supporting colour). Either add: +.PP +.Vb 1 +\& TERM rxvt-unicode +.Ve +.PP +to \f(CW\*(C`/etc/DIR_COLORS\*(C'\fR or simply add: +.PP +.Vb 1 +\& alias ls='ls --color=auto' +.Ve +.PP +to your \f(CW\*(C`.profile\*(C'\fR or \f(CW\*(C`.bashrc\*(C'\fR. +.PP +\fIWhy doesn't vim/emacs etc. use the 88 colour mode?\fR +.IX Subsection "Why doesn't vim/emacs etc. use the 88 colour mode?" +.PP +See next entry. +.PP +\fIWhy doesn't vim/emacs etc. make use of italic?\fR +.IX Subsection "Why doesn't vim/emacs etc. make use of italic?" +.PP +See next entry. +.PP +\fIWhy are the secondary screen-related options not working properly?\fR +.IX Subsection "Why are the secondary screen-related options not working properly?" +.PP +Make sure you are using \f(CW\*(C`TERM=rxvt\-unicode\*(C'\fR. Some pre-packaged +distributions (most notably Debian GNU/Linux) break rxvt-unicode +by setting \f(CW\*(C`TERM\*(C'\fR to \f(CW\*(C`rxvt\*(C'\fR, which doesn't have these extra +features. Unfortunately, some of these (most notably, again, Debian +GNU/Linux) furthermore fail to even install the \f(CW\*(C`rxvt\-unicode\*(C'\fR terminfo +file, so you will need to install it on your own (See the question \fBWhen +I log-in to another system it tells me about missing terminfo data?\fR on +how to do this). +.Sh "Encoding / Locale / Input Method Issues" +.IX Subsection "Encoding / Locale / Input Method Issues" +\fIRxvt-unicode does not seem to understand the selected encoding?\fR +.IX Subsection "Rxvt-unicode does not seem to understand the selected encoding?" +.PP +See next entry. +.PP +\fIUnicode does not seem to work?\fR +.IX Subsection "Unicode does not seem to work?" +.PP +If you encounter strange problems like typing an accented character but +getting two unrelated other characters or similar, or if program output is +subtly garbled, then you should check your locale settings. +.PP +Rxvt-unicode must be started with the same \f(CW\*(C`LC_CTYPE\*(C'\fR setting as the +programs. Often rxvt-unicode is started in the \f(CW\*(C`C\*(C'\fR locale, while the +login script running within the rxvt-unicode window changes the locale to +something else, e.g. \f(CW\*(C`en_GB.UTF\-8\*(C'\fR. Needless to say, this is not going to work. +.PP +The best thing is to fix your startup environment, as you will likely run +into other problems. If nothing works you can try this in your .profile. +.PP +.Vb 1 +\& printf '\ee]701;%s\e007' "$LC_CTYPE" +.Ve +.PP +If this doesn't work, then maybe you use a \f(CW\*(C`LC_CTYPE\*(C'\fR specification not +supported on your systems. Some systems have a \f(CW\*(C`locale\*(C'\fR command which +displays this (also, \f(CW\*(C`perl \-e0\*(C'\fR can be used to check locale settings, as +it will complain loudly if it cannot set the locale). If it displays something +like: +.PP +.Vb 1 +\& locale: Cannot set LC_CTYPE to default locale: ... +.Ve +.PP +Then the locale you specified is not supported on your system. +.PP +If nothing works and you are sure that everything is set correctly then +you will need to remember a little known fact: Some programs just don't +support locales :( +.PP +\fIHow does rxvt-unicode determine the encoding to use?\fR +.IX Subsection "How does rxvt-unicode determine the encoding to use?" +.PP +See next entry. +.PP +\fIIs there an option to switch encodings?\fR +.IX Subsection "Is there an option to switch encodings?" +.PP +Unlike some other terminals, rxvt-unicode has no encoding switch, and no +specific \*(L"utf\-8\*(R" mode, such as xterm. In fact, it doesn't even know about +\&\s-1UTF\-8\s0 or any other encodings with respect to terminal I/O. +.PP +The reasons is that there exists a perfectly fine mechanism for selecting +the encoding, doing I/O and (most important) communicating this to all +applications so everybody agrees on character properties such as width +and code number. This mechanism is the \fIlocale\fR. Applications not using +that info will have problems (for example, \f(CW\*(C`xterm\*(C'\fR gets the width of +characters wrong as it uses it's own, locale-independent table under all +locales). +.PP +Rxvt-unicode uses the \f(CW\*(C`LC_CTYPE\*(C'\fR locale category to select encoding. All +programs doing the same (that is, most) will automatically agree in the +interpretation of characters. +.PP +Unfortunately, there is no system-independent way to select locales, nor +is there a standard on how locale specifiers will look like. +.PP +On most systems, the content of the \f(CW\*(C`LC_CTYPE\*(C'\fR environment variable +contains an arbitrary string which corresponds to an already-installed +locale. Common names for locales are \f(CW\*(C`en_US.UTF\-8\*(C'\fR, \f(CW\*(C`de_DE.ISO\-8859\-15\*(C'\fR, +\&\f(CW\*(C`ja_JP.EUC\-JP\*(C'\fR, i.e. \f(CW\*(C`language_country.encoding\*(C'\fR, but other forms +(i.e. \f(CW\*(C`de\*(C'\fR or \f(CW\*(C`german\*(C'\fR) are also common. +.PP +Rxvt-unicode ignores all other locale categories, and except for +the encoding, ignores country or language-specific settings, +i.e. \f(CW\*(C`de_DE.UTF\-8\*(C'\fR and \f(CW\*(C`ja_JP.UTF\-8\*(C'\fR are the normally same to +rxvt\-unicode. +.PP +If you want to use a specific encoding you have to make sure you start +rxvt-unicode with the correct \f(CW\*(C`LC_CTYPE\*(C'\fR category. +.PP +\fICan I switch locales at runtime?\fR +.IX Subsection "Can I switch locales at runtime?" +.PP +Yes, using an escape sequence. Try something like this, which sets +rxvt\-unicode's idea of \f(CW\*(C`LC_CTYPE\*(C'\fR. +.PP +.Vb 1 +\& printf '\ee]701;%s\e007' ja_JP.SJIS +.Ve +.PP +See also the previous answer. +.PP +Sometimes this capability is rather handy when you want to work in +one locale (e.g. \f(CW\*(C`de_DE.UTF\-8\*(C'\fR) but some programs don't support it +(e.g. \s-1UTF\-8\s0). For example, I use this script to start \f(CW\*(C`xjdic\*(C'\fR, which +first switches to a locale supported by xjdic and back later: +.PP +.Vb 3 +\& printf '\ee]701;%s\e007' ja_JP.SJIS +\& xjdic -js +\& printf '\ee]701;%s\e007' de_DE.UTF-8 +.Ve +.PP +You can also use xterm's \f(CW\*(C`luit\*(C'\fR program, which usually works fine, except +for some locales where character width differs between program\- and +rxvt\-unicode\-locales. +.PP +\fIMy input method wants but I want \s-1UTF\-8\s0, what can I do?\fR +.IX Subsection "My input method wants but I want UTF-8, what can I do?" +.PP +You can specify separate locales for the input method and the rest of the +terminal, using the resource \f(CW\*(C`imlocale\*(C'\fR: +.PP +.Vb 1 +\& URxvt.imlocale: ja_JP.EUC-JP +.Ve +.PP +Now you can start your terminal with \f(CW\*(C`LC_CTYPE=ja_JP.UTF\-8\*(C'\fR and still +use your input method. Please note, however, that you will not be able to +input characters outside \f(CW\*(C`EUC\-JP\*(C'\fR in a normal way then, as your input +method limits you. +.PP +\fIRxvt-unicode crashes when the X Input Method changes or exits.\fR +.IX Subsection "Rxvt-unicode crashes when the X Input Method changes or exits." +.PP +Unfortunately, this is unavoidable, as the \s-1XIM\s0 protocol is racy by +design. Applications can avoid some crashes at the expense of memory +leaks, and Input Methods can avoid some crashes by careful ordering at +exit time. \fBkinput2\fR (and derived input methods) generally succeeds, +while \fB\s-1SCIM\s0\fR (or similar input methods) fails. In the end, however, +crashes cannot be completely avoided even if both sides cooperate. +.PP +So the only workaround is not to kill your Input Method Servers. +.Sh "Operating Systems / Package Maintaining" +.IX Subsection "Operating Systems / Package Maintaining" +\fII am using Debian GNU/Linux and have a problem...\fR +.IX Subsection "I am using Debian GNU/Linux and have a problem..." +.PP +The Debian GNU/Linux package of rxvt-unicode in sarge contains large +patches that considerably change the behaviour of rxvt-unicode (but +unfortunately this notice has been removed). Before reporting a bug to +the original rxvt-unicode author please download and install the genuine +version () and try to reproduce +the problem. If you cannot, chances are that the problems are specific to +Debian GNU/Linux, in which case it should be reported via the Debian Bug +Tracking System (use \f(CW\*(C`reportbug\*(C'\fR to report the bug). +.PP +For other problems that also affect the Debian package, you can and +probably should use the Debian \s-1BTS\s0, too, because, after all, it's also a +bug in the Debian version and it serves as a reminder for other users that +might encounter the same issue. +.PP +\fII am maintaining rxvt-unicode for distribution/OS \s-1XXX\s0, any recommendation?\fR +.IX Subsection "I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation?" +.PP +You should build one binary with the default options. \fIconfigure\fR +now enables most useful options, and the trend goes to making them +runtime\-switchable, too, so there is usually no drawback to enbaling them, +except higher disk and possibly memory usage. The perl interpreter should +be enabled, as important functionality (menus, selection, likely more in +the future) depends on it. +.PP +You should not overwrite the \f(CW\*(C`perl\-ext\-common\*(C'\fR snd \f(CW\*(C`perl\-ext\*(C'\fR resources +system-wide (except maybe with \f(CW\*(C`defaults\*(C'\fR). This will result in useful +behaviour. If your distribution aims at low memory, add an empty +\&\f(CW\*(C`perl\-ext\-common\*(C'\fR resource to the app-defaults file. This will keep the +perl interpreter disabled until the user enables it. +.PP +If you can/want build more binaries, I recommend building a minimal +one with \f(CW\*(C`\-\-disable\-everything\*(C'\fR (very useful) and a maximal one with +\&\f(CW\*(C`\-\-enable\-everything\*(C'\fR (less useful, it will be very big due to a lot of +encodings built-in that increase download times and are rarely used). +.PP +\fII need to make it setuid/setgid to support utmp/ptys on my \s-1OS\s0, is this safe?\fR +.IX Subsection "I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe?" +.PP +It should be, starting with release 7.1. You are encouraged to properly +install urxvt with privileges necessary for your \s-1OS\s0 now. +.PP +When rxvt-unicode detects that it runs setuid or setgid, it will fork +into a helper process for privileged operations (pty handling on some +systems, utmp/wtmp/lastlog handling on others) and drop privileges +immediately. This is much safer than most other terminals that keep +privileges while running (but is more relevant to urxvt, as it contains +things as perl interpreters, which might be \*(L"helpful\*(R" to attackers). +.PP +This forking is done as the very first within \fImain()\fR, which is very early +and reduces possible bugs to initialisation code run before \fImain()\fR, or +things like the dynamic loader of your system, which should result in very +little risk. +.PP +\fIOn Solaris 9, many line-drawing characters are too wide.\fR +.IX Subsection "On Solaris 9, many line-drawing characters are too wide." +.PP +Seems to be a known bug, read +. Some people use the +following ugly workaround to get non-double-wide-characters working: +.PP +.Vb 1 +\& #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) +.Ve +.PP +\fII am on FreeBSD and rxvt-unicode does not seem to work at all.\fR +.IX Subsection "I am on FreeBSD and rxvt-unicode does not seem to work at all." +.PP +Rxvt-unicode requires the symbol \f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR to be defined +in your compile environment, or an implementation that implements it, +wether it defines the symbol or not. \f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR requires that +\&\fBwchar_t\fR is represented as unicode. +.PP +As you might have guessed, FreeBSD does neither define this symobl nor +does it support it. Instead, it uses it's own internal representation of +\&\fBwchar_t\fR. This is, of course, completely fine with respect to standards. +.PP +However, that means rxvt-unicode only works in \f(CW\*(C`POSIX\*(C'\fR, \f(CW\*(C`ISO\-8859\-1\*(C'\fR and +\&\f(CW\*(C`UTF\-8\*(C'\fR locales under FreeBSD (which all use Unicode as \fBwchar_t\fR. +.PP +\&\f(CW\*(C`_\|_STDC_ISO_10646_\|_\*(C'\fR is the only sane way to support multi-language +apps in an \s-1OS\s0, as using a locale-dependent (and non\-standardized) +representation of \fBwchar_t\fR makes it impossible to convert between +\&\fBwchar_t\fR (as used by X11 and your applications) and any other encoding +without implementing OS-specific-wrappers for each and every locale. There +simply are no APIs to convert \fBwchar_t\fR into anything except the current +locale encoding. +.PP +Some applications (such as the formidable \fBmlterm\fR) work around this +by carrying their own replacement functions for character set handling +with them, and either implementing OS-dependent hacks or doing multiple +conversions (which is slow and unreliable in case the \s-1OS\s0 implements +encodings slightly different than the terminal emulator). +.PP +The rxvt-unicode author insists that the right way to fix this is in the +system libraries once and for all, instead of forcing every app to carry +complete replacements for them :) +.PP +\fII use Solaris 9 and it doesn't compile/work/etc.\fR +.IX Subsection "I use Solaris 9 and it doesn't compile/work/etc." +.PP +Try the diff in \fIdoc/solaris9.patch\fR as a base. It fixes the worst +problems with \f(CW\*(C`wcwidth\*(C'\fR and a compile problem. +.PP +\fIHow can I use rxvt-unicode under cygwin?\fR +.IX Subsection "How can I use rxvt-unicode under cygwin?" +.PP +rxvt-unicode should compile and run out of the box on cygwin, using +the X11 libraries that come with cygwin. libW11 emulation is no +longer supported (and makes no sense, either, as it only supported a +single font). I recommend starting the X\-server in \f(CW\*(C`\-multiwindow\*(C'\fR or +\&\f(CW\*(C`\-rootless\*(C'\fR mode instead, which will result in similar look&feel as the +old libW11 emulation. +.PP +At the time of this writing, cygwin didn't seem to support any multi-byte +encodings (you might try \f(CW\*(C`LC_CTYPE=C\-UTF\-8\*(C'\fR), so you are likely limited +to 8\-bit encodings. .SH "RXVT TECHNICAL REFERENCE" .IX Header "RXVT TECHNICAL REFERENCE" .SH "DESCRIPTION" diff --git a/doc/rxvt.7.pod b/doc/rxvt.7.pod index 0d0c9514..20d055af 100644 --- a/doc/rxvt.7.pod +++ b/doc/rxvt.7.pod @@ -23,133 +23,97 @@ L. =head1 FREQUENTLY ASKED QUESTIONS -=head2 The new selection selects pieces that are too big, how can I select -single words? -If you want to select e.g. alphanumeric words, you can use the following -setting: - - URxvt.selection.pattern-0: ([[:word:]]+) - -If you click more than twice, the selection will be extended -more and more. - -To get a selection that is very similar to the old code, try this pattern: - - URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+) - -Please also note that the I combination also -selects words like the old code. - -=head2 I don't like the new selection/popups/hotkeys/perl, how do I -change/disable it? - -You can disable the perl extension completely by setting the -B resource to the empty string, which also keeps -rxvt-unicode from initialising perl, saving memory. - -If you only want to disable specific features, you first have to -identify which perl extension is responsible. For this, read the section -B in the @@RXVT_NAME@@perl(3) manpage. For -example, to disable the B and B, specify -this B resource: - - URxvt.perl-ext-common: default,-selection-popup,-option-popup - -This will keep the default extensions, but disable the two popup -extensions. Some extensions can also be configured, for example, -scrollback search mode is triggered by B. You can move it to any -other combination either by setting the B resource: +=head2 Meta, Features & Commandline Issues - URxvt.searchable-scrollback: CM-s - -=head3 The cursor moves when selecting text in the current input line, how -do I switch this off? - -See next entry. - -=head2 During rlogin/ssh/telnet/etc. sessions, clicking near the cursor -outputs strange escape sequences, how do I fix this? - -These are caused by the C perl extension. Under normal -circumstances, it will move your cursor around when you click into the -line that contains it. It tries hard not to do this at the wrong moment, -but when running a program that doesn't parse cursor movements or in some -cases during rlogin sessions, it fails to detect this properly. - -You can permamently switch this feature off by disabling the C -extension: - - URxvt.perl-ext-common: default,-readline +=head3 My question isn't answered here, can I ask a human? -=head2 Why doesn't rxvt-unicode read my resources? +Before sending me mail, you could go to IRC: C, +channel C<#rxvt-unicode> has some rxvt-unicode enthusiasts that might be +interested in learning about new and exciting problems (but not FAQs :). -Well, why, indeed? It does, in a way very similar to other X -applications. Most importantly, this means that if you or your OS loads -resources into the X display (the right way to do it), rxvt-unicode will -ignore any resource files in your home directory. It will only read -F<$HOME/.Xdefaults> when no resources are attached to the display. +=head3 Does it support tabs, can I have a tabbed rxvt-unicode? -If you have or use an F<$HOME/.Xresources> file, chances are that -resources are loaded into your X-server. In this case, you have to -re-login after every change (or run F). +Beginning with version 7.3, there is a perl extension that implements a +simple tabbed terminal. It is installed by default, so any of these should +give you tabs: -Also consider the form resources have to use: + @@RXVT_NAME@@ -pe tabbed - URxvt.resource: value + URxvt.perl-ext-common: default,tabbed -If you want to use another form (there are lots of different ways of -specifying resources), make sure you understand wether and why it -works. If unsure, use the form above. +It will also work fine with tabbing functionality of many window managers +or similar tabbing programs, and its embedding-features allow it to be +embedded into other programs, as witnessed by F or +the upcoming C perl module, which features a tabbed urxvt +(murxvt) terminal as an example embedding application. -=head2 I can't get transparency working, what am I doing wrong? +=head3 How do I know which rxvt-unicode version I'm using? -First of all, transparency isn't officially supported in rxvt-unicode, so -you are mostly on your own. Do not bug the author about it (but you may -bug everybody else). Also, if you can't get it working consider it a rite -of passage: ... and you failed. +The version number is displayed with the usage (-h). Also the escape +sequence C sets the window title to the version number. When +using the @@RXVT_NAME@@c client, the version displayed is that of the +daemon. -Here are four ways to get transparency. B read the manpage and option -descriptions for the programs mentioned and rxvt-unicode. Really, do it! +=head3 Rxvt-unicode uses gobs of memory, how can I reduce that? -1. Use inheritPixmap: +Rxvt-unicode tries to obey the rule of not charging you for something you +don't use. One thing you should try is to configure out all settings that +you don't need, for example, Xft support is a resource hog by design, +when used. Compiling it out ensures that no Xft font will be loaded +accidentally when rxvt-unicode tries to find a font for your characters. - Esetroot wallpaper.jpg - @@RXVT_NAME@@ -ip -tint red -sh 40 +Also, many people (me included) like large windows and even larger +scrollback buffers: Without C<--enable-unicode3>, rxvt-unicode will use +6 bytes per screen cell. For a 160x?? window this amounts to almost a +kilobyte per line. A scrollback buffer of 10000 lines will then (if full) +use 10 Megabytes of memory. With C<--enable-unicode3> it gets worse, as +rxvt-unicode then uses 8 bytes per screen cell. -That works. If you think it doesn't, you lack transparency and tinting -support, or you are unable to read. +=head3 How can I start @@RXVT_NAME@@d in a race-free way? -2. Use a simple pixmap and emulate pseudo-transparency. This enables you -to use effects other than tinting and shading: Just shade/tint/whatever -your picture with gimp: +Try C<@@RXVT_NAME@@d -f -o>, which tells @@RXVT_NAME@@d to open the +display, create the listening socket and then fork. - convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm - @@RXVT_NAME@@ -pixmap background.xpm -pe automove-background +=head3 How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc. -That works. If you think it doesn't, you lack XPM and