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 Noah Friedman <friedman@splode.com>
44 Friedrich Delgado Friedrichs <friedel@nomaden.org>
45 IRIE Tetsuya <irie@t.email.ne.jp>
46 Yoshiaki Kasahara <kasahara@nc.kyushu-u.ac.jp>
47 Bas Kok <nekkobassu@yahoo.com>
48 Jurej Kubelka <Juraj.Kubelka@email.cz>
49 David Lichteblau <david@lichteblau.com>
50 Richard Lewis <rtf@jabble.com>
51 mace <mace@kirjakaapeli.lib.hel.fi>
52 Suresh Madhu <madhu@cs.unm.edu>
53 Xavier Mallard <zedek@gnu-rox.org>
54 Istvan Marko <mi-mtty@kismala.com>
55 Ted Morse <morse@ciholas.com>
56 Gergely Nagy <algernon@debian.org>
57 Dan Nicolaescu <dann@ics.uci.edu>
58 Kalle Olavi Niemitalo <kon@iki.fi>
59 Mark Plaksin <happy@mcplaksin.org>
60 Frank Ruell <stoerte@dreamwarrior.net>
61 Tom Schutzer-Weissmann <trmsw@yahoo.co.uk>
62 Joakim Verona <joakim@verona.se>
63 Dan Waber <dwaber@logolalia.com>
66 Richard Stallman was kind enough to review an earlier version of my
73 The multi-tty mailing list (discussion & bug reports):
75 Address: multi-tty@lists.fnord.hu
76 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty
77 Archive: http://news.gmane.org/gmane.emacs.multi-tty/
79 Commit notifications (read-only):
81 Address: multi-tty-commits@lists.fnord.hu
82 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty-commits
88 The branch is now very stable and almost full-featured. All of the
89 major problems have been fixed, only a few minor issues remain. (It
90 still needs to be ported to Windows/Mac/DOS, though.) Both multiple
91 tty device support and simultaneous X and tty frame support works
92 fine. Emacsclient has been extended to support opening new tty and X
93 frames. It has been changed to open new Emacs frames by default.
95 The multi-tty branch has been scheduled for inclusion in the next
96 major release of Emacs (version 23). I expect the merge into the
97 development trunk to occur sometime during next year (2006), after the
98 merge of the Unicode branch.
100 Tested on GNU/Linux, Solaris 8, FreeBSD and OpenBSD. Please let me
101 know if you succeed or fail to use it on other platforms---I'll have a
102 few tricky test cases for you.
106 * GTK support. If you compile your Emacs with the GTK
107 toolkit, some functionality of multi-tty will be lost. In
108 particular, you will not be able to work on multiple X
109 displays at once. Current releases of GTK have limitations
110 and bugs that prevent full-blown multi-display support in
111 Emacs. (GTK crashes when Emacs tries to disconnect from an
112 X server.) Use the Lucid toolkit if you want to see a
113 complete feature set.
115 * The single-kboard mode.
117 If your multi-tty Emacs session seems to be frozen, you
118 probably have a recursive editing session or a pending
119 minibuffer prompt (which is a kind of recursive editing) on
120 another display. To unfreeze your session, switch to that
121 display and complete the recursive edit, for example by
122 pressing C-] (`abort-recursive-edit').
124 I am sorry to say that currently there is no way to break
125 out of this "single-kboard mode" from a frozen display. If
126 you are unable to switch to the display that locks the
127 others (for example because it is on a remote computer),
128 then you can use emacsclient to break out of all recursive
131 emacsclient -e '(top-level)'
133 Note that this (perhaps) unintuitive behaviour is by design.
134 Single-kboard mode is required because of an intrinsic Emacs
135 limitation that is very hard to eliminate. (This limitation
136 is related to the single-threaded nature of Emacs.)
138 I plan to implement better user notification and support for
139 breaking out of single-kboard mode from locked displays.
141 * Mac, Windows and DOS support is broken, doesn't even
142 compile. Multiple display support will probably not provide
143 new Emacs features on these systems, but the multi-tty
144 branch changed a few low-level interfaces, and the
145 system-dependent source files need to be adapted
146 accordingly. The changes are mostly trivial, so almost
147 anyone can help, if only by compiling the branch and
148 reporting the compiler errors.
151 HOW TO GET THE BRANCH
152 ---------------------
154 The branch uses Bazaar 1 (http://bazaar.canonical.com) for version control.
156 Retrieving the latest version of the branch:
158 baz register-archive -f http://aszt.inf.elte.hu/~lorentey/mirror/arch/2004
159 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
161 This incantation uses an archive mirror that is hosted on a
162 high-bandwidth site. Please note that on average there is a two-hour
163 delay for commits to arrive on this mirror. My primary mirror is on the
164 low-bandwidth http://lorentey.hu/ site:
166 baz register-archive -f http://lorentey.hu/arch/2004/
167 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
169 This is "instantly" updated, but very slow from outside Hungary.
170 (By "instantly" I mean as soon as I connect the notebook I work on to
171 a network. It could take days.)
