st

Mahdi's build of st
git clone git://mahdi.pw/st.git
Log | Files | Refs | README | LICENSE

FAQ (9923B)


      1 ## Why does st not handle utmp entries?
      2 
      3 Use the excellent tool of [utmp](https://git.suckless.org/utmp/) for this task.
      4 
      5 
      6 ## Some _random program_ complains that st is unknown/not recognised/unsupported/whatever!
      7 
      8 It means that st doesn’t have any terminfo entry on your system. Chances are
      9 you did not `make install`. If you just want to test it without installing it,
     10 you can manually run `tic -sx st.info`.
     11 
     12 
     13 ## Nothing works, and nothing is said about an unknown terminal!
     14 
     15 * Some programs just assume they’re running in xterm i.e. they don’t rely on
     16   terminfo. What you see is the current state of the “xterm compliance”.
     17 * Some programs don’t complain about the lacking st description and default to
     18   another terminal. In that case see the question about terminfo.
     19 
     20 
     21 ## How do I scroll back up?
     22 
     23 * Using a terminal multiplexer.
     24 	* `st -e tmux` using C-b [
     25 	* `st -e screen` using C-a ESC
     26 * Using the excellent tool of [scroll](https://git.suckless.org/scroll/).
     27 * Using the scrollback [patch](https://st.suckless.org/patches/scrollback/).
     28 
     29 
     30 ## I would like to have utmp and/or scroll functionality by default
     31 
     32 You can add the absolute path of both programs in your config.h file. You only
     33 have to modify the value of utmp and scroll variables.
     34 
     35 
     36 ## Why doesn't the Del key work in some programs?
     37 
     38 Taken from the terminfo manpage:
     39 
     40 	If the terminal has a keypad that transmits codes when the keys
     41 	are pressed, this information can be given. Note that it is not
     42 	possible to handle terminals where the keypad only works in
     43 	local (this applies, for example, to the unshifted HP 2621 keys).
     44 	If the keypad can be set to transmit or not transmit, give these
     45 	codes as smkx and rmkx. Otherwise the keypad is assumed to
     46 	always transmit.
     47 
     48 In the st case smkx=E[?1hE= and rmkx=E[?1lE>, so it is mandatory that
     49 applications which want to test against keypad keys send these
     50 sequences.
     51 
     52 But buggy applications (like bash and irssi, for example) don't do this. A fast
     53 solution for them is to use the following command:
     54 
     55 	$ printf '\033[?1h\033=' >/dev/tty
     56 
     57 or
     58 	$ tput smkx
     59 
     60 In the case of bash, readline is used. Readline has a different note in its
     61 manpage about this issue:
     62 
     63 	enable-keypad (Off)
     64 		When set to On, readline will try to enable the
     65 		application keypad when it is called. Some systems
     66 		need this to enable arrow keys.
     67 
     68 Adding this option to your .inputrc will fix the keypad problem for all
     69 applications using readline.
     70 
     71 If you are using zsh, then read the zsh FAQ
     72 <http://zsh.sourceforge.net/FAQ/zshfaq03.html#l25>:
     73 
     74 	It should be noted that the O / [ confusion can occur with other keys
     75 	such as Home and End. Some systems let you query the key sequences
     76 	sent by these keys from the system's terminal database, terminfo.
     77 	Unfortunately, the key sequences given there typically apply to the
     78 	mode that is not the one zsh uses by default (it's the "application"
     79 	mode rather than the "raw" mode). Explaining the use of terminfo is
     80 	outside of the scope of this FAQ, but if you wish to use the key
     81 	sequences given there you can tell the line editor to turn on
     82 	"application" mode when it starts and turn it off when it stops:
     83 
     84 		function zle-line-init () { echoti smkx }
     85 		function zle-line-finish () { echoti rmkx }
     86 		zle -N zle-line-init
     87 		zle -N zle-line-finish
     88 
     89 Putting these lines into your .zshrc will fix the problems.
     90 
     91 
     92 ## How can I use meta in 8bit mode?