Perl support, or you -are unable to read. +rxvt and rxvt-unicode always export the variable "COLORTERM", so you can +check and see if that is set. Note that several programs, JED, slrn, +Midnight Commander automatically check this variable to decide whether or +not to use color. -3. Use an ARGB visual: +=head3 How do I set the correct, full IP address for the DISPLAY variable? - @@RXVT_NAME@@ -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc +If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled +insecure mode then it is possible to use the following shell script +snippets to correctly set the display. If your version of rxvt-unicode +wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) then +the COLORTERM variable can be used to distinguish rxvt-unicode from a +regular xterm. -This requires XFT support, and the support of your X-server. If that -doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't -there yet, no matter what they claim. Rxvt-Unicode contains the neccessary -bugfixes and workarounds for Xft and Xlib to make it work, but that -doesn't mean that your WM has the required kludges in place. +Courtesy of Chuck Blake with the following shell script +snippets: -4. Use xcompmgr and let it do the job: + # Bourne/Korn/POSIX family of shells: + [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know + if [ ${TERM:-foo} = xterm ]; then + stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not + echo -n '^[Z' + read term_id + stty icanon echo + if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then + echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string + read DISPLAY # set it in our local shell + fi + fi - xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ - -set _NET_WM_WINDOW_OPACITY 0xc0000000 +=head3 How do I compile the manual pages on my own? -Then click on a window you want to make transparent. Replace C<0xc0000000> -by other values to change the degree of opacity. If it doesn't work and -your server crashes, you got to keep the pieces. +You need to have a recent version of perl installed as F, +one that comes with F, F and F. Then go to +the doc subdirectory and enter C. -=head2 Isn't rxvt supposed to be small? Don't all those features bloat? +=head3 Isn't rxvt-unicode supposed to be small? Don't all those features bloat? I often get asked about this, and I think, no, they didn't cause extra bloat. If you compare a minimal rxvt and a minimal urxvt, you can see @@ -191,7 +155,7 @@ still fares rather well. And compared to some monsters like gnome-terminal startup time, including the hundreds of warnings it spits out), it fares extremely well *g*. -=head2 Why C++, isn't that unportable/bloated/uncool? +=head3 Why C++, isn't that unportable/bloated/uncool? Is this a question? :) It comes up very often. The simple answer is: I had to write it, and C++ allowed me to write and maintain it in a fraction @@ -225,321 +189,329 @@ And here is rxvt-unicode: No large bloated libraries (of course, none were linked in statically), except maybe libX11 :) -=head2 Does it support tabs, can I have a tabbed rxvt-unicode? -Beginning with version 7.3, there is a perl extension that implements a -simple tabbed terminal. It is installed by default, so any of these should -give you tabs: +=head2 Rendering, Font & Look and Feel Issues - @@RXVT_NAME@@ -pe tabbed +=head3 I can't get transparency working, what am I doing wrong? - URxvt.perl-ext-common: default,tabbed +First of all, transparency isn't officially supported in rxvt-unicode, so +you are mostly on your own. Do not bug the author about it (but you may +bug everybody else). Also, if you can't get it working consider it a rite +of passage: ... and you failed. -It will also work fine with tabbing functionality of many window managers -or similar tabbing programs, and its embedding-features allow it to be -embedded into other programs, as witnessed by F or -the upcoming C perl module, which features a tabbed urxvt -(murxvt) terminal as an example embedding application. +Here are four ways to get transparency. B read the manpage and option +descriptions for the programs mentioned and rxvt-unicode. Really, do it! -=head2 How do I know which rxvt-unicode version I'm using? +1. Use inheritPixmap: -The version number is displayed with the usage (-h). Also the escape -sequence C sets the window title to the version number. When -using the @@RXVT_NAME@@c client, the version displayed is that of the -daemon. + Esetroot wallpaper.jpg + @@RXVT_NAME@@ -ip -tint red -sh 40 -=head2 I am using Debian GNU/Linux and have a problem... +That works. If you think it doesn't, you lack transparency and tinting +support, or you are unable to read. -The Debian GNU/Linux package of rxvt-unicode in sarge contains large -patches that considerably change the behaviour of rxvt-unicode (but -unfortunately this notice has been removed). Before reporting a bug to -the original rxvt-unicode author please download and install the genuine -version (L) and try to reproduce -the problem. If you cannot, chances are that the problems are specific to -Debian GNU/Linux, in which case it should be reported via the Debian Bug -Tracking System (use C to report the bug). +2. Use a simple pixmap and emulate pseudo-transparency. This enables you +to use effects other than tinting and shading: Just shade/tint/whatever +your picture with gimp or any other tool: -For other problems that also affect the Debian package, you can and -probably should use the Debian BTS, too, because, after all, it's also a -bug in the Debian version and it serves as a reminder for other users that -might encounter the same issue. + convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm + @@RXVT_NAME@@ -pixmap background.xpm -pe automove-background -=head2 I am maintaining rxvt-unicode for distribution/OS XXX, any -recommendation? +That works. If you think it doesn't, you lack XPM and Perl support, or you +are unable to read. -You should build one binary with the default options. F -now enables most useful options, and the trend goes to making them -runtime-switchable, too, so there is usually no drawback to enbaling them, -except higher disk and possibly memory usage. The perl interpreter should -be enabled, as important functionality (menus, selection, likely more in -the future) depends on it. +3. Use an ARGB visual: -You should not overwrite the C snd C resources -system-wide (except maybe with C). This will result in useful -behaviour. If your distribution aims at low memory, add an empty -C resource to the app-defaults file. This will keep the -perl interpreter disabled until the user enables it. + @@RXVT_NAME@@ -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc -If you can/want build more binaries, I recommend building a minimal -one with C<--disable-everything> (very useful) and a maximal one with -C<--enable-everything> (less useful, it will be very big due to a lot of -encodings built-in that increase download times and are rarely used). +This requires XFT support, and the support of your X-server. If that +doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't +there yet, no matter what they claim. Rxvt-Unicode contains the neccessary +bugfixes and workarounds for Xft and Xlib to make it work, but that +doesn't mean that your WM has the required kludges in place. -=head2 I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe? +4. Use xcompmgr and let it do the job: -It should be, starting with release 7.1. You are encouraged to properly -install urxvt with privileges necessary for your OS now. + xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ + -set _NET_WM_WINDOW_OPACITY 0xc0000000 -When rxvt-unicode detects that it runs setuid or setgid, it will fork -into a helper process for privileged operations (pty handling on some -systems, utmp/wtmp/lastlog handling on others) and drop privileges -immediately. This is much safer than most other terminals that keep -privileges while running (but is more relevant to urxvt, as it contains -things as perl interpreters, which might be "helpful" to attackers). +Then click on a window you want to make transparent. Replace C<0xc0000000> +by other values to change the degree of opacity. If it doesn't work and +your server crashes, you got to keep the pieces. -This forking is done as the very first within main(), which is very early -and reduces possible bugs to initialisation code run before main(), or -things like the dynamic loader of your system, which should result in very -little risk. +=head3 Why do some chinese characters look so different than others? -=head2 When I log-in to another system it tells me about missing terminfo data? +This is because there is a difference between script and language -- +rxvt-unicode does not know which language the text that is output is, +as it only knows the unicode character codes. If rxvt-unicode first +sees a japanese/chinese character, it might choose a japanese font for +display. Subsequent japanese characters will use that font. Now, many +chinese characters aren't represented in japanese fonts, so when the first +non-japanese character comes up, rxvt-unicode will look for a chinese font +-- unfortunately at this point, it will still use the japanese font for +chinese characters that are also in the japanese font. -The terminal description used by rxvt-unicode is not as widely available -as that for xterm, or even rxvt (for which the same problem often arises). +The workaround is easy: just tag a chinese font at the end of your font +list (see the previous question). The key is to view the font list as +a preference list: If you expect more japanese, list a japanese font +first. If you expect more chinese, put a chinese font first. -The correct solution for this problem is to install the terminfo, this can -be done like this (with ncurses' infocmp): +In the future it might be possible to switch language preferences at +runtime (the internal data structure has no problem with using different +fonts for the same character at the same time, but no interface for this +has been designed yet). - REMOTE=remotesystem.domain - infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" +Until then, you might get away with switching fonts at runtime (see L later in this document). -... or by installing rxvt-unicode normally on the remote system, +=head3 Why does rxvt-unicode sometimes leave pixel droppings? -If you cannot or do not want to do this, then you can simply set -C or even C, and live with the small number of -problems arising, which includes wrong keymapping, less and different -colours and some refresh errors in fullscreen applications. It's a nice -quick-and-dirty workaround for rare cases, though. - -If you always want to do this (and are fine with the consequences) you -can either recompile rxvt-unicode with the desired TERM value or use a -resource to set it: - - URxvt.termName: rxvt - -If you don't plan to use B (quite common...) you could also replace -the rxvt terminfo file with the rxvt-unicode one. - -=head2 C outputs some error when compiling the terminfo entry. +Most fonts were not designed for terminal use, which means that character +size varies a lot. A font that is otherwise fine for terminal use might +contain some characters that are simply too wide. Rxvt-unicode will avoid +these characters. For characters that are just "a bit" too wide a special +"careful" rendering mode is used that redraws adjacent characters. -Most likely it's the empty definition for C. Just replace it by -C and try again. +All of this requires that fonts do not lie about character sizes, +however: Xft fonts often draw glyphs larger than their acclaimed bounding +box, and rxvt-unicode has no way of detecting this (the correct way is to +ask for the character bounding box, which unfortunately is wrong in these +cases). -=head2 C's readline does not work correctly under @@RXVT_NAME@@. +It's not clear (to me at least), wether this is a bug in Xft, freetype, +or the respective font. If you encounter this problem you might try using +the C<-lsp> option to give the font more height. If that doesn't work, you +might be forced to use a different font. -See next entry. +All of this is not a problem when using X11 core fonts, as their bounding +box data is correct. -=head2 I need a termcap file entry. +=head3 How can I keep rxvt-unicode from using reverse video so much? -One reason you might want this is that some distributions or operating -systems still compile some programs using the long-obsoleted termcap -library (Fedora Core's bash is one example) and rely on a termcap entry -for C. +First of all, make sure you are running with the right terminal settings +(C), which will get rid of most of these effects. Then +make sure you have specified colours for italic and bold, as otherwise +rxvt-unicode might use reverse video to simulate the effect: -You could use rxvt's termcap entry with resonable results in many cases. -You can also create a termcap entry by using terminfo's infocmp program -like this: + URxvt.colorBD: white + URxvt.colorIT: green - infocmp -C rxvt-unicode +=head3 Some programs assume totally weird colours (red instead of blue), how can I fix that? -Or you could use this termcap entry, generated by the command above: +For some unexplainable reason, some rare programs assume a very weird +colour palette when confronted with a terminal with more than the standard +8 colours (rxvt-unicode supports 88). The right fix is, of course, to fix +these programs not to assume non-ISO colours without very good reasons. - rxvt-unicode|rxvt-unicode terminal (X Window System):\ - :am:bw:eo:km:mi:ms:xn:xo:\ - :co#80:it#8:li#24:lm#0:\ - :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ - :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ - :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ - :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ - :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ - :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ - :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ - :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ - :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ - :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ - :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ - :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ - :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ - :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ - :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ - :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ - :vs=\E[?25h: +In the meantime, you can either edit your C terminfo +definition to only claim 8 colour support or use C, which will +fix colours but keep you from using other rxvt-unicode features. -=head2 Why does C no longer have coloured output? +=head3 Can I switch the fonts at runtime? -The C in the GNU coreutils unfortunately doesn't use terminfo to -decide wether a terminal has colour, but uses it's own configuration -file. Needless to say, C is not in it's default file (among -with most other terminals supporting colour). Either add: +Yes, using an escape sequence. Try something like this, which has the same +effect as using the C<-fn> switch, and takes effect immediately: - TERM rxvt-unicode + printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" -to C or simply add: +This is useful if you e.g. work primarily with japanese (and prefer a +japanese font), but you have to switch to chinese temporarily, where +japanese fonts would only be in your way. - alias ls='ls --color=auto' +You can think of this as a kind of manual ISO-2022 switching. -to your C<.profile> or C<.bashrc>. +=head3 Why do italic characters look as if clipped? -=head2 Why doesn't vim/emacs etc. use the 88 colour mode? +Many fonts have difficulties with italic characters and hinting. For +example, the otherwise very nicely hinted font C completely fails in it's italic face. A workaround might be to +enable freetype autohinting, i.e. like this: -See next entry. + URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true + URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true -=head2 Why doesn't vim/emacs etc. make use of italic? +=head3 Can I speed up Xft rendering somehow? -See next entry. +Yes, the most obvious way to speed it up is to avoid Xft entirely, as +it is simply slow. If you still want Xft fonts you might try to disable +antialiasing (by appending C<:antialias=false>), which saves lots of +memory and also speeds up rendering considerably. -=head2 Why are the secondary screen-related options not working properly? +=head3 Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? -Make sure you are using C. Some pre-packaged -distributions (most notably Debian GNU/Linux) break rxvt-unicode -by setting C to C, which doesn't have these extra -features. Unfortunately, some of these (most notably, again, Debian -GNU/Linux) furthermore fail to even install the C terminfo -file, so you will need to install it on your own (See the question B on -how to do this). +Rxvt-unicode will use whatever you specify as a font. If it needs to +fall back to it's default font search list it will prefer X11 core +fonts, because they are small and fast, and then use Xft fonts. It has +antialiasing disabled for most of them, because the author thinks they +look best that way. -=head2 My numerical keypad acts weird and generates differing output? +If you want antialiasing, you have to specify the fonts manually. -Some Debian GNUL/Linux users seem to have this problem, although no -specific details were reported so far. It is possible that this is caused -by the wrong C setting, although the details of wether and how -this can happen are unknown, as C should offer a compatible -keymap. See the answer to the previous question, and please report if that -helped. +=head3 What's with this bold/blink stuff? -=head2 Rxvt-unicode does not seem to understand the selected encoding? +If no bold colour is set via C, bold will invert text using the +standard foreground colour. -See next entry. +For the standard background colour, blinking will actually make the +text blink when compiled with C<--enable-blinking>. with standard +colours. Without C<--enable-blinking>, the blink attribute will be +ignored. -=head2 Unicode does not seem to work? +On ANSI colours, bold/blink attributes are used to set high-intensity +foreground/background colors. -If you encounter strange problems like typing an accented character but -getting two unrelated other characters or similar, or if program output is -subtly garbled, then you should check your locale settings. +color0-7 are the low-intensity colors. -Rxvt-unicode must be started with the same C setting as the -programs. Often rxvt-unicode is started in the C locale, while the -login script running within the rxvt-unicode window changes the locale to -something else, e.g. C. Needless to say, this is not going to work. +color8-15 are the corresponding high-intensity colors. -The best thing is to fix your startup environment, as you will likely run -into other problems. If nothing works you can try this in your .profile. +=head3 I don't like the screen colors. How do I change them? - printf '\e]701;%s\007' "$LC_CTYPE" +You can change the screen colors at run-time using F<~/.Xdefaults> +resources (or as long-options). -If this doesn't work, then maybe you use a C specification not -supported on your systems. Some systems have a C command which -displays this (also, C can be used to check locale settings, as -it will complain loudly if it cannot set the locale). If it displays something -like: +Here are values that are supposed to resemble a VGA screen, +including the murky brown that passes for low-intensity yellow: - locale: Cannot set LC_CTYPE to default locale: ... + URxvt.color0: #000000 + URxvt.color1: #A80000 + URxvt.color2: #00A800 + URxvt.color3: #A8A800 + URxvt.color4: #0000A8 + URxvt.color5: #A800A8 + URxvt.color6: #00A8A8 + URxvt.color7: #A8A8A8 -Then the locale you specified is not supported on your system. + URxvt.color8: #000054 + URxvt.color9: #FF0054 + URxvt.color10: #00FF54 + URxvt.color11: #FFFF54 + URxvt.color12: #0000FF + URxvt.color13: #FF00FF + URxvt.color14: #00FFFF + URxvt.color15: #FFFFFF -If nothing works and you are sure that everything is set correctly then -you will need to remember a little known fact: Some programs just don't -support locales :( +And here is a more complete set of non-standard colors described (not by +me) as "pretty girly". -=head2 