173 The Arch supermirror provides mirroring services for all public Arch
174 repositories. We have a mirror there, too, if you prefer.
176 baz register-archive -f http://mirrors.sourcecontrol.net/lorentey%40elte.hu--2004
177 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
179 My GPG key id is 0FB27A3F; it is available from
180 hkp://wwwkeys.eu.pgp.net/, or from my homepage at
181 http://lorentey.hu/rolam/gpg.html)
183 Don't worry if the above checkout takes a few minutes to complete;
184 once you have a source tree, updating it to the latest revision will
185 be _much_ faster. Use the following command for the update:
189 You can find more information about Bazaar on
190 http://bazaar.canonical.com/. It's a distributed source control
191 system that is somewhat less broken than competing projects.
193 If you don't have Bazaar, the branch has a homepage from which you can
194 download conventional patches against Emacs CVS HEAD:
196 http://lorentey.hu/project/emacs.html
198 I suggest you use Bazaar whenever feasible.
204 If you run Debian, or a distribution based on Debian, you are welcome
205 to use our binary packages; put these lines in your /etc/apt/sources.list:
208 deb http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
209 deb-src http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
211 Note that these packages are intended solely to provide an easy way to
212 test the new multi-tty features. They are not to be taken as Emacs
213 releases, and it's a mistake to expect robust operation or any kind of
214 timely support for them. Do not install them, unless you'd like to
215 have your editor crash on you.
221 The multi-tty branch is compiled the same way as Emacs itself:
223 make maintainer-clean # (If you have compiled Emacs before)
225 ./configure --without-gtk <your favourite options>
229 If you have strange compilation errors, they may be caused by old
230 *.elc files that are left over from an earlier bootstrap. The `make
231 maintainer-clean' target deletes them, so it is a good idea to run
232 that before reporting a bug. (Emacs requires a clean recompilation
233 after certain kinds of source code changes.)
238 To test the multi-tty branch, start up the Emacs server with the
244 and then (from a shell prompt on another terminal) start emacsclient
246 emacsclient -t /optional/file/names... (for a tty frame)
247 emacsclient /optional/file/names... (for an X frame)
249 (Make sure both emacs and emacsclient are multi-tty versions.)
250 You'll hopefully have two fully working, independent frames on
251 separate terminals. The new frame is closed automatically when you
252 finish editing the specified files (C-x #), but delete-frame (C-x 5 0)
253 also works. Of course, you can create frames on more than two tty
256 Creating new frames on the same tty with C-x 5 2 (make-frame-command)
257 works, and behaves the same way as in previous Emacs versions. If you
258 exit emacs, all terminals should be restored to their previous states.
260 This is work in progress, and probably full of bugs. It is a good
261 idea to run emacs from gdb, so that you'll have a live instance to
262 debug if something goes wrong. Please send me your bug reports on our
263 mailing list: multi-tty@lists.fnord.hu
268 I think the best way to use the new Emacs is to have it running inside
269 a disconnected GNU screen session, and always use emacsclient for
270 normal work. One advantage of this is that not a single keystroke of
271 your work will be lost if the display device that you are using
272 crashes, or the network connection times out, or whatever. (I had an
273 extremely unstable X server for some time while I was developing these
274 patches, and running Emacs this way has saved me a number of M-x
275 recover-session invocations.)
277 I use the following two bash scripts to handle my Emacs sessions:
279 -------------------------------------------------------connect-emacs--
281 # Usage: connect-emacs <name> <args>...
283 # Connects to the Emacs instance called NAME. Starts up the instance
284 # if it is not already running. The rest of the arguments are passed
290 if [ -z "$name" ]; then
291 echo "Usage: connect_emacs <name> <args>..." >&2
294 preload-emacs "$name" wait
295 /usr/bin/emacsclient.emacs-multi-tty -s "$name" "$@"
296 ----------------------------------------------------------------------
298 -------------------------------------------------------preload-emacs--
300 # Usage: preload-emacs <name> [<waitp>]
302 # Preloads the Emacs instance called NAME in a detached screen
303 # session. Does nothing if the instance is already running. If WAITP
304 # is non-empty, the function waits until the server starts up and
305 # creates its socket; otherwise it returns immediately.