     93 
     94 St supports meta in 8bit mode, but the default terminfo entry doesn't
     95 use this capability. If you want it, you have to use the 'st-meta' value
     96 in TERM.
     97 
     98 
     99 ## I cannot compile st in OpenBSD
    100 
    101 OpenBSD lacks librt, despite it being mandatory in POSIX
    102 <http://pubs.opengroup.org/onlinepubs/9699919799/utilities/c99.html#tag_20_11_13>.
    103 If you want to compile st for OpenBSD you have to remove -lrt from config.mk, and
    104 st will compile without any loss of functionality, because all the functions are
    105 included in libc on this platform.
    106 
    107 
    108 ## The Backspace Case
    109 
    110 St is emulating the Linux way of handling backspace being delete and delete being
    111 backspace.
    112 
    113 This is an issue that was discussed in suckless mailing list
    114 <https://lists.suckless.org/dev/1404/20697.html>. Here is why some old grumpy
    115 terminal users wants its backspace to be how he feels it:
    116 
    117 	Well, I am going to comment why I want to change the behaviour
    118 	of this key. When ASCII was defined in 1968, communication
    119 	with computers was done using punched cards, or hardcopy
    120 	terminals (basically a typewriter machine connected with the
    121 	computer using a serial port).  ASCII defines DELETE as 7F,
    122 	because, in punched-card terms, it means all the holes of the
    123 	card punched; it is thus a kind of 'physical delete'. In the
    124 	same way, the BACKSPACE key was a non-destructive backspace,
    125 	as on a typewriter.  So, if you wanted to delete a character,
    126 	you had to BACKSPACE and then DELETE.  Another use of BACKSPACE
    127 	was to type accented characters, for example 'a BACKSPACE `'.
    128 	The VT100 had no BACKSPACE key; it was generated using the
    129 	CONTROL key as another control character (CONTROL key sets to
    130 	0 b7 b6 b5, so it converts H (code 0x48) into BACKSPACE (code
    131 	0x08)), but it had a DELETE key in a similar position where
    132 	the BACKSPACE key is located today on common PC keyboards.
    133 	All the terminal emulators emulated the difference between
    134 	these keys correctly: the backspace key generated a BACKSPACE
    135 	(^H) and delete key generated a DELETE (^?).
    136 
    137 	But a problem arose when Linus Torvalds wrote Linux. Unlike
    138 	earlier terminals, the Linux virtual terminal (the terminal
    139 	emulator integrated in the kernel) returned a DELETE when
    140 	backspace was pressed, due to the VT100 having a DELETE key in
    141 	the same position.  This created a lot of problems (see [1]
    142 	and [2]). Since Linux has become the king, a lot of terminal
    143 	emulators today generate a DELETE when the backspace key is
    144 	pressed in order to avoid problems with Linux. The result is
    145 	that the only way of generating a BACKSPACE on these systems
    146 	is by using CONTROL + H. (I also think that emacs had an
    147 	important point here because the CONTROL + H prefix is used
    148 	in emacs in some commands (help commands).)
    149 
    150 	From point of view of the kernel, you can change the key
    151 	for deleting a previous character with stty erase. When you
    152 	connect a real terminal into a machine you describe the type
    153 	of terminal, so getty configures the correct value of stty
    154 	erase for this terminal. In the case of terminal emulators,
    155 	however, you don't have any getty that can set the correct
    156 	value of stty erase, so you always get the default value.
    157 	For this reason, it is necessary to add 'stty erase ^H' to your
    158 	profile if you have changed the value of the backspace key.
    159 	Of course, another solution is for st itself to modify the
    160 	value of stty erase.  I usually have the inverse problem:
    161 	when I connect to non-Unix machines, I have to press CONTROL +
    162 	h to get a BACKSPACE. The inverse problem occurs when a user
    163 	connects to my Unix machines from a different system with a
    164 	correct backspace key.
    165 
    166 	[1] http://www.ibb.net/~anne/keyboard.html
    167 	[2] http://www.tldp.org/HOWTO/Keyboard-and-Console-HOWTO-5.html
    168 
    169 
    170 ## But I really want the old grumpy behaviour of my terminal
    171 
    172 Apply [1].
    173 
    174 [1] https://st.suckless.org/patches/delkey
    175 
    176 
    177 ## Why do images not work in st using the w3m image hack?