Why do some characters look so much different than others? + URxvt.cursorColor: #dc74d1 + URxvt.pointerColor: #dc74d1 + URxvt.background: #0e0e0e + URxvt.foreground: #4ad5e1 + URxvt.color0: #000000 + URxvt.color8: #8b8f93 + URxvt.color1: #dc74d1 + URxvt.color9: #dc74d1 + URxvt.color2: #0eb8c7 + URxvt.color10: #0eb8c7 + URxvt.color3: #dfe37e + URxvt.color11: #dfe37e + URxvt.color5: #9e88f0 + URxvt.color13: #9e88f0 + URxvt.color6: #73f7ff + URxvt.color14: #73f7ff + URxvt.color7: #e1dddd + URxvt.color15: #e1dddd +=head3 Why do some characters look so much different than others? + See next entry. - -=head2 How does rxvt-unicode choose fonts? - + +=head3 How does rxvt-unicode choose fonts? + Most fonts do not contain the full range of Unicode, which is fine. Chances are that the font you (or the admin/package maintainer of your system/os) have specified does not cover all the characters you want to display. - + B makes a best-effort try at finding a replacement font. Often the result is fine, but sometimes the chosen font looks bad/ugly/wrong. Some fonts have totally strange characters that don't resemble the correct glyph at all, and rxvt-unicode lacks the artificial intelligence to detect that a specific glyph is wrong: it has to believe the font that the characters it claims to contain indeed look correct. - + In that case, select a font of your taste and add it to the font list, e.g.: - + @@RXVT_NAME@@ -fn basefont,font2,font3... - + When rxvt-unicode sees a character, it will first look at the base font. If the base font does not contain the character, it will go to the next font, and so on. Specifying your own fonts will also speed up this search and use less resources within rxvt-unicode and the X-server. - + The only limitation is that none of the fonts may be larger than the base font, as the base font defines the terminal character cell size, which must be the same due to the way terminals work. -=head2 Why do some chinese characters look so different than others? -This is because there is a difference between script and language -- -rxvt-unicode does not know which language the text that is output is, -as it only knows the unicode character codes. If rxvt-unicode first -sees a japanese/chinese character, it might choose a japanese font for -display. Subsequent japanese characters will use that font. Now, many -chinese characters aren't represented in japanese fonts, so when the first -non-japanese character comes up, rxvt-unicode will look for a chinese font --- unfortunately at this point, it will still use the japanese font for -chinese characters that are also in the japanese font. +=head2 Keyboard, Mouse & User Interaction -The workaround is easy: just tag a chinese font at the end of your font -list (see the previous question). The key is to view the font list as -a preference list: If you expect more japanese, list a japanese font -first. If you expect more chinese, put a chinese font first. +=head3 The new selection selects pieces that are too big, how can I select single words? -In the future it might be possible to switch language preferences at -runtime (the internal data structure has no problem with using different -fonts for the same character at the same time, but no interface for this -has been designed yet). +If you want to select e.g. alphanumeric words, you can use the following +setting: -Until then, you might get away with switching fonts at runtime (see L later in this document). + URxvt.selection.pattern-0: ([[:word:]]+) -=head2 Why does rxvt-unicode sometimes leave pixel droppings? +If you click more than twice, the selection will be extended +more and more. -Most fonts were not designed for terminal use, which means that character -size varies a lot. A font that is otherwise fine for terminal use might -contain some characters that are simply too wide. Rxvt-unicode will avoid -these characters. For characters that are just "a bit" too wide a special -"careful" rendering mode is used that redraws adjacent characters. +To get a selection that is very similar to the old code, try this pattern: -All of this requires that fonts do not lie about character sizes, -however: Xft fonts often draw glyphs larger than their acclaimed bounding -box, and rxvt-unicode has no way of detecting this (the correct way is to -ask for the character bounding box, which unfortunately is wrong in these -cases). + URxvt.selection.pattern-0: ([^"&'()*,;<=>?@[\\\\]^`{|})]+) -It's not clear (to me at least), wether this is a bug in Xft, freetype, -or the respective font. If you encounter this problem you might try using -the C<-lsp> option to give the font more height. If that doesn't work, you -might be forced to use a different font. +Please also note that the I combination also +selects words like the old code. -All of this is not a problem when using X11 core fonts, as their bounding -box data is correct. +=head3 I don't like the new selection/popups/hotkeys/perl, how do I change/disable it? -=head2 On Solaris 9, many line-drawing characters are too wide. +You can disable the perl extension completely by setting the +B resource to the empty string, which also keeps +rxvt-unicode from initialising perl, saving memory. -Seems to be a known bug, read -L. Some people use the -following ugly workaround to get non-double-wide-characters working: +If you only want to disable specific features, you first have to +identify which perl extension is responsible. For this, read the section +B in the @@RXVT_NAME@@perl(3) manpage. For +example, to disable the B and B, specify +this B resource: - #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) + URxvt.perl-ext-common: default,-selection-popup,-option-popup + +This will keep the default extensions, but disable the two popup +extensions. Some extensions can also be configured, for example, +scrollback search mode is triggered by B. You can move it to any +other combination either by setting the B resource: + + URxvt.searchable-scrollback: CM-s + +=head3 The cursor moves when selecting text in the current input line, how do I switch this off? + +See next entry. + +=head3 During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this? + +These are caused by the C perl extension. Under normal +circumstances, it will move your cursor around when you click into the +line that contains it. It tries hard not to do this at the wrong moment, +but when running a program that doesn't parse cursor movements or in some +cases during rlogin sessions, it fails to detect this properly. + +You can permamently switch this feature off by disabling the C +extension: + + URxvt.perl-ext-common: default,-readline + +=head3 My numerical keypad acts weird and generates differing output? -=head2 My Compose (Multi_key) key is no longer working. +Some Debian GNUL/Linux users seem to have this problem, although no +specific details were reported so far. It is possible that this is caused +by the wrong C setting, although the details of wether and how +this can happen are unknown, as C should offer a compatible +keymap. See the answer to the previous question, and please report if that +helped. + +=head3 My Compose (Multi_key) key is no longer working. The most common causes for this are that either your locale is not set correctly, or you specified a B that is not supported by @@ -551,7 +523,7 @@ rxvt-unicode will continue without an input method. In this case either do not specify a B or specify more than one pre-edit style, such as B. -=head2 I cannot type C to get an ASCII NUL character due to ISO 14755 +=head3 I cannot type C to get an ASCII NUL character due to ISO 14755 Either try C alone (it often is mapped to ASCII NUL even on international keyboards) or simply use ISO 14755 support to your @@ -559,440 +531,479 @@ advantage, typing to get a ASCII NUL. This works for other codes, too, such as C to type the default telnet escape character and so on. -=head2 How can I keep rxvt-unicode from using reverse video so much? +=head3 Mouse cut/paste suddenly no longer works. -First of all, make sure you are running with the right terminal settings -(C), which will get rid of most of these effects. Then -make sure you have specified colours for italic and bold, as otherwise -rxvt-unicode might use reverse video to simulate the effect: +Make sure that mouse reporting is actually turned off since killing +some editors prematurely may leave the mouse in mouse report mode. I've +heard that tcsh may use mouse reporting unless it otherwise specified. A +quick check is to see if cut/paste works when the Alt or Shift keys are +depressed. - URxvt.colorBD: white - URxvt.colorIT: green +=head3 What's with the strange Backspace/Delete key behaviour? -=head2 Some programs assume totally weird colours (red instead of blue), how can I fix that? +Assuming that the physical Backspace key corresponds to the +BackSpace keysym (not likely for Linux ... see the following +question) there are two standard values that can be used for +Backspace: C<^H> and C<^?>. -For some unexplainable reason, some rare programs assume a very weird -colour palette when confronted with a terminal with more than the standard -8 colours (rxvt-unicode supports 88). The right fix is, of course, to fix -these programs not to assume non-ISO colours without very good reasons. +Historically, either value is correct, but rxvt-unicode adopts the debian +policy of using C<^?> when unsure, because it's the one only only correct +choice :). -In the meantime, you can either edit your C terminfo -definition to only claim 8 colour support or use C, which will -fix colours but keep you from using other rxvt-unicode features. +Rxvt-unicode tries to inherit the current stty settings and uses the value +of `erase' to guess the value for backspace. If rxvt-unicode wasn't +started from a terminal (say, from a menu or by remote shell), then the +system value of `erase', which corresponds to CERASE in , will +be used (which may not be the same as your stty setting). -=head2 I am on FreeBSD and rxvt-unicode does not seem to work at all. +For starting a new rxvt-unicode: -Rxvt-unicode requires the symbol C<__STDC_ISO_10646__> to be defined -in your compile environment, or an implementation that implements it, -wether it defines the symbol or not. C<__STDC_ISO_10646__> requires that -B is represented as unicode. + # use Backspace = ^H + $ stty erase ^H + $ @@RXVT_NAME@@ -As you might have guessed, FreeBSD does neither define this symobl nor -does it support it. Instead, it uses it's own internal representation of -B. This is, of course, completely fine with respect to standards. + # use Backspace = ^? + $ stty erase ^? + $ @@RXVT_NAME@@ -However, that means rxvt-unicode only works in C, C and -C locales under FreeBSD (which all use Unicode as B. +Toggle with C / C. -C<__STDC_ISO_10646__> is the only sane way to support multi-language -apps in an OS, as using a locale-dependent (and non-standardized) -representation of B makes it impossible to convert between -B (as used by X11 and your applications) and any other encoding -without implementing OS-specific-wrappers for each and every locale. There -simply are no APIs to convert B into anything except the current -locale encoding. +For an existing rxvt-unicode: -Some applications (such as the formidable B) work around this -by carrying their own replacement functions for character set handling -with them, and either implementing OS-dependent hacks or doing multiple -conversions (which is slow and unreliable in case the OS implements -encodings slightly different than the terminal emulator). + # use Backspace = ^H + $ stty erase ^H + $ echo -n "^[[36h" -The rxvt-unicode author insists that the right way to fix this is in the -system libraries once and for all, instead of forcing every app to carry -complete replacements for them :) + # use Backspace = ^? + $ stty erase ^? + $ echo -n "^[[36l" -=head2 I use Solaris 9 and it doesn't compile/work/etc. +This helps satisfy some of the Backspace discrepancies that occur, but +if you use Backspace = C<^H>, make sure that the termcap/terminfo value +properly reflects that. -Try the diff in F as a base. It fixes the worst -problems with C and a compile problem. +The Delete key is a another casualty of the ill-defined Backspace problem. +To avoid confusion between the Backspace and Delete keys, the Delete +key has been assigned an escape sequence to match the vt100 for Execute +(C) and is in the supplied termcap/terminfo. -=head2 How can I use rxvt-unicode under cygwin? +Some other Backspace problems: -rxvt-unicode should compile and run out of the box on cygwin, using -the X11 libraries that come with cygwin. libW11 emulation is no -longer supported (and makes no sense, either, as it only supported a -single font). I recommend starting the X-server in C<-multiwindow> or -C<-rootless> mode instead, which will result in similar look&feel as the -old libW11 emulation. +some editors use termcap/terminfo, +some editors (vim I'm told) expect Backspace = ^H, +GNU Emacs (and Emacs-like editors) use ^H for help. -At the time of this writing, cygwin didn't seem to support any multi-byte -encodings (you might try C), so you are likely limited -to 8-bit encodings. +Perhaps someday this will all be resolved in a consistent manner. -=head2 How does rxvt-unicode determine the encoding to use? +=head3 I don't like the key-bindings. How do I change them? -See next entry. +There are some compile-time selections available via configure. Unless +you have run "configure" with the C<--disable-resources> option you can +use the `keysym' resource to alter the keystrings associated with keysyms. -=head2 Is there an option to switch encodings? +Here's an example for a URxvt session started using C<@@RXVT_NAME@@ -name URxvt> -Unlike some other terminals, rxvt-unicode has no encoding switch, and no -specific "utf-8" mode, such as xterm. In fact, it doesn't even know about -UTF-8 or any other encodings with respect to terminal I/O. + URxvt.keysym.Home: \033[1~ + URxvt.keysym.End: \033[4~ + URxvt.keysym.C-apostrophe: \033 + URxvt.keysym.C-slash: \033 + URxvt.keysym.C-semicolon: \033 + URxvt.keysym.C-grave: \033 + URxvt.keysym.C-comma: \033 + URxvt.keysym.C-period: \033 + URxvt.keysym.C-0x60: \033 + URxvt.keysym.C-Tab: \033 + URxvt.keysym.C-Return: \033 + URxvt.keysym.S-Return: \033 + URxvt.keysym.S-space: \033 + URxvt.keysym.M-Up: \033 + URxvt.keysym.M-Down: \033 + URxvt.keysym.M-Left: \033 + URxvt.keysym.M-Right: \033 + URxvt.keysym.M-C-0: list \033 + URxvt.keysym.M-C-a: list \033 + URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 -The reasons is that there exists a perfectly fine mechanism for selecting -the encoding, doing I/O and (most important) communicating this to all -applications so everybody agrees on character properties such as width -and code number. This mechanism is the I. Applications not using -that info will have problems (for example, C gets the width of -characters wrong as it uses it's own, locale-independent table under all -locales). +See some more examples in the documentation for the B resource. -Rxvt-unicode uses the C locale category to select encoding. All -programs doing the same (that is, most) will automatically agree in the -interpretation of characters. +=head3 I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map -Unfortunately, there is no system-independent way to select locales, nor -is there a standard on how locale specifiers will look like. + KP_Insert == Insert + F22 == Print + F27 == Home + F29 == Prior + F33 == End + F35 == Next -On most systems, the content of the C environment variable -contains an arbitrary string which corresponds to an already-installed -locale. Common names for locales are C, C, -C, i.e. C, but other forms -(i.e. C or C) are also common. +Rather than have rxvt-unicode try to accommodate all the various possible +keyboard mappings, it is better to use `xmodmap' to remap the keys as +required for your particular machine. -Rxvt-unicode ignores all other locale categories, and except for -the encoding, ignores country or language-specific settings, -i.e. C and C are the normally same to -rxvt-unicode. -If you want to use a specific encoding you have to make sure you start -rxvt-unicode with the correct C category. -=head2 Can I switch locales at runtime? +=head2 Terminal Configuration -Yes, using an escape sequence. Try something like this, which sets -rxvt-unicode's idea of C. +=head3 Why doesn't rxvt-unicode read my resources? - printf '\e]701;%s\007' ja_JP.SJIS +Well, why, indeed? It does, in a way very similar to other X +applications. Most importantly, this means that if you or your OS loads +resources into the X display (the right way to do it), rxvt-unicode will +ignore any resource files in your home directory. It will only read +F<$HOME/.Xdefaults> when no resources are attached to the display. -See also the previous answer. +If you have or use an F<$HOME/.Xresources> file, chances are that +resources are loaded into your X-server. In this case, you have to +re-login after every change (or run F). -Sometimes this capability is rather handy when you want to work in -one locale (e.g. C) but some programs don't support it -(e.g. UTF-8). For example, I use this script to start C, which -first switches to a locale supported by xjdic and back later: +Also consider the form resources have to use: - printf '\e]701;%s\007' ja_JP.SJIS - xjdic -js - printf '\e]701;%s\007' de_DE.UTF-8 + URxvt.resource: value -You can also use xterm's C program, which usually works fine, except -for some locales where character width differs between program- and -rxvt-unicode-locales. +If you want to use another form (there are lots of different ways of +specifying resources), make sure you understand wether and why it +works. If unsure, use the form above. -=head2 Can I switch the fonts at runtime? +=head3 When I log-in to another system it tells me about missing terminfo data? -Yes, using an escape sequence. Try something like this, which has the same -effect as using the C<-fn> switch, and takes effect immediately: +The terminal description used by rxvt-unicode is not as widely available +as that for xterm, or even rxvt (for which the same problem often arises). - printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" +The correct solution for this problem is to install the terminfo, this can +be done like this (with ncurses' infocmp): -This is useful if you e.g. work primarily with japanese (and prefer a -japanese font), but you have to switch to chinese temporarily, where -japanese fonts would only be in your way. + REMOTE=remotesystem.domain + infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" -You can think of this as a kind of manual ISO-2022 switching. +... or by installing rxvt-unicode normally on the remote system, -=head2 Why do italic characters look as if clipped? +If you cannot or do not want to do this, then you can simply set +C or even C, and live with the small number of +problems arising, which includes wrong keymapping, less and different +colours and some refresh errors in fullscreen applications. It's a nice +quick-and-dirty workaround for rare cases, though. -Many fonts have difficulties with italic characters and hinting. For -example, the otherwise very nicely hinted font C completely fails in it's italic face. A workaround might be to -enable freetype autohinting, i.e. like this: +If you always want to do this (and are fine with the consequences) you +can either recompile rxvt-unicode with the desired TERM value or use a +resource to set it: - URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true - URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true + URxvt.termName: rxvt -=head2 My input method wants but I want UTF-8, what can I do? +If you don't plan to use B (quite common...) you could also replace +the rxvt terminfo file with the rxvt-unicode one. -You can specify separate locales for the input method and the rest of the -terminal, using the resource C: +=head3 C outputs some error when compiling the terminfo entry. - URxvt.imlocale: ja_JP.EUC-JP +Most likely it's the empty definition for C. Just replace it by +C and try again. -Now you can start your terminal with C and still -use your input method. Please note, however, that you will not be able to -input characters outside C in a normal way then, as your input -method limits you. +=head3 C's readline does not work correctly under @@RXVT_NAME@@. -=head2 Rxvt-unicode crashes when the X Input Method changes or exits. +See