309 screendir="/var/run/screen/S-$USER"
310 serverdir="/tmp/emacs$UID"
311 emacs=/usr/bin/emacs-multi-tty # Or wherever you installed your multi-tty Emacs
313 if [ -z "$name" ]; then
314 echo "Usage: preload_emacs <name> [<waitp>]" >&2
318 if [ ! -e "$screendir"/*."$name" ]; then
319 if [ -e "$serverdir/$name" ]; then
320 # Delete leftover socket (for the wait option)
321 rm "$serverdir/$name"
323 screen -dmS "$name" "$emacs" -nw --eval "(setq server-name \"$name\")" -f server-start
325 if [ ! -z "$waitp" ]; then
326 while [ ! -e "$serverdir/$name" ]; do sleep 0.1; done
328 ----------------------------------------------------------------------
330 I have the following in my profile to have two instances automatically
331 preloaded for editing and email:
336 It is useful to set up short aliases for connect-emacs. I use the
339 alias edit="connect-emacs editor"
341 alias et="connect-emacs editor -t"
342 alias gnus="connect-emacs gnus"
354 For the NEWS file: (Needs much, much work)
356 ** Support for multiple terminal devices and simultaneous graphical
357 and tty frames has been added. You can test for the presence of
358 this feature in your Lisp code by testing for the `multi-tty'
361 *** The `window-system' variable has been made frame-local. The new
362 `initial-window-system' variable contains the `window-system'
363 value for the first frame.
365 *** You can specify a terminal device (`tty' parameter) and a terminal
366 type (`tty-type' parameter) to `make-terminal-frame'.
368 *** The new function `make-frame-on-tty' allows you to create a new
369 frame on another tty device interactively.
371 *** The function `make-frame-on-display' now works during a tty
372 session, and `make-frame-on-tty' works during a graphical session.
374 *** Emacsclient has been extended to support opening a new terminal
375 frame. Its behaviour has been changed to open a new Emacs frame by
376 default. Use the -c option to get the old behavior of opening
377 files in the currently selected Emacs frame.
379 *** C-z now invokes `suspend-frame', C-x C-c now invokes
380 `save-buffers-kill-terminal'.
382 *** New functions: frame-tty-name, frame-tty-type, delete-tty,
383 suspend-tty, resume-tty, terminal-id, terminal-parameters,
384 terminal-parameter, set-terminal-parameter,
385 modify-terminal-parameters, environment, let-environment
387 *** New variables: local-key-translation-map, local-function-key-map
389 *** The `keyboard-translate-table' variable and the terminal and
390 keyboard coding systems have been made terminal-local.
392 *** In addition to the global key-translation-map and
393 function-key-map, Emacs has terminal-local
394 local-key-translation-map and local-function-key-map variables,
395 and uses them instead of the global keymaps to set up translations
396 and function key sequences relevant to a specific terminal device.
398 *** talk.el has been extended for multiple tty support.
402 (The rest of this file consists of my development notes and as such it
403 is probably not very interesting for anyone else.)
408 ** See if `tty-defined-color-alist' needs to be terminal-local.
410 ** emacsclient -t on the console does not work after su. You have to
411 use non-root accounts or start as root to see this.
417 *ERROR*: Could not open file: /dev/tty1
419 The tty can be opened as /dev/tty by emacsclient, but not by Emacs.
420 This seems to be a serious problem. Currently my only idea is to
421 bring back the ugly pty proxy hack from the initial versions of
422 multi-tty. Suggestions would be appreciated.
424 Update: we could change emacsclient to pass its open file
425 descriptor to the Emacs process. Unfortunately, this requires a
426 new Lisp-level Emacs API, and as file descriptors are not otherwise
427 exported to Lisp, this approach seems at least as ugly as the pty
430 ** lisp/vc.el depends on the terminal type during load time.
431 `vc-annotate-color-map' is one example that needs to be fixed.
433 ** Understand how `quit_throw_to_read_char' works, and fix any bugs
436 ** See if getcjmp can be eliminated somehow. Why does Emacs allow
437 asynchronous input processing while it's reading input anyway?
439 ** `delete-frame' events are handled by `special-event-map'
440 immediately when read by `read_char'. This is fine but it prevents
441 higher-level keymaps from binding that event to get notified of the
444 Sometimes it would be useful for Lisp code to be notified of frame
445 deletions after they have happened, usually because they want to
446 clean up after the deleted frame. Not all frame-local states can
447 be stored as a frame parameter. For example,
448 `display-splash-screen' uses `recursive-edit' with a special keymap
449 override to create its buffer---and it leads to all kinds of
450 nastiness if Emacs stays in this recursive edit mode after the
451 frame containing the splash screen is deleted. Basically, the
452 splash-screen implementation wants to throw out of the recursive
453 edit when the frame is deleted; however, it is not legal to throw
454 from `delete-frame-functions' because `delete-frame' must not fail.
455 (Introducing `delete-frame-after-functions' would not help either
456 because `delete-frame' may not fail at that time either.)
458 Currently `fancy-splash-screens' installs a
459 `delete-frame-functions' hook that sets up a timer to exit the
460 recursive edit. This is an adequate solution, but it would perhaps
461 be better to have something like a `frame-deleted' event that could
462 be bound in the normal way.
