1 -*- coding: utf-8; mode: text; -*-
5 This branch implements support for opening multiple, different tty
6 devices and simultaneous X and tty frames from a single Emacs session.
9 Emacs is notoriously slow at startup, so most people use another
10 editor or emacsclient for quick editing jobs from the console.
11 Unfortunately, emacsclient was very awkward to use, because it did not
12 support opening a new Emacs frame on the current virtual tty.
13 Now, with multi-tty support, it can do that. (Emacsclient starts up
16 Some Gnus users (including me) run Gnus in an X frame in its own Emacs
17 instance, which they typically leave running for weeks. It would be
18 nice if they could connect to this instance from a remote ssh session
19 and check their messages without opening a remote X frame or resorting
25 I'm Károly Lőrentey. My address: lorentey@elte.hu.
27 Comments, bug reports, suggestions and patches are welcome; send them
28 to multi-tty@lists.fnord.hu.
30 The following is a (sadly incomplete) list of people who have
31 contributed to the project by testing, submitting patches, bug
32 reports, and suggestions. Thanks!
34 Bernard Adrian <bernadrian@free.fr>
35 ARISAWA Akihiro <ari@mbf.ocn.ne.jp>
36 Vincent Bernat <bernat@luffy.cx>
37 Han Boetes <han@mijncomputer.nl>
38 Francisco Borges <borges@let.rug.nl>
39 Damien Cassou <damien.cassou@laposte.net>
40 Robert J. Chassell <bob@rattlesnake.com>
41 Romain Francoise <romain@orebokech.com>
42 Ami Fischman <ami@fischman.org>
43 Friedrich Delgado Friedrichs <friedel@nomaden.org>
44 IRIE Tetsuya <irie@t.email.ne.jp>
45 Yoshiaki Kasahara <kasahara@nc.kyushu-u.ac.jp>
46 Bas Kok <nekkobassu@yahoo.com>
47 Jurej Kubelka <Juraj.Kubelka@email.cz>
48 David Lichteblau <david@lichteblau.com>
49 Richard Lewis <rtf@jabble.com>
50 mace <mace@kirjakaapeli.lib.hel.fi>
51 Suresh Madhu <madhu@cs.unm.edu>
52 Xavier Mallard <zedek@gnu-rox.org>
53 Istvan Marko <mi-mtty@kismala.com>
54 Ted Morse <morse@ciholas.com>
55 Gergely Nagy <algernon@debian.org>
56 Dan Nicolaescu <dann@ics.uci.edu>
57 Kalle Olavi Niemitalo <kon@iki.fi>
58 Mark Plaksin <happy@mcplaksin.org>
59 Frank Ruell <stoerte@dreamwarrior.net>
60 Tom Schutzer-Weissmann <trmsw@yahoo.co.uk>
61 Joakim Verona <joakim@verona.se>
62 Dan Waber <dwaber@logolalia.com>
65 Richard Stallman was kind enough to review an earlier version of my
72 The multi-tty mailing list (discussion & bug reports):
74 Address: multi-tty@lists.fnord.hu
75 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty
76 Archive: http://news.gmane.org/gmane.emacs.multi-tty/
78 Commit notifications (read-only):
80 Address: multi-tty-commits@lists.fnord.hu
81 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty-commits
87 The branch is now very stable and almost full-featured. All of the
88 major problems have been fixed, only a few minor issues remain. (It
89 still needs to be ported to Windows/Mac/DOS, though.) Both multiple
90 tty device support and simultaneous X and tty frame support works
91 fine. Emacsclient has been extended to support opening new tty and X
92 frames. It has been changed to open new Emacs frames by default.
94 The multi-tty branch has been scheduled for inclusion in the next
95 major release of Emacs (version 23). I expect the merge into the
96 development trunk to occur sometime during next year (2006), after the
97 merge of the Unicode branch.
99 Tested on GNU/Linux, Solaris 8, FreeBSD and OpenBSD. Please let me
100 know if you succeed or fail to use it on other platforms---I'll have a
101 few tricky test cases for you.
105 * GTK support. If you compile your Emacs with the GTK
106 toolkit, some functionality of multi-tty will be lost. In
107 particular, you will not be able to work on multiple X
108 displays at once. Current releases of GTK have limitations
109 and bugs that prevent full-blown multi-display support in
110 Emacs. (GTK crashes when Emacs tries to disconnect from an
111 X server.) Use the Lucid toolkit if you want to see a
112 complete feature set.
114 * The single-kboard mode.
116 If your multi-tty Emacs session seems to be frozen, you
117 probably have a recursive editing session or a pending
118 minibuffer prompt (which is a kind of recursive editing) on
119 another display. To unfreeze your session, switch to that
120 display and complete the recursive edit, for example by
121 pressing C-] (`abort-recursive-edit').
123 I am sorry to say that currently there is no way to break
124 out of this "single-kboard mode" from a frozen display. If
125 you are unable to switch to the display that locks the
126 others (for example because it is on a remote computer),
127 then you can use emacsclient to break out of all recursive
130 emacsclient -e '(top-level)'
132 Note that this (perhaps) unintuitive behaviour is by design.
133 Single-kboard mode is required because of an intrinsic Emacs
134 limitation that is very hard to eliminate. (This limitation
135 is related to the single-threaded nature of Emacs.)
137 I plan to implement better user notification and support for
138 breaking out of single-kboard mode from locked displays.
140 * Mac, Windows and DOS support is broken, doesn't even
141 compile. Multiple display support will probably not provide
142 new Emacs features on these systems, but the multi-tty
143 branch changed a few low-level interfaces, and the
144 system-dependent source files need to be adapted
145 accordingly. The changes are mostly trivial, so almost
146 anyone can help, if only by compiling the branch and
147 reporting the compiler errors.
150 HOW TO GET THE BRANCH
151 ---------------------
153 The branch uses Bazaar 1 (http://bazaar.canonical.com) for version control.
155 Retrieving the latest version of the branch:
157 baz register-archive -f http://aszt.inf.elte.hu/~lorentey/mirror/arch/2004
158 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
160 This incantation uses an archive mirror that is hosted on a
161 high-bandwidth site. Please note that on average there is a two-hour
162 delay for commits to arrive on this mirror. My primary mirror is on the
163 low-bandwidth http://lorentey.hu/ site:
165 baz register-archive -f http://lorentey.hu/arch/2004/
166 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
168 This is "instantly" updated, but very slow from outside Hungary.
169 (By "instantly" I mean as soon as I connect the notebook I work on to
170 a network. It could take days.)
172 The Arch supermirror provides mirroring services for all public Arch
173 repositories. We have a mirror there, too, if you prefer.