next entry. -Unfortunately, this is unavoidable, as the XIM protocol is racy by -design. Applications can avoid some crashes at the expense of memory -leaks, and Input Methods can avoid some crashes by careful ordering at -exit time. B (and derived input methods) generally succeeds, -while B (or similar input methods) fails. In the end, however, -crashes cannot be completely avoided even if both sides cooperate. +=head3 I need a termcap file entry. -So the only workaround is not to kill your Input Method Servers. +One reason you might want this is that some distributions or operating +systems still compile some programs using the long-obsoleted termcap +library (Fedora Core's bash is one example) and rely on a termcap entry +for C. -=head2 Rxvt-unicode uses gobs of memory, how can I reduce that? +You could use rxvt's termcap entry with resonable results in many cases. +You can also create a termcap entry by using terminfo's infocmp program +like this: -Rxvt-unicode tries to obey the rule of not charging you for something you -don't use. One thing you should try is to configure out all settings that -you don't need, for example, Xft support is a resource hog by design, -when used. Compiling it out ensures that no Xft font will be loaded -accidentally when rxvt-unicode tries to find a font for your characters. + infocmp -C rxvt-unicode -Also, many people (me included) like large windows and even larger -scrollback buffers: Without C<--enable-unicode3>, rxvt-unicode will use -6 bytes per screen cell. For a 160x?? window this amounts to almost a -kilobyte per line. A scrollback buffer of 10000 lines will then (if full) -use 10 Megabytes of memory. With C<--enable-unicode3> it gets worse, as -rxvt-unicode then uses 8 bytes per screen cell. +Or you could use this termcap entry, generated by the command above: -=head2 Can I speed up Xft rendering somehow? + rxvt-unicode|rxvt-unicode terminal (X Window System):\ + :am:bw:eo:km:mi:ms:xn:xo:\ + :co#80:it#8:li#24:lm#0:\ + :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ + :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ + :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ + :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ + :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ + :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ + :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ + :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ + :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ + :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ + :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ + :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ + :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ + :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ + :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ + :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ + :vs=\E[?25h: -Yes, the most obvious way to speed it up is to avoid Xft entirely, as -it is simply slow. If you still want Xft fonts you might try to disable -antialiasing (by appending C<:antialias=false>), which saves lots of -memory and also speeds up rendering considerably. +=head3 Why does C no longer have coloured output? -=head2 Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? +The C in the GNU coreutils unfortunately doesn't use terminfo to +decide wether a terminal has colour, but uses it's own configuration +file. Needless to say, C is not in it's default file (among +with most other terminals supporting colour). Either add: -Rxvt-unicode will use whatever you specify as a font. If it needs to -fall back to it's default font search list it will prefer X11 core -fonts, because they are small and fast, and then use Xft fonts. It has -antialiasing disabled for most of them, because the author thinks they -look best that way. + TERM rxvt-unicode -If you want antialiasing, you have to specify the fonts manually. +to C or simply add: -=head2 Mouse cut/paste suddenly no longer works. + alias ls='ls --color=auto' -Make sure that mouse reporting is actually turned off since killing -some editors prematurely may leave the mouse in mouse report mode. I've -heard that tcsh may use mouse reporting unless it otherwise specified. A -quick check is to see if cut/paste works when the Alt or Shift keys are -depressed. +to your C<.profile> or C<.bashrc>. -=head2 What's with this bold/blink stuff? +=head3 Why doesn't vim/emacs etc. use the 88 colour mode? -If no bold colour is set via C, bold will invert text using the -standard foreground colour. +See next entry. -For the standard background colour, blinking will actually make the -text blink when compiled with C<--enable-blinking>. with standard -colours. Without C<--enable-blinking>, the blink attribute will be -ignored. +=head3 Why doesn't vim/emacs etc. make use of italic? -On ANSI colours, bold/blink attributes are used to set high-intensity -foreground/background colors. +See next entry. -color0-7 are the low-intensity colors. +=head3 Why are the secondary screen-related options not working properly? -color8-15 are the corresponding high-intensity colors. +Make sure you are using C. Some pre-packaged +distributions (most notably Debian GNU/Linux) break rxvt-unicode +by setting C to C, which doesn't have these extra +features. Unfortunately, some of these (most notably, again, Debian +GNU/Linux) furthermore fail to even install the C terminfo +file, so you will need to install it on your own (See the question B on +how to do this). -=head2 I don't like the screen colors. How do I change them? -You can change the screen colors at run-time using F<~/.Xdefaults> -resources (or as long-options). +=head2 Encoding / Locale / Input Method Issues -Here are values that are supposed to resemble a VGA screen, -including the murky brown that passes for low-intensity yellow: +=head3 Rxvt-unicode does not seem to understand the selected encoding? - URxvt.color0: #000000 - URxvt.color1: #A80000 - URxvt.color2: #00A800 - URxvt.color3: #A8A800 - URxvt.color4: #0000A8 - URxvt.color5: #A800A8 - URxvt.color6: #00A8A8 - URxvt.color7: #A8A8A8 +See next entry. - URxvt.color8: #000054 - URxvt.color9: #FF0054 - URxvt.color10: #00FF54 - URxvt.color11: #FFFF54 - URxvt.color12: #0000FF - URxvt.color13: #FF00FF - URxvt.color14: #00FFFF - URxvt.color15: #FFFFFF +=head3 Unicode does not seem to work? -And here is a more complete set of non-standard colors described (not by -me) as "pretty girly". +If you encounter strange problems like typing an accented character but +getting two unrelated other characters or similar, or if program output is +subtly garbled, then you should check your locale settings. - URxvt.cursorColor: #dc74d1 - URxvt.pointerColor: #dc74d1 - URxvt.background: #0e0e0e - URxvt.foreground: #4ad5e1 - URxvt.color0: #000000 - URxvt.color8: #8b8f93 - URxvt.color1: #dc74d1 - URxvt.color9: #dc74d1 - URxvt.color2: #0eb8c7 - URxvt.color10: #0eb8c7 - URxvt.color3: #dfe37e - URxvt.color11: #dfe37e - URxvt.color5: #9e88f0 - URxvt.color13: #9e88f0 - URxvt.color6: #73f7ff - URxvt.color14: #73f7ff - URxvt.color7: #e1dddd - URxvt.color15: #e1dddd +Rxvt-unicode must be started with the same C setting as the +programs. Often rxvt-unicode is started in the C locale, while the +login script running within the rxvt-unicode window changes the locale to +something else, e.g. C. Needless to say, this is not going to work. -=head2 How can I start @@RXVT_NAME@@d in a race-free way? +The best thing is to fix your startup environment, as you will likely run +into other problems. If nothing works you can try this in your .profile. -Try C<@@RXVT_NAME@@d -f -o>, which tells @@RXVT_NAME@@d to open the -display, create the listening socket and then fork. + printf '\e]701;%s\007' "$LC_CTYPE" -=head2 What's with the strange Backspace/Delete key behaviour? +If this doesn't work, then maybe you use a C specification not +supported on your systems. Some systems have a C command which +displays this (also, C can be used to check locale settings, as +it will complain loudly if it cannot set the locale). If it displays something +like: -Assuming that the physical Backspace key corresponds to the -BackSpace keysym (not likely for Linux ... see the following -question) there are two standard values that can be used for -Backspace: C<^H> and C<^?>. + locale: Cannot set LC_CTYPE to default locale: ... -Historically, either value is correct, but rxvt-unicode adopts the debian -policy of using C<^?> when unsure, because it's the one only only correct -choice :). +Then the locale you specified is not supported on your system. -Rxvt-unicode tries to inherit the current stty settings and uses the value -of `erase' to guess the value for backspace. If rxvt-unicode wasn't -started from a terminal (say, from a menu or by remote shell), then the -system value of `erase', which corresponds to CERASE in , will -be used (which may not be the same as your stty setting). +If nothing works and you are sure that everything is set correctly then +you will need to remember a little known fact: Some programs just don't +support locales :( -For starting a new rxvt-unicode: +=head3 How does rxvt-unicode determine the encoding to use? - # use Backspace = ^H - $ stty erase ^H - $ @@RXVT_NAME@@ +See next entry. - # use Backspace = ^? - $ stty erase ^? - $ @@RXVT_NAME@@ +=head3 Is there an option to switch encodings? -Toggle with C / C. +Unlike some other terminals, rxvt-unicode has no encoding switch, and no +specific "utf-8" mode, such as xterm. In fact, it doesn't even know about +UTF-8 or any other encodings with respect to terminal I/O. -For an existing rxvt-unicode: +The reasons is that there exists a perfectly fine mechanism for selecting +the encoding, doing I/O and (most important) communicating this to all +applications so everybody agrees on character properties such as width +and code number. This mechanism is the I. Applications not using +that info will have problems (for example, C gets the width of +characters wrong as it uses it's own, locale-independent table under all +locales). - # use Backspace = ^H - $ stty erase ^H - $ echo -n "^[[36h" +Rxvt-unicode uses the C locale category to select encoding. All +programs doing the same (that is, most) will automatically agree in the +interpretation of characters. - # use Backspace = ^? - $ stty erase ^? - $ echo -n "^[[36l" +Unfortunately, there is no system-independent way to select locales, nor +is there a standard on how locale specifiers will look like. -This helps satisfy some of the Backspace discrepancies that occur, but -if you use Backspace = C<^H>, make sure that the termcap/terminfo value -properly reflects that. +On most systems, the content of the C environment variable +contains an arbitrary string which corresponds to an already-installed +locale. Common names for locales are C, C, +C, i.e. C, but other forms +(i.e. C or C) are also common. -The Delete key is a another casualty of the ill-defined Backspace problem. -To avoid confusion between the Backspace and Delete keys, the Delete -key has been assigned an escape sequence to match the vt100 for Execute -(C) and is in the supplied termcap/terminfo. +Rxvt-unicode ignores all other locale categories, and except for +the encoding, ignores country or language-specific settings, +i.e. C and C are the normally same to +rxvt-unicode. -Some other Backspace problems: +If you want to use a specific encoding you have to make sure you start +rxvt-unicode with the correct C category. -some editors use termcap/terminfo, -some editors (vim I'm told) expect Backspace = ^H, -GNU Emacs (and Emacs-like editors) use ^H for help. +=head3 Can I switch locales at runtime? -Perhaps someday this will all be resolved in a consistent manner. +Yes, using an escape sequence. Try something like this, which sets +rxvt-unicode's idea of C. -=head2 I don't like the key-bindings. How do I change them? + printf '\e]701;%s\007' ja_JP.SJIS -There are some compile-time selections available via configure. Unless -you have run "configure" with the C<--disable-resources> option you can -use the `keysym' resource to alter the keystrings associated with keysyms. +See also the previous answer. -Here's an example for a URxvt session started using C<@@RXVT_NAME@@ -name URxvt> +Sometimes this capability is rather handy when you want to work in +one locale (e.g. C) but some programs don't support it +(e.g. UTF-8). For example, I use this script to start C, which +first switches to a locale supported by xjdic and back later: - URxvt.keysym.Home: \033[1~ - URxvt.keysym.End: \033[4~ - URxvt.keysym.C-apostrophe: \033 - URxvt.keysym.C-slash: \033 - URxvt.keysym.C-semicolon: \033 - URxvt.keysym.C-grave: \033 - URxvt.keysym.C-comma: \033 - URxvt.keysym.C-period: \033 - URxvt.keysym.C-0x60: \033 - URxvt.keysym.C-Tab: \033 - URxvt.keysym.C-Return: \033 - URxvt.keysym.S-Return: \033 - URxvt.keysym.S-space: \033 - URxvt.keysym.M-Up: \033 - URxvt.keysym.M-Down: \033 - URxvt.keysym.M-Left: \033 - URxvt.keysym.M-Right: \033 - URxvt.keysym.M-C-0: list \033 - URxvt.keysym.M-C-a: list \033 - URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 + printf '\e]701;%s\007' ja_JP.SJIS + xjdic -js + printf '\e]701;%s\007' de_DE.UTF-8 -See some more examples in the documentation for the B resource. +You can also use xterm's C program, which usually works fine, except +for some locales where character width differs between program- and +rxvt-unicode-locales. -=head2 I'm using keyboard model XXX that has extra Prior/Next/Insert keys. -How do I make use of them? For example, the Sun Keyboard type 4 -has the following mappings that rxvt-unicode doesn't recognize. +=head3 My input method wants but I want UTF-8, what can I do? - KP_Insert == Insert - F22 == Print - F27 == Home - F29 == Prior - F33 == End - F35 == Next +You can specify separate locales for the input method and the rest of the +terminal, using the resource C: -Rather than have rxvt-unicode try to accommodate all the various possible -keyboard mappings, it is better to use `xmodmap' to remap the keys as -required for your particular machine. + URxvt.imlocale: ja_JP.EUC-JP -=head2 How do I distinguish wether I'm running rxvt-unicode or a regular xterm? -I need this to decide about setting colors etc. +Now you can start your terminal with C and still +use your input method. Please note, however, that you will not be able to +input characters outside C in a normal way then, as your input +method limits you. -rxvt and rxvt-unicode always export the variable "COLORTERM", so you can -check and see if that is set. Note that several programs, JED, slrn, -Midnight Commander automatically check this variable to decide whether or -not to use color. +=head3 Rxvt-unicode crashes when the X Input Method changes or exits. -=head2 How do I set the correct, full IP address for the DISPLAY variable? +Unfortunately, this is unavoidable, as the XIM protocol is racy by +design. Applications can avoid some crashes at the expense of memory +leaks, and Input Methods can avoid some crashes by careful ordering at +exit time. B (and derived input methods) generally succeeds, +while B (or similar input methods) fails. In the end, however, +crashes cannot be completely avoided even if both sides cooperate. -If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled -insecure mode then it is possible to use the following shell script -snippets to correctly set the display. If your version of rxvt-unicode -wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) then -the COLORTERM variable can be used to distinguish rxvt-unicode from a -regular xterm. +So the only workaround is not to kill your Input Method Servers. -Courtesy of Chuck Blake with the following shell script -snippets: - # Bourne/Korn/POSIX family of shells: - [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know - if [ ${TERM:-foo} = xterm ]; then - stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not - echo -n '^[Z' - read term_id - stty icanon echo - if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then - echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string - read DISPLAY # set it in our local shell - fi - fi +=head2 Operating Systems / Package Maintaining -=head2 How do I compile the manual pages for myself? +=head3 I am using Debian GNU/Linux and have a problem... -You need to have a recent version of perl installed as F, -one that comes with F, F and F. Then go to -the doc subdirectory and enter C. +The Debian GNU/Linux package of rxvt-unicode in sarge contains large +patches that considerably change the behaviour of rxvt-unicode (but +unfortunately this notice has been removed). Before reporting a bug to +the original rxvt-unicode author please download and install the genuine +version (L) and try to reproduce +the problem. If you cannot, chances are that the problems are specific to +Debian GNU/Linux, in which case it should be reported via the Debian Bug +Tracking System (use C to report the bug). -=head2 My question isn't answered here, can I ask a human? +For other problems that also affect the Debian package, you can and +probably should use the Debian BTS, too, because, after all, it's also a +bug in the Debian version and it serves as a reminder for other users that +might encounter the same issue. -Before sending me mail, you could go to IRC: C, -channel C<#rxvt-unicode> has some rxvt-unicode enthusiasts that might be -interested in learning about new and exciting problems (but not FAQs :). +=head3 I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation? + +You should build one binary with the default options. F +now enables most useful options, and the trend goes to making them +runtime-switchable, too, so there is usually no drawback to enbaling them, +except higher disk and possibly memory usage. The perl interpreter should +be enabled, as important functionality (menus, selection, likely more in +the future) depends on it. + +You should not overwrite the C snd C resources +system-wide (except maybe with C). This will result in useful +behaviour. If your distribution aims at low memory, add an empty +C resource to the app-defaults file. This will keep the +perl interpreter disabled until the user enables it. + +If you can/want build more binaries, I recommend building a minimal +one with C<--disable-everything> (very useful) and a maximal one with +C<--enable-everything> (less useful, it will be very big due to a lot of +encodings built-in that increase download times and are rarely used). + +=head3 I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe? + +It should be, starting with release 7.1. You are encouraged to properly +install urxvt with privileges necessary for your OS now. + +When rxvt-unicode detects that it runs setuid or setgid, it will fork +into a helper process for privileged operations (pty handling on some +systems, utmp/wtmp/lastlog handling on others) and drop privileges +immediately. This is much safer than most other terminals that keep +privileges while running (but is more relevant to urxvt, as it contains +things as perl interpreters, which might be "helpful" to attackers). + +This forking is done as the very first within main(), which is very early +and reduces possible bugs to initialisation code run before main(), or +things like the dynamic loader of your system, which should result in very +little risk. + +=head3 On Solaris 9, many line-drawing characters are too wide. + +Seems to be a known bug, read +L. Some people use the +following ugly workaround to get non-double-wide-characters working: + + #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) + +=head3 I am on FreeBSD and rxvt-unicode does not seem to work at all. + +Rxvt-unicode requires the symbol C<__STDC_ISO_10646__> to be defined +in your compile environment, or an implementation that implements it, +wether it defines the symbol or not. C<__STDC_ISO_10646__> requires that +B is represented as unicode. + +As you might have guessed, FreeBSD does neither define this symobl nor +does it support it. Instead, it uses it's own internal representation of +B. This is, of course, completely fine with respect to standards. + +However, that means rxvt-unicode only works in C, C and +C locales under FreeBSD (which all use Unicode as B. + +C<__STDC_ISO_10646__> is the only sane way to support multi-language +apps in an OS, as using a locale-dependent (and non-standardized) +representation of B makes it impossible to convert between +B (as used by X11 and your applications) and any other