    178 
    179 w3mimg uses a hack that draws an image on top of the terminal emulator Drawable
    180 window. The hack relies on the terminal to use a single buffer to draw its
    181 contents directly.
    182 
    183 st uses double-buffered drawing so the image is quickly replaced and may show a
    184 short flicker effect.
    185 
    186 Below is a patch example to change st double-buffering to a single Drawable
    187 buffer.
    188 
    189 diff --git a/x.c b/x.c
    190 --- a/x.c
    191 +++ b/x.c
    192 @@ -732,10 +732,6 @@ xresize(int col, int row)
    193  	win.tw = col * win.cw;
    194  	win.th = row * win.ch;
    195  
    196 -	XFreePixmap(xw.dpy, xw.buf);
    197 -	xw.buf = XCreatePixmap(xw.dpy, xw.win, win.w, win.h,
    198 -			DefaultDepth(xw.dpy, xw.scr));
    199 -	XftDrawChange(xw.draw, xw.buf);
    200  	xclear(0, 0, win.w, win.h);
    201  
    202  	/* resize to new width */
    203 @@ -1148,8 +1144,7 @@ xinit(int cols, int rows)
    204  	gcvalues.graphics_exposures = False;
    205  	dc.gc = XCreateGC(xw.dpy, parent, GCGraphicsExposures,
    206  			&gcvalues);
    207 -	xw.buf = XCreatePixmap(xw.dpy, xw.win, win.w, win.h,
    208 -			DefaultDepth(xw.dpy, xw.scr));
    209 +	xw.buf = xw.win;
    210  	XSetForeground(xw.dpy, dc.gc, dc.col[defaultbg].pixel);
    211  	XFillRectangle(xw.dpy, xw.buf, dc.gc, 0, 0, win.w, win.h);
    212  
    213 @@ -1632,8 +1627,6 @@ xdrawline(Line line, int x1, int y1, int x2)
    214  void
    215  xfinishdraw(void)
    216  {
    217 -	XCopyArea(xw.dpy, xw.buf, xw.win, dc.gc, 0, 0, win.w,
    218 -			win.h, 0, 0);
    219  	XSetForeground(xw.dpy, dc.gc,
    220  			dc.col[IS_SET(MODE_REVERSE)?
    221  				defaultfg : defaultbg].pixel);
    222 
    223 
    224 ## BadLength X error in Xft when trying to render emoji
    225 
    226 Xft makes st crash when rendering color emojis with the following error:
    227 
    228 "X Error of failed request:  BadLength (poly request too large or internal Xlib length error)"
    229   Major opcode of failed request:  139 (RENDER)
    230   Minor opcode of failed request:  20 (RenderAddGlyphs)
    231   Serial number of failed request: 1595
    232   Current serial number in output stream:  1818"
    233 
    234 This is a known bug in Xft (not st) which happens on some platforms and
    235 combination of particular fonts and fontconfig settings.
    236 
    237 See also:
    238 https://gitlab.freedesktop.org/xorg/lib/libxft/issues/6
    239 https://bugs.freedesktop.org/show_bug.cgi?id=107534
    240 https://bugzilla.redhat.com/show_bug.cgi?id=1498269
    241 
    242 The solution is to remove color emoji fonts or disable this in the fontconfig
    243 XML configuration.  As an ugly workaround (which may work only on newer
    244 fontconfig versions (FC_COLOR)), the following code can be used to mask color
    245 fonts:
    246 
    247 	FcPatternAddBool(fcpattern, FC_COLOR, FcFalse);
    248 
    249 Please don't bother reporting this bug to st, but notify the upstream Xft
    250 developers about fixing this bug.
    251 
    252 As of 2022-09-05 this now seems to be finally fixed in libXft 2.3.5:
    253 https://gitlab.freedesktop.org/xorg/lib/libxft/-/blob/libXft-2.3.5/NEWS