175 baz register-archive -f http://mirrors.sourcecontrol.net/lorentey%40elte.hu--2004
176 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
178 My GPG key id is 0FB27A3F; it is available from
179 hkp://wwwkeys.eu.pgp.net/, or from my homepage at
180 http://lorentey.hu/rolam/gpg.html)
182 Don't worry if the above checkout takes a few minutes to complete;
183 once you have a source tree, updating it to the latest revision will
184 be _much_ faster. Use the following command for the update:
188 You can find more information about Bazaar on
189 http://bazaar.canonical.com/. It's a distributed source control
190 system that is somewhat less broken than competing projects.
192 If you don't have Bazaar, the branch has a homepage from which you can
193 download conventional patches against Emacs CVS HEAD:
195 http://lorentey.hu/project/emacs.html
197 I suggest you use Bazaar whenever feasible.
203 If you run Debian, or a distribution based on Debian, you are welcome
204 to use our binary packages; put these lines in your /etc/apt/sources.list:
207 deb http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
208 deb-src http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
210 Note that these packages are intended solely to provide an easy way to
211 test the new multi-tty features. They are not to be taken as Emacs
212 releases, and it's a mistake to expect robust operation or any kind of
213 timely support for them. Do not install them, unless you'd like to
214 have your editor crash on you.
220 The multi-tty branch is compiled the same way as Emacs itself:
222 make maintainer-clean # (If you have compiled Emacs before)
224 ./configure --without-gtk <your favourite options>
228 If you have strange compilation errors, they may be caused by old
229 *.elc files that are left over from an earlier bootstrap. The `make
230 maintainer-clean' target deletes them, so it is a good idea to run
231 that before reporting a bug. (Emacs requires a clean recompilation
232 after certain kinds of source code changes.)
237 To test the multi-tty branch, start up the Emacs server with the
243 and then (from a shell prompt on another terminal) start emacsclient
245 emacsclient -t /optional/file/names... (for a tty frame)
246 emacsclient /optional/file/names... (for an X frame)
248 (Make sure both emacs and emacsclient are multi-tty versions.)
249 You'll hopefully have two fully working, independent frames on
250 separate terminals. The new frame is closed automatically when you
251 finish editing the specified files (C-x #), but delete-frame (C-x 5 0)
252 also works. Of course, you can create frames on more than two tty
255 Creating new frames on the same tty with C-x 5 2 (make-frame-command)
256 works, and behaves the same way as in previous Emacs versions. If you
257 exit emacs, all terminals should be restored to their previous states.
259 This is work in progress, and probably full of bugs. It is a good
260 idea to run emacs from gdb, so that you'll have a live instance to
261 debug if something goes wrong. Please send me your bug reports on our
262 mailing list: multi-tty@lists.fnord.hu
267 I think the best way to use the new Emacs is to have it running inside
268 a disconnected GNU screen session, and always use emacsclient for
269 normal work. One advantage of this is that not a single keystroke of
270 your work will be lost if the display device that you are using
271 crashes, or the network connection times out, or whatever. (I had an
272 extremely unstable X server for some time while I was developing these
273 patches, and running Emacs this way has saved me a number of M-x
274 recover-session invocations.)
276 I use the following two bash scripts to handle my Emacs sessions:
278 -------------------------------------------------------connect-emacs--
280 # Usage: connect-emacs <name> <args>...
282 # Connects to the Emacs instance called NAME. Starts up the instance
283 # if it is not already running. The rest of the arguments are passed
289 if [ -z "$name" ]; then
290 echo "Usage: connect_emacs <name> <args>..." >&2
293 preload-emacs "$name" wait
294 /usr/bin/emacsclient.emacs-multi-tty -s "$name" "$@"
295 ----------------------------------------------------------------------
297 -------------------------------------------------------preload-emacs--
299 # Usage: preload-emacs <name> [<waitp>]
301 # Preloads the Emacs instance called NAME in a detached screen
302 # session. Does nothing if the instance is already running. If WAITP
303 # is non-empty, the function waits until the server starts up and
304 # creates its socket; otherwise it returns immediately.
308 screendir="/var/run/screen/S-$USER"
309 serverdir="/tmp/emacs$UID"
310 emacs=/usr/bin/emacs-multi-tty # Or wherever you installed your multi-tty Emacs
312 if [ -z "$name" ]; then
313 echo "Usage: preload_emacs <name> [<waitp>]" >&2
317 if [ ! -e "$screendir"/*."$name" ]; then
318 if [ -e "$serverdir/$name" ]; then
319 # Delete leftover socket (for the wait option)
320 rm "$serverdir/$name"
322 screen -dmS "$name" "$emacs" -nw --eval "(setq server-name \"$name\")" -f server-start
324 if [ ! -z "$waitp" ]; then
325 while [ ! -e "$serverdir/$name" ]; do sleep 0.1; done
327 ----------------------------------------------------------------------
329 I have the following in my profile to have two instances automatically
330 preloaded for editing and email:
335 It is useful to set up short aliases for connect-emacs. I use the
338 alias edit="connect-emacs editor"
340 alias et="connect-emacs editor -t"
341 alias gnus="connect-emacs gnus"
353 For the NEWS file: (Needs much, much work)
355 ** Support for multiple terminal devices and simultaneous graphical
356 and tty frames has been added. You can test for the presence of
357 this feature in your Lisp code by testing for the `multi-tty'
360 *** The `window-system' variable has been made frame-local. The new
361 `initial-window-system' variable contains the `window-system'
362 value for the first frame.
364 *** You can specify a terminal device (`tty' parameter) and a terminal
365 type (`tty-type' parameter) to `make-terminal-frame'.
367 *** The new function `make-frame-on-tty' allows you to create a new
368 frame on another tty device interactively.
370 *** The function `make-frame-on-display' now works during a tty
371 session, and `make-frame-on-tty' works during a graphical session.
373 *** Emacsclient has been extended to support opening a new terminal
374 frame. Its behaviour has been changed to open a new Emacs frame by
375 default. Use the -c option to get the old behavior of opening
376 files in the currently selected Emacs frame.
378 *** C-z now invokes `suspend-frame', C-x C-c now invokes
379 `save-buffers-kill-terminal'.
381 *** New functions: frame-tty-name, frame-tty-type, delete-tty,
382 suspend-tty, resume-tty, terminal-id, terminal-parameters,
383 terminal-parameter, set-terminal-parameter,
384 modify-terminal-parameters, environment, let-environment
386 *** New variables: local-key-translation-map, local-function-key-map
388 *** The `keyboard-translate-table' variable and the terminal and
389 keyboard coding systems have been made terminal-local.
391 *** In addition to the global key-translation-map and
392 function-key-map, Emacs has terminal-local
393 local-key-translation-map and local-function-key-map variables,
394 and uses them instead of the global keymaps to set up translations
395 and function key sequences relevant to a specific terminal device.
397 *** talk.el has been extended for multiple tty support.
401 (The rest of this file consists of my development notes and as such it
402 is probably not very interesting for anyone else.)
407 ** emacsclient -t on the console does not work after su. You have to
408 use non-root accounts or start as root to see this.