encoding +without implementing OS-specific-wrappers for each and every locale. There +simply are no APIs to convert B into anything except the current +locale encoding. + +Some applications (such as the formidable B) work around this +by carrying their own replacement functions for character set handling +with them, and either implementing OS-dependent hacks or doing multiple +conversions (which is slow and unreliable in case the OS implements +encodings slightly different than the terminal emulator). + +The rxvt-unicode author insists that the right way to fix this is in the +system libraries once and for all, instead of forcing every app to carry +complete replacements for them :) + +=head3 I use Solaris 9 and it doesn't compile/work/etc. + +Try the diff in F as a base. It fixes the worst +problems with C and a compile problem. + +=head3 How can I use rxvt-unicode under cygwin? + +rxvt-unicode should compile and run out of the box on cygwin, using +the X11 libraries that come with cygwin. libW11 emulation is no +longer supported (and makes no sense, either, as it only supported a +single font). I recommend starting the X-server in C<-multiwindow> or +C<-rootless> mode instead, which will result in similar look&feel as the +old libW11 emulation. + +At the time of this writing, cygwin didn't seem to support any multi-byte +encodings (you might try C), so you are likely limited +to 8-bit encodings. =head1 RXVT TECHNICAL REFERENCE diff --git a/doc/rxvt.7.txt b/doc/rxvt.7.txt index 876bc9ed..c3193704 100644 --- a/doc/rxvt.7.txt +++ b/doc/rxvt.7.txt @@ -20,10 +20,405 @@ DESCRIPTION . FREQUENTLY ASKED QUESTIONS - The new selection selects pieces that are too big, how can I select -single words? - Yes. For example, if you want to select alphanumeric words, you can use - the following resource: + Meta, Features & Commandline Issues + My question isn't answered here, can I ask a human? + Before sending me mail, you could go to IRC: "irc.freenode.net", channel + "#rxvt-unicode" has some rxvt-unicode enthusiasts that might be + interested in learning about new and exciting problems (but not FAQs :). + + Does it support tabs, can I have a tabbed rxvt-unicode? + Beginning with version 7.3, there is a perl extension that implements a + simple tabbed terminal. It is installed by default, so any of these + should give you tabs: + + rxvt -pe tabbed + + URxvt.perl-ext-common: default,tabbed + + It will also work fine with tabbing functionality of many window + managers or similar tabbing programs, and its embedding-features allow + it to be embedded into other programs, as witnessed by doc/rxvt-tabbed + or the upcoming "Gtk2::URxvt" perl module, which features a tabbed urxvt + (murxvt) terminal as an example embedding application. + + How do I know which rxvt-unicode version I'm using? + The version number is displayed with the usage (-h). Also the escape + sequence "ESC [ 8 n" sets the window title to the version number. When + using the rxvtc client, the version displayed is that of the daemon. + + Rxvt-unicode uses gobs of memory, how can I reduce that? + Rxvt-unicode tries to obey the rule of not charging you for something + you don't use. One thing you should try is to configure out all settings + that you don't need, for example, Xft support is a resource hog by + design, when used. Compiling it out ensures that no Xft font will be + loaded accidentally when rxvt-unicode tries to find a font for your + characters. + + Also, many people (me included) like large windows and even larger + scrollback buffers: Without "--enable-unicode3", rxvt-unicode will use 6 + bytes per screen cell. For a 160x?? window this amounts to almost a + kilobyte per line. A scrollback buffer of 10000 lines will then (if + full) use 10 Megabytes of memory. With "--enable-unicode3" it gets + worse, as rxvt-unicode then uses 8 bytes per screen cell. + + How can I start rxvtd in a race-free way? + Try "rxvtd -f -o", which tells rxvtd to open the display, create the + listening socket and then fork. + + How do I distinguish wether I'm running rxvt-unicode or a regular xterm? I need this to decide about setting colors etc. + rxvt and rxvt-unicode always export the variable "COLORTERM", so you can + check and see if that is set. Note that several programs, JED, slrn, + Midnight Commander automatically check this variable to decide whether + or not to use color. + + How do I set the correct, full IP address for the DISPLAY variable? + If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled + insecure mode then it is possible to use the following shell script + snippets to correctly set the display. If your version of rxvt-unicode + wasn't also compiled with ESCZ_ANSWER (as assumed in these snippets) + then the COLORTERM variable can be used to distinguish rxvt-unicode from + a regular xterm. + + Courtesy of Chuck Blake with the following shell script + snippets: + + # Bourne/Korn/POSIX family of shells: + [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know + if [ ${TERM:-foo} = xterm ]; then + stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not + echo -n '^[Z' + read term_id + stty icanon echo + if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then + echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string + read DISPLAY # set it in our local shell + fi + fi + + How do I compile the manual pages on my own? + You need to have a recent version of perl installed as /usr/bin/perl, + one that comes with pod2man, pod2text and pod2html. Then go to the doc + subdirectory and enter "make alldoc". + + Isn't rxvt-unicode supposed to be small? Don't all those features bloat? + I often get asked about this, and I think, no, they didn't cause extra + bloat. If you compare a minimal rxvt and a minimal urxvt, you can see + that the urxvt binary is larger (due to some encoding tables always + being compiled in), but it actually uses less memory (RSS) after + startup. Even with "--disable-everything", this comparison is a bit + unfair, as many features unique to urxvt (locale, encoding conversion, + iso14755 etc.) are already in use in this mode. + + text data bss drs rss filename + 98398 1664 24 15695 1824 rxvt --disable-everything + 188985 9048 66616 18222 1788 urxvt --disable-everything + + When you "--enable-everything" (which _is_ unfair, as this involves xft + and full locale/XIM support which are quite bloaty inside libX11 and my + libc), the two diverge, but not unreasnobaly so. + + text data bss drs rss filename + 163431 2152 24 20123 2060 rxvt --enable-everything + 1035683 49680 66648 29096 3680 urxvt --enable-everything + + The very large size of the text section is explained by the east-asian + encoding tables, which, if unused, take up disk space but nothing else + and can be compiled out unless you rely on X11 core fonts that use those + encodings. The BSS size comes from the 64k emergency buffer that my c++ + compiler allocates (but of course doesn't use unless you are out of + memory). Also, using an xft font instead of a core font immediately adds + a few megabytes of RSS. Xft indeed is responsible for a lot of RSS even + when not used. + + Of course, due to every character using two or four bytes instead of + one, a large scrollback buffer will ultimately make rxvt-unicode use + more memory. + + Compared to e.g. Eterm (5112k), aterm (3132k) and xterm (4680k), this + still fares rather well. And compared to some monsters like + gnome-terminal (21152k + extra 4204k in separate processes) or konsole + (22200k + extra 43180k in daemons that stay around after exit, plus half + a minute of startup time, including the hundreds of warnings it spits + out), it fares extremely well *g*. + + Why C++, isn't that unportable/bloated/uncool? + Is this a question? :) It comes up very often. The simple answer is: I + had to write it, and C++ allowed me to write and maintain it in a + fraction of the time and effort (which is a scarce resource for me). Put + even shorter: It simply wouldn't exist without C++. + + My personal stance on this is that C++ is less portable than C, but in + the case of rxvt-unicode this hardly matters, as its portability limits + are defined by things like X11, pseudo terminals, locale support and + unix domain sockets, which are all less portable than C++ itself. + + Regarding the bloat, see the above question: It's easy to write programs + in C that use gobs of memory, an certainly possible to write programs in + C++ that don't. C++ also often comes with large libraries, but this is + not necessarily the case with GCC. Here is what rxvt links against on my + system with a minimal config: + + libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) + libc.so.6 => /lib/libc.so.6 (0x00002aaaaadde000) + libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab01d000) + /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) + + And here is rxvt-unicode: + + libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) + libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x00002aaaaada2000) + libc.so.6 => /lib/libc.so.6 (0x00002aaaaaeb0000) + libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab0ee000) + /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) + + No large bloated libraries (of course, none were linked in statically), + except maybe libX11 :) + + Rendering, Font & Look and Feel Issues + I can't get transparency working, what am I doing wrong? + First of all, transparency isn't officially supported in rxvt-unicode, + so you are mostly on your own. Do not bug the author about it (but you + may bug everybody else). Also, if you can't get it working consider it a + rite of passage: ... and you failed. + + Here are four ways to get transparency. Do read the manpage and option + descriptions for the programs mentioned and rxvt-unicode. Really, do it! + + 1. Use inheritPixmap: + + Esetroot wallpaper.jpg + rxvt -ip -tint red -sh 40 + + That works. If you think it doesn't, you lack transparency and tinting + support, or you are unable to read. + + 2. Use a simple pixmap and emulate pseudo-transparency. This enables you + to use effects other than tinting and shading: Just shade/tint/whatever + your picture with gimp or any other tool: + + convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm + rxvt -pixmap background.xpm -pe automove-background + + That works. If you think it doesn't, you lack XPM and Perl support, or + you are unable to read. + + 3. Use an ARGB visual: + + rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc + + This requires XFT support, and the support of your X-server. If that + doesn't work for you, blame Xorg and Keith Packard. ARGB visuals aren't + there yet, no matter what they claim. Rxvt-Unicode contains the + neccessary bugfixes and workarounds for Xft and Xlib to make it work, + but that doesn't mean that your WM has the required kludges in place. + + 4. Use xcompmgr and let it do the job: + + xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ + -set _NET_WM_WINDOW_OPACITY 0xc0000000 + + Then click on a window you want to make transparent. Replace 0xc0000000 + by other values to change the degree of opacity. If it doesn't work and + your server crashes, you got to keep the pieces. + + Why do some chinese characters look so different than others? + This is because there is a difference between script and language -- + rxvt-unicode does not know which language the text that is output is, as + it only knows the unicode character codes. If rxvt-unicode first sees a + japanese/chinese character, it might choose a japanese font for display. + Subsequent japanese characters will use that font. Now, many chinese + characters aren't represented in japanese fonts, so when the first + non-japanese character comes up, rxvt-unicode will look for a chinese + font -- unfortunately at this point, it will still use the japanese font + for chinese characters that are also in the japanese font. + + The workaround is easy: just tag a chinese font at the end of your font + list (see the previous question). The key is to view the font list as a + preference list: If you expect more japanese, list a japanese font + first. If you expect more chinese, put a chinese font first. + + In the future it might be possible to switch language preferences at + runtime (the internal data structure has no problem with using different + fonts for the same character at the same time, but no interface for this + has been designed yet). + + Until then, you might get away with switching fonts at runtime (see "Can + I switch the fonts at runtime?" later in this document). + + Why does rxvt-unicode sometimes leave pixel droppings? + Most fonts were not designed for terminal use, which means that + character size varies a lot. A font that is otherwise fine for terminal + use might contain some characters that are simply too wide. Rxvt-unicode + will avoid these characters. For characters that are just "a bit" too + wide a special "careful" rendering mode is used that redraws adjacent + characters. + + All of this requires that fonts do not lie about character sizes, + however: Xft fonts often draw glyphs larger than their acclaimed + bounding box, and rxvt-unicode has no way of detecting this (the correct + way is to ask for the character bounding box, which unfortunately is + wrong in these cases). + + It's not clear (to me at least), wether this is a bug in Xft, freetype, + or the respective font. If you encounter this problem you might try + using the "-lsp" option to give the font more height. If that doesn't + work, you might be forced to use a different font. + + All of this is not a problem when using X11 core fonts, as their + bounding box data is correct. + + How can I keep rxvt-unicode from using reverse video so much? + First of all, make sure you are running with the right terminal settings + ("TERM=rxvt-unicode"), which will get rid of most of these effects. Then + make sure you have specified colours for italic and bold, as otherwise + rxvt-unicode might use reverse video to simulate the effect: + + URxvt.colorBD: white + URxvt.colorIT: green + + Some programs assume totally weird colours (red instead of blue), how can I fix that? + For some unexplainable reason, some rare programs assume a very weird + colour palette when confronted with a terminal with more than the + standard 8 colours (rxvt-unicode supports 88). The right fix is, of + course, to fix these programs not to assume non-ISO colours without very + good reasons. + + In the meantime, you can either edit your "rxvt-unicode" terminfo + definition to only claim 8 colour support or use "TERM=rxvt", which will + fix colours but keep you from using other rxvt-unicode features. + + Can I switch the fonts at runtime? + Yes, using an escape sequence. Try something like this, which has the + same effect as using the "-fn" switch, and takes effect immediately: + + printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" + + This is useful if you e.g. work primarily with japanese (and prefer a + japanese font), but you have to switch to chinese temporarily, where + japanese fonts would only be in your way. + + You can think of this as a kind of manual ISO-2022 switching. + + Why do italic characters look as if clipped? + Many fonts have difficulties with italic characters and hinting. For + example, the otherwise very nicely hinted font "xft:Bitstream Vera Sans + Mono" completely fails in it's italic face. A workaround might be to + enable freetype autohinting, i.e. like this: + + URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true + URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true + + Can I speed up Xft rendering somehow? + Yes, the most obvious way to speed it up is to avoid Xft entirely, as it + is simply slow. If you still want Xft fonts you might try to disable + antialiasing (by appending ":antialias=false"), which saves lots of + memory and also speeds up rendering considerably. + + Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? + Rxvt-unicode will use whatever you specify as a font. If it needs to + fall back to it's default font search list it will prefer X11 core + fonts, because they are small and fast, and then use Xft fonts. It has + antialiasing disabled for most of them, because the author thinks they + look best that way. + + If you want antialiasing, you have to specify the fonts manually. + + What's with this bold/blink stuff? + If no bold colour is set via "colorBD:", bold will invert text using the + standard foreground colour. + + For the standard background colour, blinking will actually make the text + blink when compiled with "--enable-blinking". with standard colours. + Without "--enable-blinking", the blink attribute will be ignored. + + On ANSI colours, bold/blink attributes are used to set high-intensity + foreground/background colors. + + color0-7 are the low-intensity colors. + + color8-15 are the corresponding high-intensity colors. + + I don't like the screen colors. How do I change them? + You can change the screen colors at run-time using ~/.Xdefaults + resources (or as long-options). + + Here are values that are supposed to resemble a VGA screen, including + the murky brown that passes for low-intensity yellow: + + URxvt.color0: #000000 + URxvt.color1: #A80000 + URxvt.color2: #00A800 + URxvt.color3: #A8A800 + URxvt.color4: #0000A8 + URxvt.color5: #A800A8 + URxvt.color6: #00A8A8 + URxvt.color7: #A8A8A8 + + URxvt.color8: #000054 + URxvt.color9: #FF0054 + URxvt.color10: #00FF54 + URxvt.color11: #FFFF54 + URxvt.color12: #0000FF + URxvt.color13: #FF00FF + URxvt.color14: #00FFFF + URxvt.color15: #FFFFFF + + And here is a more complete set of non-standard colors described (not by + me) as "pretty girly". + + URxvt.cursorColor: #dc74d1 + URxvt.pointerColor: #dc74d1 + URxvt.background: #0e0e0e + URxvt.foreground: #4ad5e1 + URxvt.color0: #000000 + URxvt.color8: #8b8f93 + URxvt.color1: #dc74d1 + URxvt.color9: #dc74d1 + URxvt.color2: #0eb8c7 + URxvt.color10: #0eb8c7 + URxvt.color3: #dfe37e + URxvt.color11: #dfe37e + URxvt.color5: #9e88f0 + URxvt.color13: #9e88f0 + URxvt.color6: #73f7ff + URxvt.color14: #73f7ff + URxvt.color7: #e1dddd + URxvt.color15: #e1dddd + + Why do some characters look so much different than others? + See next entry. + + How does rxvt-unicode choose fonts? + Most fonts do not contain the full range of Unicode, which is fine. + Chances are that the font you (or the admin/package maintainer of your + system/os) have specified does not cover all the characters you want to + display. + + rxvt-unicode makes a best-effort try at finding a replacement font. + Often the result is fine, but sometimes the chosen font looks + bad/ugly/wrong. Some fonts have totally strange characters that don't + resemble the correct glyph at all, and rxvt-unicode lacks the artificial + intelligence to detect that a specific glyph is wrong: it has to believe + the font that the characters it claims to contain indeed look correct. + + In that case, select a font of your taste and add it to the font list, + e.g.: + + rxvt -fn basefont,font2,font3... + + When rxvt-unicode sees a character, it will first look at the base font. + If the base font does not contain the character, it will go to the next + font, and so on. Specifying your own fonts will also speed up this + search and use less resources within rxvt-unicode and the X-server. + + The only limitation is that none of the fonts may be larger than the + base font, as the base font defines the terminal character cell size, + which must be the same due to the way terminals work. + + Keyboard, Mouse & User Interaction + The new selection selects pieces that are too big, how can I select single words? + If you want to select e.g. alphanumeric words, you can use the following + setting: URxvt.selection.pattern-0: ([[:word:]]+) @@ -38,927 +433,508 @@ single words? Please also note that the *LeftClick Shift-LeftClik* combination also selects words like the old code. - I don't like the new selection/popups/hotkeys/perl, how do I - change/disable it? - You can disable the perl extension completely by setting the - perl-ext-common resource to the empty string, which also keeps - rxvt-unicode from initialising perl, saving memory. + I don't like the new selection/popups/hotkeys/perl, how do I change/disable it? + You can disable the perl extension completely by setting the + perl-ext-common resource to the empty string, which also keeps + rxvt-unicode from initialising perl, saving memory. + + If you only want to disable specific features, you