414 *ERROR*: Could not open file: /dev/tty1
416 The tty can be opened as /dev/tty by emacsclient, but not by Emacs.
417 This seems to be a serious problem. Currently my only idea is to
418 bring back the ugly pty proxy hack from the initial versions of
419 multi-tty. Suggestions would be appreciated.
421 Update: we could change emacsclient to pass its open file
422 descriptor to the Emacs process. Unfortunately, this requires a
423 new Lisp-level Emacs API, and as file descriptors are not otherwise
424 exported to Lisp, this approach seems at least as ugly as the pty
427 ** Understand how `quit_throw_to_read_char' works, and fix any bugs
430 ** See if getcjmp can be eliminated somehow. Why does Emacs allow
431 asynchronous input processing while it's reading input anyway?
433 ** `delete-frame' events are handled by `special-event-map'
434 immediately when read by `read_char'. This is fine but it prevents
435 higher-level keymaps from binding that event to get notified of the
438 Sometimes it would be useful for Lisp code to be notified of frame
439 deletions after they have happened, usually because they want to
440 clean up after the deleted frame. Not all frame-local states can
441 be stored as a frame parameter. For example,
442 `display-splash-screen' uses `recursive-edit' with a special keymap
443 override to create its buffer---and it leads to all kinds of
444 nastiness if Emacs stays in this recursive edit mode after the
445 frame containing the splash screen is deleted. Basically, the
446 splash-screen implementation wants to throw out of the recursive
447 edit when the frame is deleted; however, it is not legal to throw
448 from `delete-frame-functions' because `delete-frame' must not fail.
449 (Introducing `delete-frame-after-functions' would not help either
450 because `delete-frame' may not fail at that time either.)
452 Currently `fancy-splash-screens' installs a
453 `delete-frame-functions' hook that sets up a timer to exit the
454 recursive edit. This is an adequate solution, but it would perhaps
455 be better to have something like a `frame-deleted' event that could
456 be bound in the normal way.
458 ** Trouble: `setenv' doesn't actually set environment variables in the
459 Emacs process. This defeats the purpose of the elaborate
460 `server-with-environment' magic around the `tgetent' call in
463 ** (Possibly) create hooks in struct device for creating frames on a
464 specific terminal, and eliminate the hackish terminal-related frame
465 parameters (display, tty, tty-type).
470 ** Decide whether to keep the C implementation of terminal parameters,
471 or revert to the previous, purely Lisp code. It turned out that
472 local environments do not need terminal parameters after all.
474 ** Move Fsend_string_to_terminal to term.c, and declare get_named_tty
475 as static, removing it from dispextern.h.
476 Move fatal to emacs.c and declare it somewhere.
478 ** Search for `suspend-emacs' references and replace them with
479 `suspend-frame', if necessary. Ditto for `save-buffers-kill-emacs'
480 vs. `save-buffers-kill-display'.
482 ** Emacs crashes when a tty frame is resized so that there is no space
483 for all its windows. (Tom Schutzer-Weissmann)
485 ** Report GTK multi-display problems to GTK maintainers. For extra
488 Currently you can not connect to new X displays when you compile
489 Emacs with GTK support. If you want to play around with GTK
490 multidisplay (and don't mind core dumps), you can edit src/config.h
491 and define HAVE_GTK_MULTIDISPLAY there by hand.
493 http://bugzilla.gnome.org/show_bug.cgi?id=85715
495 Update: Han reports that GTK+ version 2.8.9 almost gets display
496 disconnects right. GTK will probably be fully fixed by the time
497 multi-tty gets into the trunk.
499 Update: I am still having problems with GTK+ 2.8.10. I have the
500 impression that the various multidisplay fixes will only get
501 released in GTK+ 2.10.
503 ** Audit `face-valid-attribute-values' usage in customize and
504 elsewhere. Its return value depends on the current window system.
505 Replace static initializers using it with runtime functions. For
506 example, custom's buttons are broken on non-initial device types.
508 ** Possibly turn off the double C-g feature when there is an X frame.
509 C.f. (emacs)Emergency Escape.
511 ** frames-on-display-list should also accept frames.
513 ** Consider the `tty-type' frame parameter and the `display-tty-type'
514 function. They serve the exact same purpose. I think it may be
515 a good idea to eliminate one of them, preferably `tty-type'.
517 ** The handling of lisp/term/*.el, and frame creation in general, is a
518 big, big mess. How come the terminal-specific file is loaded by
519 tty-create-frame-with-faces? I don't think it is necessary to load
520 these files for each frame; once per terminal should be enough.
521 Update: lisp/term/*.el is not loaded repeatedly anymore, but
522 faces.el still needs to be cleaned up.
524 ** Fix frame-set-background-mode in this branch. It was recently
525 changed in CVS, and frame.el in multi-tty has not yet been adapted
526 for the changes. (It needs to look at
527 default-frame-background-mode.) (Update: maybe it is fixed now;
528 needs testing.) (Note that the byte compiler has this to say about
531 term/rxvt.el:309:17:Warning: assignment to free variable
532 `default-frame-background-mode'
534 ** I think `(set-)terminal-local-value' and the terminal parameter
535 mechanism should be integrated into a single framework.
537 (Update: `(set-)terminal-local-value' is now eliminated, but the
538 terminal-local variables should still be accessible as terminal
539 parameters. This also applies to `display-name' and similar
542 ** Add the following hooks: after-delete-frame-hook (for server.el,
543 instead of delete-frame-functions),
544 after-delete-terminal-functions, after-create-terminal-functions.
546 ** BULK RENAME: The `display-' prefix of new Lisp-level functions
547 conflicts with stuff like `display-time-mode'. Use `device-'
548 or `terminal-' instead. I think I prefer `terminal-'.
550 It turns out that most of the offending Lisp functions were defined
551 in the trunk. Therefore, compatibility aliases should be defined
552 for the following names:
554 display-color-cells terminal-color-cells
555 display-color-p terminal-color-p
556 display-graphic-p terminal-graphic-p
557 display-grayscale-p terminal-grayscale-p
558 display-images-p terminal-images-p
559 display-mm-height terminal-mm-height
560 display-mm-width terminal-mm-width
561 display-mouse-p terminal-mouse-p
562 display-multi-font-p terminal-multi-font-p
563 display-multi-frame-p terminal-multi-frame-p
564 display-pixel-height terminal-pixel-height
565 display-pixel-width terminal-pixel-width
566 display-pixels-per-inch terminal-pixels-per-inch
567 display-planes terminal-planes
568 display-popup-menus-p terminal-popup-menus-p
569 display-save-under terminal-save-under
570 display-screens terminal-screens
571 display-supports-face-attributes-p terminal-supports-face-attributes-p
572 display-visual-class terminal-visual-class
573 framep-on-display framep-on-terminal
574 frames-on-display-list frames-on-terminal-list
576 The following functions were introduced in the multi-tty branch, and
577 were renamed without aliases:
579 delete-display delete-terminal
580 display-controlling-tty-p controlling-tty-p
581 display-list terminal-list
582 display-live-p terminal-live-p
583 display-name terminal-name
584 display-tty-type tty-type
585 frame-display frame-terminal
586 selected-display selected-terminal
588 ** The single-keyboard mode of MULTI_KBOARD is extremely confusing
589 sometimes; Emacs does not respond to stimuli from other keyboards.
590 At least a beep or a message would be important, if the single-mode
591 is still required to prevent interference. (Reported by Dan
594 Update: selecting a region with the mouse enables single_kboard
595 under X. This is very confusing.
597 Update: After discussions with Richard Stallman, this will be
598 resolved by having locked displays warn the user to wait, and
599 introducing a complex protocol to remotely bail out of
600 single-kboard mode by pressing C-g.
602 Update: Warning the user is not trivial to implement, as Emacs has
603 only one echo area, shared by all frames. Ideally the warning
604 should not be displayed on the display that is locking the others.
605 Perhaps the high probability of user confusion caused by
606 single_kboard mode deserves a special case in the display code.
607 Alternatively, it might be good enough to signal single_kboard mode
608 by changing the modelines or some other frame-local display element
609 on the locked out displays.
611 Update: In fact struct kboard does have an echo_string slot.
613 ** The session management module is prone to crashes when the X
614 connection is closed and then later I try to connect to a new X
617 #0 0xb7ebc806 in SmcGetIceConnection () from /usr/X11R6/lib/libSM.so.6
618 #1 0x080e6641 in x_session_check_input (bufp=0xbf86c9c0) at xsmfns.c:144
619 #2 0x080d3bbc in XTread_socket (device=0xa722ff8, expected=1, hold_quit=0xbf86ca90) at xterm.c:7037
620 #3 0x080fa404 in read_avail_input (expected=1) at keyboard.c:6696
621 #4 0x080fa4ca in handle_async_input () at keyboard.c:6900
622 #5 0x080d51fa in x_term_init (display_name=162628899, xrm_option=0x0, resource_name=0x857068c "emacs") at xterm.c:10622
623 #6 0x080d920e in x_display_info_for_name (name=162628899) at xfns.c:3975
624 #7 0x080d92f9 in check_x_display_info (object=1) at xfns.c:274
625 #8 0x080d97b8 in Fx_create_frame (parms=151221485) at xfns.c:3016
626 #9 0x0815bf72 in Ffuncall (nargs=2, args=0xbf86ceec) at eval.c:2851
628 I installed a workaround to prevent this. The X session manager is
629 only contacted when the very first display in the Emacs session is
630 an X display. Also, x_delete_display() on this display aborts
631 session management, and XTread_socket only calls
632 x_session_check_input when it is called for the display that the
633 session was opened on. While this does not really fix the bug, it
634 makes it much less frequent, because session manager support will
635 not normally be enabled when Emacs can survive the shutdown of the
638 See if xsmfns.c should be updated.
640 ** Hunt down display-related functions in frame.el and extend them all
641 to accept display ids.
643 ** rif->flush_display_optional (NULL) calls should be replaced by a
646 ** The set-locale-environment hack (adding the DISPLAY option) should
647 be replaced with a clean design.
649 ** standard-display-table should be display-local.
650 standard-display-european should be display-local.
652 ** With iswitchb-default-method set to 'always-frame, only frames on
653 the current display should be considered. This might involve
654 extending `get-buffer-window'.
656 ** Have a look at Vlocale_coding_system. Seems like it would be a
657 tedious job to localize it, although most references use it for
658 interfacing with libc and are therefore OK with the global
661 Exceptions found so far: x-select-text and
662 x-cut-buffer-or-selection-value.
664 ** Have a look at fatal_error_hook.
666 ** Have a look at set_frame_matrix_frame.
668 ** Check if we got term-setup-hook right.
670 ** I think tip_frame should be display-local.
672 ** Check display reference count handling in x_create_tip_frame.
674 ** make-frame does not correctly handle extra parameters in its
677 (frame-parameter (make-frame (list (cons 'foobar 42))) 'foobar)
680 (This is likely an error in the CVS trunk.)
682 ** Dan Nicolaescu suggests that -nw should be added as an alias for -t
683 in emacsclient. Good idea. (Alas, implementing this is not
684 trivial, getopt_long does not seem to support two-letter ``short''
685 options. Patches are welcome.)
687 ** Mark Plaksin suggests that emacsclient should accept the same
688 X-related command-line arguments as Emacs. Most of the X-related
689 argument-handling is done in Lisp, so this should be quite easy to
692 ** Gergely Nagy suggests that C-x # should only kill the current
693 frame, not any other emacsclient frame that may have the same file
694 opened for editing. I think I agree with him.
696 ** Very strange bug: visible-bell does not work on secondary
697 terminals in xterm and konsole. The screen does flicker a bit,
698 but it's so quick it isn't noticable.
700 (Update: This is probably some problem with padding or whatnot on
701 the secondary terminals.)
703 ** Move baud_rate to struct display.
705 ** Implement support for starting an interactive Emacs session without
706 an initial frame. (The user would connect to it and open frames
707 later, with emacsclient.)
709 ** Fix Mac support (I can't do this entirely myself). Note that the
710 current state of Mac-specific source files in the multi-tty tree
711 are not useful; before starting work on Mac support, revert to
712 pristine, pre-multi-tty versions.
714 ** Fix W32 support (I can't do this entirely myself). Note that the
715 current state of W32-specific source files in the multi-tty tree
716 are not useful; before starting work on W32 support, revert to
717 pristine, pre-multi-tty versions.
719 ** Fix DOS support (I can't do this entirely myself). Note that the
720 current state of DOS-specific source files in the multi-tty tree
721 are not useful; before starting work on DOS support, revert to
722 pristine, pre-multi-tty versions.
724 ** Do a grep on XXX and ?? for more issues.
726 ** flow-ctrl.el must be updated.
728 ** Fix stuff_char for multi-tty. Doesn't seem to be of high priority.
733 (ex-TODO items with explanations.)
735 -- Introduce a new struct for terminal devices.
737 (Done, see struct tty_output. The list of members is not yet
740 -- Change the bootstrap procedure to initialize tty_list.
742 (Done, but needs review.)
744 -- Change make-terminal-frame to support specifying another tty.
746 (Done, new frame parameters: `tty' and `tty-type'.)
748 -- Implement support for reading from multiple terminals.
750 (Done, read_avail_input tries to read from each terminal, until one
751 succeeds. MULTI_KBOARD is not used. Secondary terminals don't send
754 (Update: They do, now.)
756 (Update2: After enabling X, they don't.)
758 -- other-frame should cycle through the frames on the `current'
761 (Done, by trivially modifiying next_frame and prev_frame.)
763 -- Support different terminal sizes.