first have to + identify which perl extension is responsible. For this, read the section + PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For example, to + disable the selection-popup and option-popup, specify this + perl-ext-common resource: + + URxvt.perl-ext-common: default,-selection-popup,-option-popup + + This will keep the default extensions, but disable the two popup + extensions. Some extensions can also be configured, for example, + scrollback search mode is triggered by M-s. You can move it to any other + combination either by setting the searchable-scrollback resource: + + URxvt.searchable-scrollback: CM-s + + The cursor moves when selecting text in the current input line, how do I switch this off? + See next entry. + + During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs strange escape sequences, how do I fix this? + These are caused by the "readline" perl extension. Under normal + circumstances, it will move your cursor around when you click into the + line that contains it. It tries hard not to do this at the wrong moment, + but when running a program that doesn't parse cursor movements or in + some cases during rlogin sessions, it fails to detect this properly. - If you only want to disable specific features, you first have to - identify which perl extension is responsible. For this, read the - section PREPACKAGED EXTENSIONS in the rxvtperl(3) manpage. For - example, to disable the selection-popup and option-popup, specify - this perl-ext-common resource: - - URxvt.perl-ext-common: default,-selection-popup,-option-popup - - This will keep the default extensions, but disable the two popup - extensions. Some extensions can also be configured, for example, - scrollback search mode is triggered by M-s. You can move it to any - other combination either by setting the searchable-scrollback - resource: - - URxvt.searchable-scrollback: CM-s - - The cursor moves when selecting text in the current input line, how do I - switch this off? - During rlogin/ssh/telnet/etc. sessions, clicking near the cursor outputs - strange escape sequences, how do I fix this? - These are caused by the "readline" perl extension. Under normal - circumstances, it will move your cursor around when you click into - the line that contains it. It tries hard not to do this at the wrong - moment, but when running a program that doesn't parse cursor - movements or in some cases during rlogin sessions, it fails to - detect this properly. - - You can permamently switch this feature off by disabling the - "readline" extension: - - URxvt.perl-ext-common: default,-readline - - Why doesn't rxvt-unicode read my resources? - Well, why, indeed? It does, in a way very similar to other X - applications. Most importantly, this means that if you or your OS - loads resources into the X display (the right way to do it), - rxvt-unicode will ignore any resource files in your home directory. - It will only read $HOME/.Xdefaults when no resources are attached to - the display. - - If you have or use an $HOME/.Xresources file, chances are that - resources are loaded into your X-server. In this case, you have to - re-login after every change (or run xrdb -merge $HOME/.Xresources). - - Also consider the form resources have to use: - - URxvt.resource: value - - If you want to use another form (there are lots of different ways of - specifying resources), make sure you understand wether and why it - works. If unsure, use the form above. - - I can't get transparency working, what am I doing wrong? - First of all, transparency isn't officially supported in - rxvt-unicode, so you are mostly on your own. Do not bug the author - about it (but you may bug everybody else). Also, if you can't get it - working consider it a rite of passage: ... and you failed. - - Here are four ways to get transparency. Do read the manpage and - option descriptions for the programs mentioned and rxvt-unicode. - Really, do it! - - 1. Use inheritPixmap: - - Esetroot wallpaper.jpg - rxvt -ip -tint red -sh 40 - - That works. If you think it doesn't, you lack transparency and - tinting support, or you are unable to read. - - 2. Use a simple pixmap and emulate pseudo-transparency. This enables - you to use effects other than tinting and shading: Just - shade/tint/whatever your picture with gimp: - - convert wallpaper.jpg -blur 20x20 -modulate 30 background.xpm - rxvt -pixmap background.xpm -pe automove-background - - That works. If you think it doesn't, you lack XPM and Perl support, - or you are unable to read. - - 3. Use an ARGB visual: - - rxvt -depth 32 -fg grey90 -bg rgba:0000/0000/4444/cccc - - This requires XFT support, and the support of your X-server. If that - doesn't work for you, blame Xorg and Keith Packard. ARGB visuals - aren't there yet, no matter what they claim. Rxvt-Unicode contains - the neccessary bugfixes and workarounds for Xft and Xlib to make it - work, but that doesn't mean that your WM has the required kludges in - place. - - 4. Use xcompmgr and let it do the job: - - xprop -frame -f _NET_WM_WINDOW_OPACITY 32c \ - -set _NET_WM_WINDOW_OPACITY 0xc0000000 - - Then click on a window you want to make transparent. Replace - 0xc0000000 by other values to change the degree of opacity. If it - doesn't work and your server crashes, you got to keep the pieces. - - Isn't rxvt supposed to be small? Don't all those features bloat? - I often get asked about this, and I think, no, they didn't cause - extra bloat. If you compare a minimal rxvt and a minimal urxvt, you - can see that the urxvt binary is larger (due to some encoding tables - always being compiled in), but it actually uses less memory (RSS) - after startup. Even with "--disable-everything", this comparison is - a bit unfair, as many features unique to urxvt (locale, encoding - conversion, iso14755 etc.) are already in use in this mode. - - text data bss drs rss filename - 98398 1664 24 15695 1824 rxvt --disable-everything - 188985 9048 66616 18222 1788 urxvt --disable-everything - - When you "--enable-everything" (which _is_ unfair, as this involves - xft and full locale/XIM support which are quite bloaty inside libX11 - and my libc), the two diverge, but not unreasnobaly so. - - text data bss drs rss filename - 163431 2152 24 20123 2060 rxvt --enable-everything - 1035683 49680 66648 29096 3680 urxvt --enable-everything - - The very large size of the text section is explained by the - east-asian encoding tables, which, if unused, take up disk space but - nothing else and can be compiled out unless you rely on X11 core - fonts that use those encodings. The BSS size comes from the 64k - emergency buffer that my c++ compiler allocates (but of course - doesn't use unless you are out of memory). Also, using an xft font - instead of a core font immediately adds a few megabytes of RSS. Xft - indeed is responsible for a lot of RSS even when not used. - - Of course, due to every character using two or four bytes instead of - one, a large scrollback buffer will ultimately make rxvt-unicode use - more memory. - - Compared to e.g. Eterm (5112k), aterm (3132k) and xterm (4680k), - this still fares rather well. And compared to some monsters like - gnome-terminal (21152k + extra 4204k in separate processes) or - konsole (22200k + extra 43180k in daemons that stay around after - exit, plus half a minute of startup time, including the hundreds of - warnings it spits out), it fares extremely well *g*. - - Why C++, isn't that unportable/bloated/uncool? - Is this a question? :) It comes up very often. The simple answer is: - I had to write it, and C++ allowed me to write and maintain it in a - fraction of the time and effort (which is a scarce resource for me). - Put even shorter: It simply wouldn't exist without C++. - - My personal stance on this is that C++ is less portable than C, but - in the case of rxvt-unicode this hardly matters, as its portability - limits are defined by things like X11, pseudo terminals, locale - support and unix domain sockets, which are all less portable than - C++ itself. - - Regarding the bloat, see the above question: It's easy to write - programs in C that use gobs of memory, an certainly possible to - write programs in C++ that don't. C++ also often comes with large - libraries, but this is not necessarily the case with GCC. Here is - what rxvt links against on my system with a minimal config: - - libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) - libc.so.6 => /lib/libc.so.6 (0x00002aaaaadde000) - libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab01d000) - /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) - - And here is rxvt-unicode: - - libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0x00002aaaaabc3000) - libgcc_s.so.1 => /lib/libgcc_s.so.1 (0x00002aaaaada2000) - libc.so.6 => /lib/libc.so.6 (0x00002aaaaaeb0000) - libdl.so.2 => /lib/libdl.so.2 (0x00002aaaab0ee000) - /lib64/ld-linux-x86-64.so.2 (0x00002aaaaaaab000) - - No large bloated libraries (of course, none were linked in - statically), except maybe libX11 :) - - Does it support tabs, can I have a tabbed rxvt-unicode? - Beginning with version 7.3, there is a perl extension that - implements a simple tabbed terminal. It is installed by default, so - any of these should give you tabs: - - rxvt -pe tabbed - - URxvt.perl-ext-common: default,tabbed - - It will also work fine with tabbing functionality of many window - managers or similar tabbing programs, and its embedding-features - allow it to be embedded into other programs, as witnessed by - doc/rxvt-tabbed or the upcoming "Gtk2::URxvt" perl module, which - features a tabbed urxvt (murxvt) terminal as an example embedding - application. - - How do I know which rxvt-unicode version I'm using? - The version number is displayed with the usage (-h). Also the escape - sequence "ESC [ 8 n" sets the window title to the version number. - When using the rxvtc client, the version displayed is that of the - daemon. - - I am using Debian GNU/Linux and have a problem... - The Debian GNU/Linux package of rxvt-unicode in sarge contains large - patches that considerably change the behaviour of rxvt-unicode (but - unfortunately this notice has been removed). Before reporting a bug - to the original rxvt-unicode author please download and install the - genuine version () and try - to reproduce the problem. If you cannot, chances are that the - problems are specific to Debian GNU/Linux, in which case it should - be reported via the Debian Bug Tracking System (use "reportbug" to - report the bug). - - For other problems that also affect the Debian package, you can and - probably should use the Debian BTS, too, because, after all, it's - also a bug in the Debian version and it serves as a reminder for - other users that might encounter the same issue. - - I am maintaining rxvt-unicode for distribution/OS XXX, any - recommendation? - You should build one binary with the default options. configure now - enables most useful options, and the trend goes to making them - runtime-switchable, too, so there is usually no drawback to enbaling - them, except higher disk and possibly memory usage. The perl - interpreter should be enabled, as important functionality (menus, - selection, likely more in the future) depends on it. - - You should not overwrite the "perl-ext-common" snd "perl-ext" - resources system-wide (except maybe with "defaults"). This will - result in useful behaviour. If your distribution aims at low memory, - add an empty "perl-ext-common" resource to the app-defaults file. - This will keep the perl interpreter disabled until the user enables - it. - - If you can/want build more binaries, I recommend building a minimal - one with "--disable-everything" (very useful) and a maximal one with - "--enable-everything" (less useful, it will be very big due to a lot - of encodings built-in that increase download times and are rarely - used). - - I need to make it setuid/setgid to support utmp/ptys on my OS, is this - safe? - It should be, starting with release 7.1. You are encouraged to - properly install urxvt with privileges necessary for your OS now. - - When rxvt-unicode detects that it runs setuid or setgid, it will - fork into a helper process for privileged operations (pty handling - on some systems, utmp/wtmp/lastlog handling on others) and drop - privileges immediately. This is much safer than most other terminals - that keep privileges while running (but is more relevant to urxvt, - as it contains things as perl interpreters, which might be "helpful" - to attackers). - - This forking is done as the very first within main(), which is very - early and reduces possible bugs to initialisation code run before - main(), or things like the dynamic loader of your system, which - should result in very little risk. - - When I log-in to another system it tells me about missing terminfo data? - The terminal description used by rxvt-unicode is not as widely - available as that for xterm, or even rxvt (for which the same - problem often arises). - - The correct solution for this problem is to install the terminfo, - this can be done like this (with ncurses' infocmp): - - REMOTE=remotesystem.domain - infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" - - ... or by installing rxvt-unicode normally on the remote system, - - If you cannot or do not want to do this, then you can simply set - "TERM=rxvt" or even "TERM=xterm", and live with the small number of - problems arising, which includes wrong keymapping, less and - different colours and some refresh errors in fullscreen - applications. It's a nice quick-and-dirty workaround for rare cases, - though. - - If you always want to do this (and are fine with the consequences) - you can either recompile rxvt-unicode with the desired TERM value or - use a resource to set it: - - URxvt.termName: rxvt - - If you don't plan to use rxvt (quite common...) you could also - replace the rxvt terminfo file with the rxvt-unicode one. - - "tic" outputs some error when compiling the terminfo entry. - Most likely it's the empty definition for "enacs=". Just replace it - by "enacs=\E[0@" and try again. - - "bash"'s readline does not work correctly under rxvt. - I need a termcap file entry. - One reason you might want this is that some distributions or - operating systems still compile some programs using the - long-obsoleted termcap library (Fedora Core's bash is one example) - and rely on a termcap entry for "rxvt-unicode". - - You could use rxvt's termcap entry with resonable results in many - cases. You can also create a termcap entry by using terminfo's - infocmp program like this: - - infocmp -C rxvt-unicode - - Or you could use this termcap entry, generated by the command above: - - rxvt-unicode|rxvt-unicode terminal (X Window System):\ - :am:bw:eo:km:mi:ms:xn:xo:\ - :co#80:it#8:li#24:lm#0:\ - :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ - :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ - :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ - :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ - :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ - :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ - :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ - :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ - :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ - :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ - :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ - :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ - :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ - :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ - :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ - :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ - :vs=\E[?25h: - - Why does "ls" no longer have coloured output? - The "ls" in the GNU coreutils unfortunately doesn't use terminfo to - decide wether a terminal has colour, but uses it's own configuration - file. Needless to say, "rxvt-unicode" is not in it's default file - (among with most other terminals supporting colour). Either add: - - TERM rxvt-unicode - - to "/etc/DIR_COLORS" or simply add: - - alias ls='ls --color=auto' - - to your ".profile" or ".bashrc". - - Why doesn't vim/emacs etc. use the 88 colour mode? - Why doesn't vim/emacs etc. make use of italic? - Why are the secondary screen-related options not working properly? - Make sure you are using "TERM=rxvt-unicode". Some pre-packaged - distributions (most notably Debian GNU/Linux) break rxvt-unicode by - setting "TERM" to "rxvt", which doesn't have these extra features. - Unfortunately, some of these (most notably, again, Debian GNU/Linux) - furthermore fail to even install the "rxvt-unicode" terminfo file, - so you will need to install it on your own (See the question When I - log-in to another system it tells me about missing terminfo data? on - how to do this). - - My numerical keypad acts weird and generates differing output? - Some Debian GNUL/Linux users seem to have this problem, although no - specific details were reported so far. It is possible that this is - caused by the wrong "TERM" setting, although the details of wether - and how this can happen are unknown, as "TERM=rxvt" should offer a - compatible keymap. See the answer to the previous question, and - please report if that helped. - - Rxvt-unicode does not seem to understand the selected encoding? - Unicode does not seem to work? - If you encounter strange problems like typing an accented character - but getting two unrelated other characters or similar, or if program - output is subtly garbled, then you should check your locale - settings. - - Rxvt-unicode must be started with the same "LC_CTYPE" setting as the - programs. Often rxvt-unicode is started in the "C" locale, while the - login script running within the rxvt-unicode window changes the - locale to something else, e.g. "en_GB.UTF-8". Needless to say, this - is not going to work. - - The best thing is to fix your startup environment, as you will - likely run into other problems. If nothing works you can try this in - your .profile. - - printf '\e]701;%s\007' "$LC_CTYPE" - - If this doesn't work, then maybe you use a "LC_CTYPE" specification - not supported on your systems. Some systems have a "locale" command - which displays this (also, "perl -e0" can be used to check locale - settings, as it will complain loudly if it cannot set the locale). - If it displays something like: - - locale: Cannot set LC_CTYPE to default locale: ... - - Then the locale you specified is not supported on your system. - - If nothing works and you are sure that everything is set correctly - then you will need to remember a little known fact: Some programs - just don't support locales :( - - Why do some characters look so much different than others? - How does rxvt-unicode choose fonts? - Most fonts do not contain the full range of Unicode, which is fine. - Chances are that the font you (or the admin/package maintainer of - your system/os) have specified does not cover all the characters you - want to display. - - rxvt-unicode makes a best-effort try at finding a replacement font. - Often the result is fine, but sometimes the chosen font looks - bad/ugly/wrong. Some fonts have totally strange characters that - don't resemble the correct glyph at all, and rxvt-unicode lacks the - artificial intelligence to detect that a specific glyph is wrong: it - has to believe the font that the characters it claims to contain - indeed look correct. - - In that case, select a font of your taste and add it to the font - list, e.g.: - - rxvt -fn basefont,font2,font3... - - When rxvt-unicode sees a character, it will first look at the base - font. If the base font does not contain the character, it will go to - the next font, and so on. Specifying your own fonts will also speed - up this search and use less resources within rxvt-unicode and the - X-server. - - The only limitation is that none of the fonts may be larger than the - base font, as the base font defines the terminal character cell - size, which must be the same due to the way terminals work. - - Why do some chinese characters look so different than others? - This is because there is a difference between script and language -- - rxvt-unicode does not know which language the text that is output - is, as it only knows the unicode character codes. If rxvt-unicode - first sees a japanese/chinese character, it might choose a japanese - font for display. Subsequent japanese characters will use that font. - Now, many chinese