767 -- Make sure terminal resizes are handled gracefully. (Could be
770 (Done. We don't get automatic SIGWINCH for additional ttys,
773 -- Extend emacsclient to automatically open a new tty when it connects
776 (Done. It's an ugly hack, needs more work.)
778 -- Redisplay must refresh the topmost frame on *all* terminals, not
779 just the initial terminal.
781 (Done, but introduced an ugly redisplay problems. Ugh.)
783 -- Fix redisplay problems.
785 (Done; it turned out that the entire Wcm structure must be moved
786 inside tty_output. Why didn't I catch this earlier?)
788 -- Provide a way for emacsclient to tell Emacs that the tty has been
791 (Done, simply forward the SIGWINCH signal.)
793 -- Each keypress should automatically select the frame corresponding
794 to the terminal that it was coming from. This means that Emacs
795 must know from which terminal the last keyboard event came from.
797 (Done, it was quite simple, the input event system already
798 supported multiple frames.)
800 -- Fix SIGIO issue with secondary terminals.
802 (Done, emacsclient signals Emacs after writing to the proxy pseudo
803 terminal. Note that this means that multi-tty does not work with
806 (Update: This is bullshit. There is a read_input_waiting function,
807 extend that somehow.)
809 (Update of update: The first update was not right either, extending
810 read_input_waiting was not necessary. Secondary ttys do seem to
811 send signals on input.)
813 (Update^3: Not any more.)
815 -- Make make-terminal-frame look up the `tty' and `tty-type' frame
816 parameters from the currently selected terminal before the global
821 -- Put all cached terminal escape sequences into struct tty_output.
822 Currently, they are still stored in global variables, so we don't
823 really support multiple terminal types.
825 (Done. It was not fun.)
827 -- Implement sane error handling after initialization. (Currently
828 emacs exits if you specify a bad terminal type.) The helpful error
829 messages must still be provided when Emacs starts.
833 -- Implement terminal deletion, i.e., deleting local frames, closing
834 the tty device and restoring its previous state without exiting
837 (Done, but at the moment only called when an error happens during
838 initialization. There is a memory corruption error around this
839 somewhere.) (Update: now it is fully enabled.)
841 -- Implement automatic deletion of terminals when the last frame on
842 that terminal is closed.
846 -- Restore tty screen after closing the terminal.
848 (Done, we do the same as Emacs 21.2 for all terminals.)
850 -- 'TERM=dumb src/emacs' does not restore the terminal state.
854 -- C-g should work on secondary terminals.
856 (Done, but the binding is not configurable.)
858 -- Deal with SIGHUP in Emacs and in emacsclient. (After this, the
859 server-frames may be removed from server.el.)
861 (Done, nothing to do. It seems that Emacs does not receive SIGHUP
862 from secondary ttys, which is actually a good thing.) (Update: I
863 think it would be a bad idea to remove server-frames.)
865 -- Change emacsclient/server.el to support the -t argument better,
866 i.e. automatically close the socket when the frame is closed.
868 (Seems to be working OK.)
870 -- Fix mysterious memory corruption error with tty deletion. To
871 trigger it, try the following shell command:
873 while true; do TERM=no-such-terminal-definition emacsclient -h; done
875 Emacs usually dumps core after a few dozen iterations. (The bug
876 seems to be related to the xfreeing or bzeroing of
877 tty_output.Wcm. Maybe there are outside references to struct Wcm?
878 Why were these vars collected into a struct before multi-tty
881 (Done. Whew. It turned out that the problem had nothing to do
882 with hypothetical external references to Wcm, or any other
883 tty_output component; it was simply that delete_tty closed the
884 filehandles of secondary ttys twice, resulting in fclose doubly
885 freeing memory. Utterly trivial matter. I love the C's memory
886 management, it puts hair on your chest.)
888 -- Support raw secondary terminals. (Note that SIGIO works only on
889 the controlling terminal.) Hint: extend read_input_waiting for
890 multiple ttys and hopefully this will be fixed.
892 (Done, it seems to have been working already for some time. It
893 seems F_SETOWN does work, after all. Not sure what made it fail
894 earlier, but it seems to be fixed (there were several changes
895 around request_sigio, maybe one of them did it).
896 read_input_waiting is only used in sys_select, don't change
897 it.) (Update: After adding X support, it's broken again.)
898 (Update^2: No it isn't.) :-)
900 -- Find out why does Emacs abort when it wants to close its
901 controlling tty. Hint: chan_process[] array. Hey, maybe
902 noninterrupt-IO would work, too? Update: no, there is no process
905 (Done. Added add/delete_keyboard_wait_descriptor to
906 term_init/delete_tty. The hint was right, in a way.)
908 -- Issue with SIGIO: it needs to be disabled during redisplay. See if
909 fcntl kernel behaviour could be emulated by emacsclient.
911 (Done. Simply disabled the SIGIO emulation hack in emacsclient.)
912 (Update: it was added back.) (Update^2: and removed again.)
914 -- server.el: There are issues with saving files in buffers of closed
915 clients. Try editing a file with emacsclient -f, and (without
916 saving it) do a delete-frame. The frame is closed without
917 question, and a surprising confirmation prompt appears in another
920 (Done. delete-frame now asks for confirmation if it still has
921 pending buffers, and modified buffers don't seem to be deleted.)
923 -- emacsclient.el, server.el: Handle eval or file open errors when
928 -- Make parts of struct tty_output accessible from Lisp. The device
929 name and the type is sufficient.
931 (Done, see frame-tty-name and frame-tty-type.)
933 -- Export delete_tty to the Lisp environment, for emacsclient.
935 (Done, see delete-tty.)
937 -- Get rid of the accessor macros in termchar.h, or define macros for
942 -- Move device-specific parameters (like costs) commonly used by
943 device backends to a common, device-dependent structure.
945 (Done. See struct display_method in termhooks.h.)
949 (Done. Well, it seems to be working.)
951 -- Allow simultaneous X and tty frames. (Handling input could be
952 tricky. Or maybe not.)
954 (Done. Allowed, that is. It is currently extremely unstable, to
955 the point of being unusable. The rif variable causes constant
956 core dumps. Handling input is indeed tricky.)
958 -- Rewrite multi-tty input in terms of MULTI_KBOARD.
960 (Done. In fact, there was no need to rewrite anything, I just
961 added a kboard member to tty_display_info, and initialized the
962 frame's kboard from there.)
964 -- Fix rif issue with X-tty combo sessions. IMHO the best thing to do
965 is to get rid of that global variable (and use the value value in
966 display_method, which is guaranteed to be correct).
968 (Done, did exactly that. Core dumps during combo sessions became
969 much rarer. In fact, I have not yet met a single one.)
971 -- Add multi-tty support to talk.el.
975 -- Clean up the source of emacsclient. It is a mess.
977 (Done, eliminated stupid proxy-pty kludge.)
979 -- Fix faces on tty frames during X-tty combo sessions. There is an
980 init_frame_faces call in init_sys_modes, see if there is a problem
983 (Done, there was a stupid mistake in
984 Ftty_supports_face_attributes_p. Colors are broken, though.)
986 -- C-x 5 2, C-x 5 o, C-x 5 0 on an emacsclient frame unexpectedly
987 exits emacsclient. This is a result of trying to be clever with
988 delete-frame-functions.
990 (Fixed, added delete-tty-after-functions, and changed server.el to
993 -- Something with (maybe) multi-keyboard support broke function keys
994 and arrows on ttys during X+tty combo sessions. Debug this.
996 (I can't reproduce it, maybe the terminal type was wrong.)
998 -- Fix input from raw ttys (again).
1000 (Now it seems to work all right.)
1002 -- During an X-tty combo session, a (message "Hello") from a tty frame
1003 goes to the X frame. Fix this.
1005 (Done. There was a safeguard against writing to the initial
1006 terminal frame during bootstrap which prevented echo_area_display
1007 from working correctly on a tty frame during a combo session.)
1009 -- If there are no frames on its controlling terminal, Emacs should
1010 exit if the user presses C-c there.
1012 (Done, as far as possible. See the SIGTERM comment in
1013 interrupt_signal on why this seems to be impossible to solve this
1016 -- During an X session, Emacs seems to read from stdin. Also, Emacs
1017 fails to start without a controlling tty.
1019 (Fixed by replacing the troublesome termcap display with a dummy
1020 bootstrap display during bootstrap.
1022 -- Do tty output through struct display, like graphical display
1027 -- Define an output_initial value for output_method for the initial
1028 frame that is dumped with Emacs. Checking for this frame (e.g. in
1029 cmd_error_internal) is ugly.
1031 (Done, breaking interactive temacs.)
1033 -- The command `emacsclient -t -e '(delete-frame)'' fails to exit.
1037 -- frame-creation-function should always create a frame that is on the
1038 same display as the selected frame. Maybe frame-creation-function
1039 should simply be removed and make-frame changed to do the right
1042 (Done, with a nice hack. frame-creation-function is now frame-local.)
1044 -- Fix C-g on raw ttys.
1046 (Done. I disabled the interrupt/quit keys on all secondary
1047 terminals, so Emacs sees C-g as normal input. This looks like an
1048 overkill, because emacsclient has extra code to pass SIGINT to
1049 Emacs, so C-g should remain the interrupt/quit key on emacsclient
1050 frames. See the next entry why implementing this distinction would
1053 -- Make sure C-g goes to the right frame with ttys. This is hard, as
1054 SIGINT doesn't have a tty parameter. :-(
1056 (Done, the previous change fixes this as a pleasant side effect.)
1058 -- I have seen a case when Emacs with multiple ttys fell into a loop
1059 eating 100% of CPU time. Strace showed this loop:
1062 kill(30284, SIGIO) = 0
1063 --- SIGIO (I/O possible) @ 0 (0) ---
1064 ioctl(6, FIONREAD, [0]) = -1 EIO (Input/output error)
1065 ioctl(5, FIONREAD, [0]) = -1 EIO (Input/output error)
1066 ioctl(0, FIONREAD, [0]) = 0
1067 sigreturn() = ? (mask now [])
1068 gettimeofday({1072842297, 747760}, NULL) = 0
1069 gettimeofday({1072842297, 747806}, NULL) = 0
1070 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1071 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1072 gettimeofday({1072842297, 748245}, NULL) = 0
1074 I have seen something similar with a single X frame, but have not
1075 been able to reproduce it for debugging.
1077 Update: This may have been caused by checking for nread != 0
1078 instead of nread > 0 after calling read_socket_hook in
1081 (Fixed. This was caused by unconditionally including stdin in
1082 input_wait_mask in init_process. The select call in
1083 wait_reading_process_input always returned immediately, indicating
1084 that there is pending input from stdin, which nobody read.
1086 Note that the above strace output seems to be an unrelated but
1087 similar bug. I think that is now fixed.)
1089 -- Exiting Emacs while there are emacsclient frames doesn't restore the
1090 ttys to their default states.
1092 (This seems to be fixed by some previous change.)
1094 -- Allow opening an X session after -nw.
1098 -- Fix color handling during tty+X combo sessions. (It seems that tty
1099 sessions automatically convert the face colors to terminal colors
1100 when the face is loaded. This conversion must happen instead on
1101 the fly in write_glyphs, which might be problematic, as color
1102 approximation is currently done in lisp (term/tty-colors.el).)
1103 (Update: hm, colors seem to work fine if I start emacs with -nw and
1104 then create an X frame. Maybe it's just a small buglet somewhere.)
1106 (Seems to be fixed. The problem was in startup.el, it did not
1107 initialize tty colors when the initial window system was
1110 -- emacs -nw --eval '(y-or-n-p "Foobar")' segfaults. (Reported by
1113 (Fixed, there was a keyboard initialization problem.)
1115 -- Fix interactive use of temacs. There are face-related SEGVs, most
1116 likely because of changes in realize_default_face, realize_face.
1120 -- Don't exit Emacs when the last X connection fails during a
1121 multi-display session.
1125 -- Dan Nicolaescu noticed that starting emacsclient on the same
1126 terminal device that is the controlling tty of the Emacs process
1127 gives unexpected results.
1131 -- Istvan Marko reported that Emacs hang on ttys if it was started
1132 from a shell script.
1134 (Fixed. There was a bug in the multi-tty version of
1135 narrow_foreground_group. tcsetpgrp blocks if it is called from a
1136 process that is not in the same process group as the tty.)
1138 -- emacsclient -t from an Emacs term buffer does not work, complains
1139 about face problems. This can even lock up Emacs (if the recursive
1140 frame sets single_kboard). Update: the face problems are caused by
1141 bugs in term.el, not in multi-tty. The lockup is caused by
1142 single_kboard mode, and is not easily resolvable. The best thing to
1143 do is to simply refuse to create a tty frame of type `eterm'.
1145 (Fixed, changed emacsclient to check for TERM=eterm. The face
1146 complaints seem to be caused by bugs in term.el; they are not
1147 related to multi-tty.)
1149 -- Find out the best way to support suspending Emacs with multiple
1150 ttys. My guess: disable it on the controlling tty, but from other
1151 ttys pass it on to emacsclient somehow. (It is (I hope) trivial to
1152 extend emacsclient to handle suspend/resume. A `kill -STOP' almost
1155 (Done. I needed to play with signal handling and the server
1156 protocol a bit to make emacsclient behave as a normal UNIX program
1157 wrt foreground/background process groups.)
1159 -- There is a flicker during the startup of `emacs -nw'; it's as if
1160 the terminal is initialized, reset and then initialialized again.