characters aren't represented in japanese fonts, - so when the first non-japanese character comes up, rxvt-unicode will - look for a chinese font -- unfortunately at this point, it will - still use the japanese font for chinese characters that are also in - the japanese font. - - The workaround is easy: just tag a chinese font at the end of your - font list (see the previous question). The key is to view the font - list as a preference list: If you expect more japanese, list a - japanese font first. If you expect more chinese, put a chinese font - first. - - In the future it might be possible to switch language preferences at - runtime (the internal data structure has no problem with using - different fonts for the same character at the same time, but no - interface for this has been designed yet). - - Until then, you might get away with switching fonts at runtime (see - "Can I switch the fonts at runtime?" later in this document). - - Why does rxvt-unicode sometimes leave pixel droppings? - Most fonts were not designed for terminal use, which means that - character size varies a lot. A font that is otherwise fine for - terminal use might contain some characters that are simply too wide. - Rxvt-unicode will avoid these characters. For characters that are - just "a bit" too wide a special "careful" rendering mode is used - that redraws adjacent characters. - - All of this requires that fonts do not lie about character sizes, - however: Xft fonts often draw glyphs larger than their acclaimed - bounding box, and rxvt-unicode has no way of detecting this (the - correct way is to ask for the character bounding box, which - unfortunately is wrong in these cases). - - It's not clear (to me at least), wether this is a bug in Xft, - freetype, or the respective font. If you encounter this problem you - might try using the "-lsp" option to give the font more height. If - that doesn't work, you might be forced to use a different font. - - All of this is not a problem when using X11 core fonts, as their - bounding box data is correct. - - On Solaris 9, many line-drawing characters are too wide. - Seems to be a known bug, read - . Some people use the - following ugly workaround to get non-double-wide-characters working: - - #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) - - My Compose (Multi_key) key is no longer working. - The most common causes for this are that either your locale is not - set correctly, or you specified a preeditStyle that is not supported - by your input method. For example, if you specified OverTheSpot and - your input method (e.g. the default input method handling Compose - keys) does not support this (for instance because it is not visual), - then rxvt-unicode will continue without an input method. - - In this case either do not specify a preeditStyle or specify more - than one pre-edit style, such as OverTheSpot,Root,None. - - I cannot type "Ctrl-Shift-2" to get an ASCII NUL character due to ISO - 14755 - Either try "Ctrl-2" alone (it often is mapped to ASCII NUL even on - international keyboards) or simply use ISO 14755 support to your - advantage, typing to get a ASCII NUL. This works for - other codes, too, such as "Ctrl-Shift-1-d" to type the default - telnet escape character and so on. - - How can I keep rxvt-unicode from using reverse video so much? - First of all, make sure you are running with the right terminal - settings ("TERM=rxvt-unicode"), which will get rid of most of these - effects. Then make sure you have specified colours for italic and - bold, as otherwise rxvt-unicode might use reverse video to simulate - the effect: - - URxvt.colorBD: white - URxvt.colorIT: green - - Some programs assume totally weird colours (red instead of blue), how - can I fix that? - For some unexplainable reason, some rare programs assume a very - weird colour palette when confronted with a terminal with more than - the standard 8 colours (rxvt-unicode supports 88). The right fix is, - of course, to fix these programs not to assume non-ISO colours - without very good reasons. - - In the meantime, you can either edit your "rxvt-unicode" terminfo - definition to only claim 8 colour support or use "TERM=rxvt", which - will fix colours but keep you from using other rxvt-unicode - features. - - I am on FreeBSD and rxvt-unicode does not seem to work at all. - Rxvt-unicode requires the symbol "__STDC_ISO_10646__" to be defined - in your compile environment, or an implementation that implements - it, wether it defines the symbol or not. "__STDC_ISO_10646__" - requires that wchar_t is represented as unicode. - - As you might have guessed, FreeBSD does neither define this symobl - nor does it support it. Instead, it uses it's own internal - representation of wchar_t. This is, of course, completely fine with - respect to standards. - - However, that means rxvt-unicode only works in "POSIX", "ISO-8859-1" - and "UTF-8" locales under FreeBSD (which all use Unicode as wchar_t. - - "__STDC_ISO_10646__" is the only sane way to support multi-language - apps in an OS, as using a locale-dependent (and non-standardized) - representation of wchar_t makes it impossible to convert between - wchar_t (as used by X11 and your applications) and any other - encoding without implementing OS-specific-wrappers for each and - every locale. There simply are no APIs to convert wchar_t into - anything except the current locale encoding. - - Some applications (such as the formidable mlterm) work around this - by carrying their own replacement functions for character set - handling with them, and either implementing OS-dependent hacks or - doing multiple conversions (which is slow and unreliable in case the - OS implements encodings slightly different than the terminal - emulator). - - The rxvt-unicode author insists that the right way to fix this is in - the system libraries once and for all, instead of forcing every app - to carry complete replacements for them :) - - I use Solaris 9 and it doesn't compile/work/etc. - Try the diff in doc/solaris9.patch as a base. It fixes the worst - problems with "wcwidth" and a compile problem. - - How can I use rxvt-unicode under cygwin? - rxvt-unicode should compile and run out of the box on cygwin, using - the X11 libraries that come with cygwin. libW11 emulation is no - longer supported (and makes no sense, either, as it only supported a - single font). I recommend starting the X-server in "-multiwindow" or - "-rootless" mode instead, which will result in similar look&feel as - the old libW11 emulation. - - At the time of this writing, cygwin didn't seem to support any - multi-byte encodings (you might try "LC_CTYPE=C-UTF-8"), so you are - likely limited to 8-bit encodings. - - How does rxvt-unicode determine the encoding to use? - Is there an option to switch encodings? - Unlike some other terminals, rxvt-unicode has no encoding switch, - and no specific "utf-8" mode, such as xterm. In fact, it doesn't - even know about UTF-8 or any other encodings with respect to - terminal I/O. - - The reasons is that there exists a perfectly fine mechanism for - selecting the encoding, doing I/O and (most important) communicating - this to all applications so everybody agrees on character properties - such as width and code number. This mechanism is the *locale*. - Applications not using that info will have problems (for example, - "xterm" gets the width of characters wrong as it uses it's own, - locale-independent table under all locales). - - Rxvt-unicode uses the "LC_CTYPE" locale category to select encoding. - All programs doing the same (that is, most) will automatically agree - in the interpretation of characters. - - Unfortunately, there is no system-independent way to select locales, - nor is there a standard on how locale specifiers will look like. - - On most systems, the content of the "LC_CTYPE" environment variable - contains an arbitrary string which corresponds to an - already-installed locale. Common names for locales are - "en_US.UTF-8", "de_DE.ISO-8859-15", "ja_JP.EUC-JP", i.e. - "language_country.encoding", but other forms (i.e. "de" or "german") - are also common. - - Rxvt-unicode ignores all other locale categories, and except for the - encoding, ignores country or language-specific settings, i.e. - "de_DE.UTF-8" and "ja_JP.UTF-8" are the normally same to - rxvt-unicode. - - If you want to use a specific encoding you have to make sure you - start rxvt-unicode with the correct "LC_CTYPE" category. - - Can I switch locales at runtime? - Yes, using an escape sequence. Try something like this, which sets - rxvt-unicode's idea of "LC_CTYPE". - - printf '\e]701;%s\007' ja_JP.SJIS - - See also the previous answer. - - Sometimes this capability is rather handy when you want to work in - one locale (e.g. "de_DE.UTF-8") but some programs don't support it - (e.g. UTF-8). For example, I use this script to start "xjdic", which - first switches to a locale supported by xjdic and back later: - - printf '\e]701;%s\007' ja_JP.SJIS - xjdic -js - printf '\e]701;%s\007' de_DE.UTF-8 - - You can also use xterm's "luit" program, which usually works fine, - except for some locales where character width differs between - program- and rxvt-unicode-locales. - - Can I switch the fonts at runtime? - Yes, using an escape sequence. Try something like this, which has - the same effect as using the "-fn" switch, and takes effect - immediately: - - printf '\e]50;%s\007' "9x15bold,xft:Kochi Gothic" - - This is useful if you e.g. work primarily with japanese (and prefer - a japanese font), but you have to switch to chinese temporarily, - where japanese fonts would only be in your way. - - You can think of this as a kind of manual ISO-2022 switching. - - Why do italic characters look as if clipped? - Many fonts have difficulties with italic characters and hinting. For - example, the otherwise very nicely hinted font "xft:Bitstream Vera - Sans Mono" completely fails in it's italic face. A workaround might - be to enable freetype autohinting, i.e. like this: - - URxvt.italicFont: xft:Bitstream Vera Sans Mono:italic:autohint=true - URxvt.boldItalicFont: xft:Bitstream Vera Sans Mono:bold:italic:autohint=true - - My input method wants but I want UTF-8, what can I do? - You can specify separate locales for the input method and the rest - of the terminal, using the resource "imlocale": - - URxvt.imlocale: ja_JP.EUC-JP - - Now you can start your terminal with "LC_CTYPE=ja_JP.UTF-8" and - still use your input method. Please note, however, that you will not - be able to input characters outside "EUC-JP" in a normal way then, - as your input method limits you. - - Rxvt-unicode crashes when the X Input Method changes or exits. - Unfortunately, this is unavoidable, as the XIM protocol is racy by - design. Applications can avoid some crashes at the expense of memory - leaks, and Input Methods can avoid some crashes by careful ordering - at exit time. kinput2 (and derived input methods) generally - succeeds, while SCIM (or similar input methods) fails. In the end, - however, crashes cannot be completely avoided even if both sides - cooperate. - - So the only workaround is not to kill your Input Method Servers. - - Rxvt-unicode uses gobs of memory, how can I reduce that? - Rxvt-unicode tries to obey the rule of not charging you for - something you don't use. One thing you should try is to configure - out all settings that you don't need, for example, Xft support is a - resource hog by design, when used. Compiling it out ensures that no - Xft font will be loaded accidentally when rxvt-unicode tries to find - a font for your characters. - - Also, many people (me included) like large windows and even larger - scrollback buffers: Without "--enable-unicode3", rxvt-unicode will - use 6 bytes per screen cell. For a 160x?? window this amounts to - almost a kilobyte per line. A scrollback buffer of 10000 lines will - then (if full) use 10 Megabytes of memory. With "--enable-unicode3" - it gets worse, as rxvt-unicode then uses 8 bytes per screen cell. - - Can I speed up Xft rendering somehow? - Yes, the most obvious way to speed it up is to avoid Xft entirely, - as it is simply slow. If you still want Xft fonts you might try to - disable antialiasing (by appending ":antialias=false"), which saves - lots of memory and also speeds up rendering considerably. - - Rxvt-unicode doesn't seem to anti-alias its fonts, what is wrong? - Rxvt-unicode will use whatever you specify as a font. If it needs to - fall back to it's default font search list it will prefer X11 core - fonts, because they are small and fast, and then use Xft fonts. It - has antialiasing disabled for most of them, because the author - thinks they look best that way. - - If you want antialiasing, you have to specify the fonts manually. - - Mouse cut/paste suddenly no longer works. - Make sure that mouse reporting is actually turned off since killing - some editors prematurely may leave the mouse in mouse report mode. - I've heard that tcsh may use mouse reporting unless it otherwise - specified. A quick check is to see if cut/paste works when the Alt - or Shift keys are depressed. - - What's with this bold/blink stuff? - If no bold colour is set via "colorBD:", bold will invert text using - the standard foreground colour. - - For the standard background colour, blinking will actually make the - text blink when compiled with "--enable-blinking". with standard - colours. Without "--enable-blinking", the blink attribute will be - ignored. - - On ANSI colours, bold/blink attributes are used to set - high-intensity foreground/background colors. - - color0-7 are the low-intensity colors. - - color8-15 are the corresponding high-intensity colors. - - I don't like the screen colors. How do I change them? - You can change the screen colors at run-time using ~/.Xdefaults - resources (or as long-options). - - Here are values that are supposed to resemble a VGA screen, - including the murky brown that passes for low-intensity yellow: - - URxvt.color0: #000000 - URxvt.color1: #A80000 - URxvt.color2: #00A800 - URxvt.color3: #A8A800 - URxvt.color4: #0000A8 - URxvt.color5: #A800A8 - URxvt.color6: #00A8A8 - URxvt.color7: #A8A8A8 - - URxvt.color8: #000054 - URxvt.color9: #FF0054 - URxvt.color10: #00FF54 - URxvt.color11: #FFFF54 - URxvt.color12: #0000FF - URxvt.color13: #FF00FF - URxvt.color14: #00FFFF - URxvt.color15: #FFFFFF - - And here is a more complete set of non-standard colors described - (not by me) as "pretty girly". - - URxvt.cursorColor: #dc74d1 - URxvt.pointerColor: #dc74d1 - URxvt.background: #0e0e0e - URxvt.foreground: #4ad5e1 - URxvt.color0: #000000 - URxvt.color8: #8b8f93 - URxvt.color1: #dc74d1 - URxvt.color9: #dc74d1 - URxvt.color2: #0eb8c7 - URxvt.color10: #0eb8c7 - URxvt.color3: #dfe37e - URxvt.color11: #dfe37e - URxvt.color5: #9e88f0 - URxvt.color13: #9e88f0 - URxvt.color6: #73f7ff - URxvt.color14: #73f7ff - URxvt.color7: #e1dddd - URxvt.color15: #e1dddd - - How can I start rxvtd in a race-free way? - Try "rxvtd -f -o", which tells rxvtd to open the display, create the - listening socket and then fork. - - What's with the strange Backspace/Delete key behaviour? - Assuming that the physical Backspace key corresponds to the - BackSpace keysym (not likely for Linux ... see the following - question) there are two standard values that can be used for - Backspace: "^H" and "^?". - - Historically, either value is correct, but rxvt-unicode adopts the - debian policy of using "^?" when unsure, because it's the one only - only correct choice :). - - Rxvt-unicode tries to inherit the current stty settings and uses the - value of `erase' to guess the value for backspace. If rxvt-unicode - wasn't started from a terminal (say, from a menu or by remote - shell), then the system value of `erase', which corresponds to - CERASE in , will be used (which may not be the same as - your stty setting). - - For starting a new rxvt-unicode: - - # use Backspace = ^H - $ stty erase ^H - $ rxvt - - # use Backspace = ^? - $ stty erase ^? - $ rxvt - - Toggle with "ESC [ 36 h" / "ESC [ 36 l". - - For an existing rxvt-unicode: - - # use Backspace = ^H - $ stty erase ^H - $ echo -n "^[[36h" - - # use Backspace = ^? - $ stty erase ^? - $ echo -n "^[[36l" - - This helps satisfy some of the Backspace discrepancies that occur, - but if you use Backspace = "^H", make sure that the termcap/terminfo - value properly reflects that. - - The Delete key is a another casualty of the ill-defined Backspace - problem. To avoid confusion between the Backspace and Delete keys, - the Delete key has been assigned an escape sequence to match the - vt100 for Execute ("ESC [ 3 ~") and is in the supplied - termcap/terminfo. - - Some other Backspace problems: - - some editors use termcap/terminfo, some editors (vim I'm told) - expect Backspace = ^H, GNU Emacs (and Emacs-like editors) use ^H for - help. - - Perhaps someday this will all be resolved in a consistent manner. - - I don't like the key-bindings. How do I change them? - There are some compile-time selections available via configure. - Unless you have run "configure" with the "--disable-resources" - option you can use the `keysym' resource to alter the keystrings - associated with keysyms. - - Here's an example for a URxvt session started using "rxvt -name - URxvt" - - URxvt.keysym.Home: \033[1~ - URxvt.keysym.End: \033[4~ - URxvt.keysym.C-apostrophe: \033 - URxvt.keysym.C-slash: \033 - URxvt.keysym.C-semicolon: \033 - URxvt.keysym.C-grave: \033 - URxvt.keysym.C-comma: \033 - URxvt.keysym.C-period: \033 - URxvt.keysym.C-0x60: \033 - URxvt.keysym.C-Tab: \033 - URxvt.keysym.C-Return: \033 - URxvt.keysym.S-Return: \033 - URxvt.keysym.S-space: \033 - URxvt.keysym.M-Up: \033 - URxvt.keysym.M-Down: \033 - URxvt.keysym.M-Left: \033 - URxvt.keysym.M-Right: \033 - URxvt.keysym.M-C-0: list \033 - URxvt.keysym.M-C-a: list \033 - URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 - - See some more examples in the documentation for the keysym resource. - - I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How - do I make use of them? For example, the Sun Keyboard type 4 has the - following mappings that rxvt-unicode doesn't recognize. - KP_Insert == Insert - F22 == Print - F27 == Home - F29 == Prior - F33 == End - F35 == Next - - Rather than have rxvt-unicode try to accommodate all the various - possible keyboard mappings, it is better to use `xmodmap' to remap - the keys as required for your particular machine. - - How do I distinguish wether I'm running rxvt-unicode or a regular xterm? - I need this to decide about setting colors etc. - rxvt and rxvt-unicode always export the variable "COLORTERM", so you - can check and see if that is set. Note that several programs, JED, - slrn, Midnight Commander automatically check this variable to decide - whether or not to use color. - - How do I set the correct, full IP address for the DISPLAY variable? - If you've compiled rxvt-unicode with DISPLAY_IS_IP and have enabled - insecure mode then it is possible to use the following shell script - snippets to correctly set the display. If your version of - rxvt-unicode wasn't also compiled with ESCZ_ANSWER (as assumed in - these snippets) then the COLORTERM variable can be used to - distinguish rxvt-unicode from a regular xterm. - - Courtesy of Chuck Blake with the following shell - script snippets: - - # Bourne/Korn/POSIX family of shells: - [ ${TERM:-foo} = foo ] && TERM=xterm # assume an xterm if we don't know - if [ ${TERM:-foo} = xterm ]; then - stty -icanon -echo min 0 time 15 # see if enhanced rxvt or not - echo -n '^[Z' - read term_id - stty icanon echo - if [ ""${term_id} = '^[[?1;2C' -a ${DISPLAY:-foo} = foo ]; then - echo -n '^[[7n' # query the rxvt we are in for the DISPLAY string - read DISPLAY # set it in our local shell - fi - fi - - How do I compile the manual pages for myself? - You need to have a recent version of perl installed as - /usr/bin/perl, one that comes with pod2man, pod2text and pod2html. - Then go to the doc subdirectory and enter "make alldoc". - - My question isn't answered here, can I ask a human? - Before sending me mail, you could go to IRC: "irc.freenode.net", - channel "#rxvt-unicode" has some rxvt-unicode enthusiasts that might - be interested in learning about new and exciting problems (but not - FAQs :). + You can permamently switch this feature off by disabling the "readline" + extension: + + URxvt.perl-ext-common: default,-readline + + My numerical keypad acts weird and generates differing output? + Some Debian GNUL/Linux users seem to have this problem, although no + specific details were reported so far. It is possible that this is + caused by the wrong "TERM" setting, although the details of wether and + how this can happen are unknown, as "TERM=rxvt" should offer a + compatible keymap. See the answer to the previous question, and please + report if that helped. + + My Compose (Multi_key) key is no longer working. + The most common causes for this are that either your locale is not set + correctly, or you specified a preeditStyle that is not supported by your + input method. For example, if you specified OverTheSpot and your input + method (e.g. the default input method handling Compose keys) does not + support this (for instance because it is not visual), then rxvt-unicode + will continue without an input method. + + In this case either do not specify a preeditStyle or specify more than + one pre-edit style, such as OverTheSpot,Root,None. + + I cannot type "Ctrl-Shift-2" to get an ASCII NUL character due to ISO 14755 + Either try "Ctrl-2" alone (it often is mapped to ASCII NUL even on + international keyboards) or simply use ISO 14755 support to your + advantage, typing to get a ASCII NUL. This works for + other codes, too, such as "Ctrl-Shift-1-d" to type the default telnet + escape character and so on. + + Mouse cut/paste suddenly no longer works. + Make sure that mouse reporting is actually turned off since killing some + editors prematurely may leave the mouse in mouse report mode. I've heard + that tcsh may use mouse reporting unless it otherwise specified. A quick + check is to see if cut/paste works when the Alt or Shift keys are + depressed. + + What's with the strange Backspace/Delete key behaviour? + Assuming that the physical Backspace key corresponds to the BackSpace + keysym (not likely for Linux ... see the following question) there are + two standard values that can be used for Backspace: "^H" and "^?". + + Historically, either value is correct, but rxvt-unicode adopts the + debian policy of using "^?" when unsure, because it's the one only only + correct choice :). + + Rxvt-unicode tries to inherit the current stty settings and uses the + value of `erase' to guess the value for backspace. If rxvt-unicode + wasn't started from a terminal (say, from a menu or by remote shell), + then the system value of `erase', which corresponds to CERASE in + , will be used (which may not be the same as your stty + setting). + + For starting a new rxvt-unicode: + + # use Backspace = ^H + $ stty erase ^H + $ rxvt + + # use Backspace = ^? + $ stty erase ^? + $ rxvt + + Toggle with "ESC [ 36 h" / "ESC [ 36 l". + + For an existing rxvt-unicode: + + # use Backspace = ^H + $ stty erase ^H + $ echo -n "^[[36h" + + # use Backspace = ^? + $ stty erase ^? + $ echo -n "^[[36l" + + This helps satisfy some of the Backspace discrepancies that occur, but + if you use Backspace = "^H", make sure that the termcap/terminfo value + properly reflects that. + + The Delete key is a another casualty of the ill-defined Backspace + problem. To avoid confusion between the Backspace and Delete keys, the + Delete key has been assigned an escape sequence to match the vt100 for + Execute ("ESC [ 3 ~") and is in the supplied termcap/terminfo. + + Some other Backspace problems: + + some editors use termcap/terminfo, some editors (vim I'm told) expect + Backspace = ^H, GNU Emacs (and Emacs-like editors) use ^H for help. + + Perhaps someday this will all be resolved in a consistent manner. + + I don't like the key-bindings. How do I change them? + There are some compile-time selections available via configure. Unless + you have run "configure" with the "--disable-resources" option you can + use the `keysym' resource to alter the keystrings associated with + keysyms. + + Here's an example for a URxvt session started using "rxvt -name URxvt" + + URxvt.keysym.Home: \033[1~ + URxvt.keysym.End: \033[4~ + URxvt.keysym.C-apostrophe: \033 + URxvt.keysym.C-slash: \033 + URxvt.keysym.C-semicolon: \033 + URxvt.keysym.C-grave: \033 + URxvt.keysym.C-comma: \033 + URxvt.keysym.C-period: \033 + URxvt.keysym.C-0x60: \033 + URxvt.keysym.C-Tab: \033 + URxvt.keysym.C-Return: \033 + URxvt.keysym.S-Return: \033 + URxvt.keysym.S-space: \033 + URxvt.keysym.M-Up: \033 + URxvt.keysym.M-Down: \033 + URxvt.keysym.M-Left: \033 + URxvt.keysym.M-Right: \033 + URxvt.keysym.M-C-0: list \033 + URxvt.keysym.M-C-a: list \033 + URxvt.keysym.F12: command:\033]701;zh_CN.GBK\007 + + See some more examples in the documentation for the keysym resource. + + I'm using keyboard model XXX that has extra Prior/Next/Insert keys. How do I make use of them? For example, the Sun Keyboard type 4 has the following map + KP_Insert == Insert + F22 == Print + F27 == Home + F29 == Prior + F33 == End + F35 == Next + + Rather than have rxvt-unicode try to accommodate all the various + possible keyboard mappings, it is better to use `xmodmap' to remap the + keys as required for your particular machine. + + Terminal Configuration + Why doesn't rxvt-unicode read my resources? + Well, why, indeed? It does, in a way very similar to other X + applications. Most importantly, this means that if you or your OS loads + resources into the X display (the right way to do it), rxvt-unicode will + ignore any resource files in your home directory. It will only read + $HOME/.Xdefaults when no resources are attached to the display. + + If you have or use an $HOME/.Xresources file, chances are that resources + are loaded into your X-server. In this case, you have to re-login after + every change (or run xrdb -merge $HOME/.Xresources). + + Also consider the form resources have to use: + + URxvt.resource: value + + If you want to use another form (there are lots of different ways of + specifying resources), make sure you understand wether and why it works. + If unsure, use the form above. + + When I log-in to another system it tells me about missing terminfo data? + The terminal description used by rxvt-unicode is not as widely available + as that for xterm, or even rxvt (for which the same problem often + arises). + + The correct solution for this problem is to install the terminfo, this + can be done like this (with ncurses' infocmp): + + REMOTE=remotesystem.domain + infocmp rxvt-unicode | ssh $REMOTE "cat >/tmp/ti && tic /tmp/ti" + + ... or by installing rxvt-unicode normally on the remote system, + + If you cannot or do not want to do this, then you can simply set + "TERM=rxvt" or even "TERM=xterm", and live with the small number of + problems arising, which includes wrong keymapping, less and different + colours and some refresh errors in fullscreen applications. It's a nice + quick-and-dirty workaround for rare cases, though. + + If you always want to do this (and are fine with the consequences) you + can either recompile rxvt-unicode with the desired TERM value or use a + resource to set it: + + URxvt.termName: rxvt + + If you don't plan to use rxvt (quite common...) you could also replace + the rxvt terminfo file with the rxvt-unicode one. + + "tic" outputs some error when compiling the terminfo entry. + Most likely it's the empty definition for "enacs=". Just replace it by + "enacs=\E[0@" and try again. + + "bash"'s readline does not work correctly under rxvt. + See next entry. + + I need a termcap file entry. + One reason you might want this is that some distributions or operating + systems still compile some programs using the long-obsoleted termcap + library (Fedora Core's bash is one example) and rely on a termcap entry + for "rxvt-unicode". + + You could use rxvt's termcap entry with resonable results in many cases. + You can also create a termcap entry by using terminfo's infocmp program + like this: + + infocmp -C rxvt-unicode + + Or you could use this termcap entry, generated by the command above: + + rxvt-unicode|rxvt-unicode terminal (X Window System):\ + :am:bw:eo:km:mi:ms:xn:xo:\ + :co#80:it#8:li#24:lm#0:\ + :AL=\E[%dL:DC=\E[%dP:DL=\E[%dM:DO=\E[%dB:IC=\E[%d@:\ + :K1=\EOw:K2=\EOu:K3=\EOy:K4=\EOq:K5=\EOs:LE=\E[%dD:\ + :RI=\E[%dC:SF=\E[%dS:SR=\E[%dT:UP=\E[%dA:ae=\E(B:al=\E[L:\ + :as=\E(0:bl=^G:cd=\E[J:ce=\E[K:cl=\E[H\E[2J:\ + :cm=\E[%i%d;%dH:cr=^M:cs=\E[%i%d;%dr:ct=\E[3g:dc=\E[P:\ + :dl=\E[M:do=^J:ec=\E[%dX:ei=\E[4l:ho=\E[H:\ + :i1=\E[?47l\E=\E[?1l:ic=\E[@:im=\E[4h:\ + :is=\E[r\E[m\E[2J\E[H\E[?7h\E[?1;3;4;6l\E[4l:\ + :k1=\E[11~:k2=\E[12~:k3=\E[13~:k4=\E[14~:k5=\E[15~:\ + :k6=\E[17~:k7=\E[18~:k8=\E[19~:k9=\E[20~:kD=\E[3~:\ + :kI=\E[2~:kN=\E[6~:kP=\E[5~:kb=\177:kd=\EOB:ke=\E[?1l\E>:\ + :kh=\E[7~:kl=\EOD:kr=\EOC:ks=\E[?1h\E=:ku=\EOA:le=^H:\ + :mb=\E[5m:md=\E[1m:me=\E[m\017:mr=\E[7m:nd=\E[C:rc=\E8:\ + :sc=\E7:se=\E[27m:sf=^J:so=\E[7m:sr=\EM:st=\EH:ta=^I:\ + :te=\E[r\E[?1049l:ti=\E[?1049h:ue=\E[24m:up=\E[A:\ + :us=\E[4m:vb=\E[?5h\E[?5l:ve=\E[?25h:vi=\E[?25l:\ + :vs=\E[?25h: + + Why does "ls" no longer have coloured output? + The "ls" in the GNU coreutils unfortunately doesn't use terminfo to + decide wether a terminal has colour, but uses it's own configuration + file. Needless to say, "rxvt-unicode" is not in it's default file (among + with most other terminals supporting colour). Either add: + + TERM rxvt-unicode + + to "/etc/DIR_COLORS" or simply add: + + alias ls='ls --color=auto' + + to your ".profile" or ".bashrc". + + Why doesn't vim/emacs etc. use the 88 colour mode? + See next entry. + + Why doesn't vim/emacs etc. make use of italic? + See next entry. + + Why are the secondary screen-related options not working properly? + Make sure you are using "TERM=rxvt-unicode". Some pre-packaged + distributions (most notably Debian GNU/Linux) break rxvt-unicode by + setting "TERM" to "rxvt", which doesn't have these extra features. + Unfortunately, some of these (most notably, again, Debian GNU/Linux) + furthermore fail to even install the "rxvt-unicode" terminfo file, so + you will need to install it on your own (See the question When I log-in + to another system it tells me about missing terminfo data? on how to do + this). + + Encoding / Locale / Input Method Issues + Rxvt-unicode does not seem to understand the selected encoding? + See next entry. + + Unicode does not seem to work? + If you encounter strange problems like typing an accented character but + getting two unrelated other characters or similar, or if program output + is subtly garbled, then you should check your locale settings. + + Rxvt-unicode must be started with the same "LC_CTYPE" setting as the + programs. Often rxvt-unicode is started in the "C" locale, while the + login script running within the rxvt-unicode window changes the locale + to something else, e.g. "en_GB.UTF-8". Needless to say, this is not + going to work. + + The best thing is to fix your startup environment, as you will likely + run into other problems. If nothing works you can try this in your + .profile. + + printf '\e]701;%s\007' "$LC_CTYPE" + + If this doesn't work, then maybe you use a "LC_CTYPE" specification not + supported on your systems. Some systems have a "locale" command which + displays this (also, "perl -e0" can be used to check locale settings, as + it will complain loudly if it cannot set the locale). If it displays + something like: + + locale: Cannot set LC_CTYPE to default locale: ... + + Then the locale you specified is not supported on your system. + + If nothing works and you are sure that everything is set correctly then + you will need to remember a little known fact: Some programs just don't + support locales :( + + How does rxvt-unicode determine the encoding to use? + See next entry. + + Is there an option to switch encodings? + Unlike some other terminals, rxvt-unicode has no encoding switch, and no + specific "utf-8" mode, such as xterm. In fact, it doesn't even know + about UTF-8 or any other encodings with respect to terminal I/O. + + The reasons is that there exists a perfectly fine mechanism for + selecting the encoding, doing I/O and (most important) communicating + this to all applications so everybody agrees on character properties + such as width and code number. This mechanism is the *locale*. + Applications not using that info will have problems (for example, + "xterm" gets the width of characters wrong as it uses it's own, + locale-independent table under all locales). + + Rxvt-unicode uses the "LC_CTYPE" locale category to select encoding. All + programs doing the same (that is, most) will automatically agree in the + interpretation of characters. + + Unfortunately, there is no system-independent way to select locales, nor + is there a standard on how locale specifiers will look like. + + On most systems, the content of the "LC_CTYPE" environment variable + contains an arbitrary string which corresponds to an already-installed + locale. Common names for locales are "en_US.UTF-8", "de_DE.ISO-8859-15", + "ja_JP.EUC-JP", i.e. "language_country.encoding", but other forms (i.e. + "de" or "german") are also common. + + Rxvt-unicode ignores all other locale categories, and except for the + encoding, ignores country or language-specific settings, i.e. + "de_DE.UTF-8" and "ja_JP.UTF-8" are the normally same to rxvt-unicode. + + If you want to use a specific encoding you have to make sure you start + rxvt-unicode with the correct "LC_CTYPE" category. + + Can I switch locales at runtime? + Yes, using an escape sequence. Try something like this, which sets + rxvt-unicode's idea of "LC_CTYPE". + + printf '\e]701;%s\007' ja_JP.SJIS + + See also the previous answer. + + Sometimes this capability is rather handy when you want to work in one + locale (e.g. "de_DE.UTF-8") but some programs don't support it (e.g. + UTF-8). For example, I use this script to start "xjdic", which first + switches to a locale supported by xjdic and back later: + + printf '\e]701;%s\007' ja_JP.SJIS + xjdic -js + printf '\e]701;%s\007' de_DE.UTF-8 + + You can also use xterm's "luit" program, which usually works fine, + except for some locales where character width differs between program- + and rxvt-unicode-locales. + + My input method wants but I want UTF-8, what can I do? + You can specify separate locales for the input method and the rest of + the terminal, using the resource "imlocale": + + URxvt.imlocale: ja_JP.EUC-JP + + Now you can start your terminal with "LC_CTYPE=ja_JP.UTF-8" and still + use your input method. Please note, however, that you will not be able + to input characters outside "EUC-JP" in a normal way then, as your input + method limits you. + + Rxvt-unicode crashes when the X Input Method changes or exits. + Unfortunately, this is unavoidable, as the XIM protocol is racy by + design. Applications can avoid some crashes at the expense of memory + leaks, and Input Methods can avoid some crashes by careful ordering at + exit time. kinput2 (and derived input methods) generally succeeds, while + SCIM (or similar input methods) fails. In the end, however, crashes + cannot be completely avoided even if both sides cooperate. + + So the only workaround is not to kill your Input Method Servers. + + Operating Systems / Package Maintaining + I am using Debian GNU/Linux and have a problem... + The Debian GNU/Linux package of rxvt-unicode in sarge contains large + patches that considerably change the behaviour of rxvt-unicode (but + unfortunately this notice has been removed). Before reporting a bug to + the original rxvt-unicode author please download and install the genuine + version () and try to reproduce + the problem. If you cannot, chances are that the problems are specific + to Debian GNU/Linux, in which case it should be reported via the Debian + Bug Tracking System (use "reportbug" to report the bug). + + For other problems that also affect the Debian package, you can and + probably should use the Debian BTS, too, because, after all, it's also a + bug in the Debian version and it serves as a reminder for other users + that might encounter the same issue. + + I am maintaining rxvt-unicode for distribution/OS XXX, any recommendation? + You should build one binary with the default options. configure now + enables most useful options, and the trend goes to making them + runtime-switchable, too, so there is usually no drawback to enbaling + them, except higher disk and possibly memory usage. The perl interpreter + should be enabled, as important functionality (menus, selection, likely + more in the future) depends on it. + + You should not overwrite the "perl-ext-common" snd "perl-ext" resources + system-wide (except maybe with "defaults"). This will result in useful + behaviour. If your distribution aims at low memory, add an empty + "perl-ext-common" resource to the app-defaults file. This will keep the + perl interpreter disabled until the user enables it. + + If you can/want build more binaries, I recommend building a minimal one + with "--disable-everything" (very useful) and a maximal one with + "--enable-everything" (less useful, it will be very big due to a lot of + encodings built-in that increase download times and are rarely used). + + I need to make it setuid/setgid to support utmp/ptys on my OS, is this safe? + It should be, starting with release 7.1. You are encouraged to properly + install urxvt with privileges necessary for your OS now. + + When rxvt-unicode detects that it runs setuid or setgid, it will fork + into a helper process for privileged operations (pty handling on some + systems, utmp/wtmp/lastlog handling on others) and drop privileges + immediately. This is much safer than most other terminals that keep + privileges while running (but is more relevant to urxvt, as it contains + things as perl interpreters, which might be "helpful" to attackers). + + This forking is done as the very first within main(), which is very + early and reduces possible bugs to initialisation code run before + main(), or things like the dynamic loader of your system, which should + result in very little risk. + + On Solaris 9, many line-drawing characters are too wide. + Seems to be a known bug, read + . Some people use the + following ugly workaround to get non-double-wide-characters working: + + #define wcwidth(x) wcwidth(x) > 1 ? 1 : wcwidth(x) + + I am on FreeBSD and rxvt-unicode does not seem to work at all. + Rxvt-unicode requires the symbol "__STDC_ISO_10646__" to be defined in + your compile environment, or an implementation that implements it, + wether it defines the symbol or not. "__STDC_ISO_10646__" requires that + wchar_t is represented as unicode. + + As you might have guessed, FreeBSD does neither define this symobl nor + does it support it. Instead, it uses it's own internal representation of + wchar_t. This is, of course, completely fine with respect to standards. + + However, that means rxvt-unicode only works in "POSIX", "ISO-8859-1" and + "UTF-8" locales under FreeBSD (which all use Unicode as wchar_t. + + "__STDC_ISO_10646__" is the only sane way to support multi-language apps + in an OS, as using a locale-dependent (and non-standardized) + representation of wchar_t makes it impossible to convert between wchar_t + (as used by X11 and your applications) and any other encoding without + implementing OS-specific-wrappers for each and every locale. There + simply are no APIs to convert wchar_t into anything except the current + locale encoding. + + Some applications (such as the formidable mlterm) work around this by + carrying their own replacement functions for character set handling with + them, and either implementing OS-dependent hacks or doing multiple + conversions (which is slow and unreliable in case the OS implements + encodings slightly different than the terminal emulator). + + The rxvt-unicode author insists that the right way to fix this is in the + system libraries once and for all, instead of forcing every app to carry + complete replacements for them :) + + I use Solaris 9 and it doesn't compile/work/etc. + Try the diff in doc/solaris9.patch as a base. It fixes the worst + problems with "wcwidth" and a compile problem. + + How can I use rxvt-unicode under cygwin? + rxvt-unicode should compile and run out of the box on cygwin, using the + X11 libraries that come with cygwin. libW11 emulation is no longer + supported (and makes no sense, either, as it only supported a single + font). I recommend starting the X-server in "-multiwindow" or + "-rootless" mode instead, which will result in similar look&feel as the + old libW11 emulation. + + At the time of this writing, cygwin didn't seem to support any + multi-byte encodings (you might try "LC_CTYPE=C-UTF-8"), so you are + likely limited to 8-bit encodings. RXVT TECHNICAL REFERENCE DESCRIPTION