1161 Debug this. (Hint: narrow_foreground_group is called twice during
1166 -- Robert Chassell has found serious copy-paste bugs with the
1167 multi-tty branch. There seem to be redisplay bugs while copying
1168 from X to a terminal frame. Copying accented characters do not
1171 (Patch-124 should fix this, by changing the interprogram-*-function
1172 variables to be frame-local, as suggested by Mark Plaksin
1173 (thanks!). I think that the redisplay bugs are in fact not bugs,
1174 but delays caused by single_kboard --> perhaps MULTI_KBOARD should
1177 -- frame-creation-function was removed, which might be a bad idea.
1178 Think up a compatible solution.
1180 (It was an internal interface that may be changed when necessary.)
1182 -- Change Lisp code not to (getenv "TERM"); use the `tty-type' frame
1183 parameter or the frame-tty-type function instead. (M-x tags-search
1184 "TERM" helps with this.) Update: Actually, all getenv invocations
1185 should be checked for multi-tty compatibility, and an interface
1186 must be implemented to get the remote client's environment.
1188 (Done. Only getenv calls in lisp/term/*.el were changed; other
1189 calls should be mostly left as they are.)
1191 -- Add an elaborate mechanism for display-local variables. (There are
1192 already a few of these; search for `terminal-local' in the Elisp
1195 (Not needed. Display-local variables could be emulated by
1196 frame-local variables.)
1198 -- Emacs assumes that all terminal frames have the same locale
1199 settings as Emacs itself. This may lead to bogus results in a
1200 multi-locale setup. (E.g., while logging in from a remote client
1201 with a different locale.)
1202 (Update after new bugreport by Friedrich Delgado Friedrichs:
1203 (at least) the structs terminal_coding and keyboard_coding in
1204 coding.c must be moved to struct display, and the Lisp interface
1205 [set-]keyboard-coding-system must be adapted for the change.)
1207 (Fixed. Emacs now uses the locale settings as seen by the
1208 emacsclient process for server tty frames.)
1209 (Update: Not really; Vlocale_coding_system is still global.)
1211 -- Make `struct display' accessible to Lisp programs. Accessor functions:
1213 (displayp OBJECT): Returns t if OBJECT is a display.
1214 => Implemented as display-live-p.
1216 (display-list): Returns list of currently active displays.
1219 (selected-display): Returns the display object of the selected frame.
1220 => Not strictly necessary, but implemented anyway.
1222 (frame-display FRAME): Returns the display object of FRAME.
1225 (display-frames DISPLAY): Returns a list of frames on DISPLAY.
1226 => Already implemented, see frames-on-display-list.
1228 (display-type DISPLAY): Returns the type of DISPLAY, as a
1229 symbol. (See `framep'.)
1230 => Implemented as display-live-p.
1232 (display-device DISPLAY): Returns the name of the device that
1233 DISPLAY uses, as a string. (E.g: "/dev/pts/16", or
1235 => Implemented as display-name.
1239 See next issue why this is necessary.
1241 (Update: The consensus on emacs-devel seems to be to do this via
1242 integer identifiers. That's fine by me.)
1246 -- The following needs to be supported:
1253 The cleanest way to solve this is to allow multiple displays on the
1254 same terminal device; each new emacsclient process should create
1255 its own display. As displays are currently identified by their
1256 device names, this is not possible until struct display becomes
1257 accessible as a Lisp-level object.
1261 -- Miles Bader suggests that C-x C-c on an emacsclient frame should
1262 only close the frame, not exit the entire Emacs session. Update:
1263 see above for a function that does this. Maybe this should be the
1266 (Done. This is the new default. No complaints so far.)
1268 -- Clean up the frame-local variable system. I think it's ugly and
1269 error-prone. But maybe I just haven't yet fully understood it.
1271 (Nothing to do. It doesn't seem ugly any more. It's rather clever.)
1273 -- Support multiple character locales. A version of
1274 `set-locale-environment' needs to be written for setting up
1275 display-local settings on ttys. I think calling
1276 set-display-table-and-terminal-coding-system and
1277 set-keyboard-coding-system would be enough. The language
1278 environment itself should remain a global setting.
1280 (Done, by an ugly hack.)
1282 -- The terminal customization files in term/*.el tend to change global
1283 parameters, which may confuse Emacs with multiple displays. Change
1284 them to tweak only frame-local settings, if possible. (They tend
1285 to call define-key to set function key sequences a lot.)
1287 (Done, by making `function-key-map' terminal-local (i.e., part of
1288 struct kboard). This has probably covered all the remaining problems.)
1290 -- Make `function-key-map' and `key-translation-map' terminal-local.
1294 -- Implement `terminal-local-value' and `set-terminal-local-value' to
1295 allow deterministic access to terminal local bindings. The
1296 encode-kb package can not set up `key-translation-map' without
1297 these. The terminal-local bindings seem to be independent of what
1302 -- xt-mouse.el needs to be adapted for multi-tty. It currently
1303 signals an error on kill-emacs under X, which prevents the user
1304 from exiting Emacs. (Reported by Mnemonikk on freenode.)
1309 -- Having {reset,init}_all_sys_modes in set-input-mode breaks arrow
1310 keys on non-selected terminals under screen, and sometimes on other
1311 terminal types as well. The other function keys continue to work
1312 fine. Sometimes faces on these screens become garbled.
1314 This only seems to affect displays that are of the same terminfo
1315 type as the selected one. Interestingly, in screen Emacs normally
1316 reports the up arrow key as `M-o A', but after the above SNAFU, it
1317 complains about `M-[ a'. UNIX ttys are a complete mystery to me,
1318 but it seems the reset-reinitialize cycle somehow leaves the
1319 non-selected terminals in a different state than usual. I have no
1320 idea how this could happen.
1322 Currently set-input-mode resets only the currently selected
1323 terminal, which seems to somehow work around the problem.
1327 Dan Nicolaescu <dann@ics.uci.edu> writes:
1328 > Some terminals have 2 modes for cursor keys: Application Mode where
1329 > the cursor keys transmit the codes defined in the terminfo entry, and
1330 > Cursor mode. Applications have to send the smkx and rmkx terminfo
1331 > strings to switch between the 2 modes. So Emacs (and emacsclient) have
1332 > to send smkx when initializing and rmkx when quitting (or on
1335 (I think patch-370 fixed this.)
1337 -- This long-standing bug (first reported by Han Boetes) seems to come
1338 and go all the time. It is time to track it down and fix it.
1343 # From another xterm:
1344 emacsclient -e '(y-or-n-p "Do you want me to crash? ")'
1345 # Notice how the answer ends up in the *scratch* buffer
1349 (Fixed in patch-414 after detailed analysis by Kalle Olavi Niemitalo.)
1351 -- normal-erase-is-backspace-mode in simple.el needs to be updated for
1352 multi-tty (rep. by Dan Waber). (The Delete key is broken on X
1355 (Fixed in patch-427.)
1357 -- I think keyboard-translate-table should be made terminal-local.
1359 (Done in patch-431.)
1361 -- The semantics of terminal-local variables are confusing; it is not
1362 clear what binding is in effect in any given time. See if
1363 current_kboard (or at least the terminal-local bindings exported to
1364 Lisp) might be changed to be tied to the selected frame instead.
1365 Currently, `function-key-map' and `key-translation-map' may be
1366 accessed reliably only using the hackish
1367 `(set-)terminal-local-value' functions.
1369 Perhaps there should be a difference between `last-command' &co.
1370 and these more conventional configuration variables.
1371 (E.g. `symbol-value' would use current_kboard to access
1372 `last-command', but SELECTED_FRAME()->display->kboard to get the
1373 value of `function-key-map'.
1375 (Fixed in patch-434.)
1377 -- If the first key pressed on a new tty terminal is a function key,
1378 it is not recognized correctly. May be related to the bug below.
1380 (Seems to have been fixed as a side effect of patch-434. "The bug
1381 below" was the set-input-mode madness.)
1383 (Update: this bug was fixed for good in patch-449. It was tracked
1384 down to a bug in `read_key_sequence': it failed to reinitialize its
1385 local function-key-map/key-translation-map references when it
1386 switched keyboards. I don't understand why did this bug only
1387 appear on brand new frames, though!)
1389 -- Disable connecting to a new X display when we use the GTK toolkit.
1391 (Disabled in patch-450.)
1393 -- Implement automatic forwarding of client environment variables to
1394 forked processes, as discussed on the multi-tty list. Terminal
1395 parameters are now accessible in C code, so the biggest obstacle is
1396 gone. The `getenv_internal' and `child_setup' functions in
1397 callproc.c must be changed to support the following variable:
1399 terminal-local-environment-variables is a variable defined in ...
1401 Enable or disable terminal-local environment variables.
1403 If set to t, `getenv', `setenv' and subprocess creation
1404 functions use the environment variables of the emacsclient
1405 process that created the selected frame, ignoring
1406 `process-environment'.
1408 If set to nil, Emacs uses `process-environment' and ignores
1409 the client environment.
1411 Otherwise, `terminal-local-environment-variables' should be a
1412 list of variable names (represented by Lisp strings) to look
1413 up in the client environment. The rest will come from
1414 `process-environment'.
1416 (Implemented in patch-461; `terminal-getenv', `terminal-setenv' and
1417 `with-terminal-environment' are now replaced by extensions to
1418 `getenv' and `setenv', and the new `local-environment-variables'
1421 (Updated in patch-465 to fix the semantics of let-binding
1422 `process-environment'. `process-environment' was changed to
1423 override all local/global environment variables, and a new variable
1424 `global-environment' was introduced to have `process-environment's
1427 (Updated in patch-466 to fix the case when two emacsclient sessions
1428 share the same terminal, but have different environment. The local
1429 environment lists are now stored as frame parameters, so the
1430 C-level terminal parameters are not strictly necessary any more.)
1432 -- `Fdelete_frame' is called from various critical places where it is
1433 not acceptable for the frame deletion to fail, e.g. from
1434 x_connection_closed after an X error. `Fdelete_frame' now protects
1435 against `delete-frame-functions' throwing an error and preventing a
1436 frame delete. (patch-475)
1438 -- Fix set-input-mode for multi-tty. It's a truly horrible interface;
1439 what if we'd blow it up into several separate functions (with a
1440 compatibility definition)?
1442 (Done. See `set-input-interrupt-mode', `set-output-flow-control',
1443 `set-input-meta-mode' and `set-quit-char'.) (patch-457)
1445 -- Let-binding `overriding-terminal-local-map' on a brand new frame
1446 does not seem to work correctly. (See `fancy-splash-screens'.)
1447 The keymap seems to be set up right, but events go to another
1448 terminal. Or is it `unread-command-events' that gets Emacs
1449 confused? Investigate.
1451 (Emacs was confused because a process filter entered
1452 `recursive-edit' while Emacs was reading input. I added support
1453 for this in the input system.) (patch-489)
1455 -- I smell something funny around pop_kboard's "deleted kboard" case.
1456 Determine what are the circumstances of this case, and fix any
1457 bug that comes to light.
1459 (It happens simply because single_kboard's terminal is sometimes
1460 deleted while executing a command on it, for example the one that
1461 kills the terminal. There was no bug here, but I rewrote the whole
1462 single_kboard mess anyway.) (patch-489)
1464 -- Understand Emacs's low-level input system (it's black magic) :-)
1465 What exactly does interrupt_input do? I tried to disable it for
1466 raw secondary tty support, but it does not seem to do anything
1467 useful. (Update: Look again. X unconditionally enables this, maybe
1468 that's why raw terminal support is broken again. I really do need
1469 to understand input.)
1470 (Update: I am starting to understand the read_key_sequence->read-char
1471 ->kbd_buffer_get_event->read_avail_input->read_socket_hook path. Yay!)
1473 (Update: OK, it all seems so easy now (NOT). Input could be done
1474 synchronously (with wait_reading_process_input), or asynchronously
1475 by SIGIO or polling (SIGALRM). C-g either sets the Vquit_flag,
1476 signals a 'quit condition (when immediate_quit), or throws to
1477 `getcjmp' when Emacs was waiting for input when the C-g event
1480 -- Replace wrong_kboard_jmpbuf with a special return value of
1481 read_char. It is absurd that we use setjmp/longjmp just to return
1482 to the immediate caller.
1484 (Done in patch-500.)
1486 -- `tool-bar-mode', `scroll-bar-mode', `menu-bar-mode' and
1487 'fringe-mode' are modes global to the entire Emacs session, not
1488 just a single frame or a single terminal. This means that their
1489 status sometimes differs from what's actually displayed on the
1490 screen. As a consequence, the Options | Show/Hide menu sometimes
1491 shows incorrect status, and you have to select an option twice for
1492 it to have any visible effect on the current frame.
1494 Change Emacs so that the status of the items in the Options |
1495 Show/Hide menu correspond to the current frame.
1497 (Done in patch-537.)
1499 -- The `default-directory' variable should somehow be set to the
1500 cwd of the emacsclient process when the user runs emacsclient
1501 without file arguments. Perhaps it is OK to just override the
1502 directory of the *scratch* buffer.
1504 (Done in patch-539.)
1506 -- The borders on tooltip frames on X are messed up. More
1507 specifically, the frame's internal border (internal-border-width
1508 frame parameter) is not filled with the correct background color.
1510 It seems the border contents are drawn onto by the
1511 update_single_window call in `x-show-tip'. After some debugging, I
1512 think the window's background color is not set up
1513 correctly---calling `x_clear_area' fills the specified area with
1514 black, not light yellow.
1516 (Done in patch-544. A background_pixel field was defined both in
1517 struct frame and struct x_output, and Emacs got confused between
1520 ;;; arch-tag: 8da1619e-2e79-41a8-9ac9-a0485daad17d