464 ** Trouble: `setenv' doesn't actually set environment variables in the
465 Emacs process. This defeats the purpose of the elaborate
466 `server-with-environment' magic around the `tgetent' call in
469 ** (Possibly) create hooks in struct device for creating frames on a
470 specific terminal, and eliminate the hackish terminal-related frame
471 parameters (display, tty, tty-type).
476 ** Decide whether to keep the C implementation of terminal parameters,
477 or revert to the previous, purely Lisp code. It turned out that
478 local environments do not need terminal parameters after all.
480 ** Move Fsend_string_to_terminal to term.c, and declare get_named_tty
481 as static, removing it from dispextern.h.
482 Move fatal to emacs.c and declare it somewhere.
484 ** Search for `suspend-emacs' references and replace them with
485 `suspend-frame', if necessary. Ditto for `save-buffers-kill-emacs'
486 vs. `save-buffers-kill-display'.
488 ** Emacs crashes when a tty frame is resized so that there is no space
489 for all its windows. (Tom Schutzer-Weissmann)
491 ** Report GTK multi-display problems to GTK maintainers. For extra
494 Currently you can not connect to new X displays when you compile
495 Emacs with GTK support. If you want to play around with GTK
496 multidisplay (and don't mind core dumps), you can edit src/config.h
497 and define HAVE_GTK_MULTIDISPLAY there by hand.
499 http://bugzilla.gnome.org/show_bug.cgi?id=85715
501 Update: Han reports that GTK+ version 2.8.9 almost gets display
502 disconnects right. GTK will probably be fully fixed by the time
503 multi-tty gets into the trunk.
505 Update: I am still having problems with GTK+ 2.8.10. I have the
506 impression that the various multidisplay fixes will only get
507 released in GTK+ 2.10.
509 ** Audit `face-valid-attribute-values' usage in customize and
510 elsewhere. Its return value depends on the current window system.
511 Replace static initializers using it with runtime functions. For
512 example, custom's buttons are broken on non-initial device types.
514 ** Possibly turn off the double C-g feature when there is an X frame.
515 C.f. (emacs)Emergency Escape.
517 ** frames-on-display-list should also accept frames.
519 ** Consider the `tty-type' frame parameter and the `display-tty-type'
520 function. They serve the exact same purpose. I think it may be
521 a good idea to eliminate one of them, preferably `tty-type'.
523 ** The handling of lisp/term/*.el, and frame creation in general, is a
524 big, big mess. How come the terminal-specific file is loaded by
525 tty-create-frame-with-faces? I don't think it is necessary to load
526 these files for each frame; once per terminal should be enough.
527 Update: lisp/term/*.el is not loaded repeatedly anymore, but
528 faces.el still needs to be cleaned up.
530 ** Fix frame-set-background-mode in this branch. It was recently
531 changed in CVS, and frame.el in multi-tty has not yet been adapted
532 for the changes. (It needs to look at
533 default-frame-background-mode.) (Update: maybe it is fixed now;
534 needs testing.) (Note that the byte compiler has this to say about
537 term/rxvt.el:309:17:Warning: assignment to free variable
538 `default-frame-background-mode'
540 ** I think `(set-)terminal-local-value' and the terminal parameter
541 mechanism should be integrated into a single framework.
543 (Update: `(set-)terminal-local-value' is now eliminated, but the
544 terminal-local variables should still be accessible as terminal
545 parameters. This also applies to `display-name' and similar
548 ** Add the following hooks: after-delete-frame-hook (for server.el,
549 instead of delete-frame-functions),
550 after-delete-terminal-functions, after-create-terminal-functions.
552 ** BULK RENAME: The `display-' prefix of new Lisp-level functions
553 conflicts with stuff like `display-time-mode'. Use `device-'
554 or `terminal-' instead. I think I prefer `terminal-'.
556 It turns out that most of the offending Lisp functions were defined
557 in the trunk. Therefore, compatibility aliases should be defined
558 for the following names:
560 display-color-cells terminal-color-cells
561 display-color-p terminal-color-p
562 display-graphic-p terminal-graphic-p
563 display-grayscale-p terminal-grayscale-p
564 display-images-p terminal-images-p
565 display-mm-height terminal-mm-height
566 display-mm-width terminal-mm-width
567 display-mouse-p terminal-mouse-p
568 display-multi-font-p terminal-multi-font-p
569 display-multi-frame-p terminal-multi-frame-p
570 display-pixel-height terminal-pixel-height
571 display-pixel-width terminal-pixel-width
572 display-pixels-per-inch terminal-pixels-per-inch
573 display-planes terminal-planes
574 display-popup-menus-p terminal-popup-menus-p
575 display-save-under terminal-save-under
576 display-screens terminal-screens
577 display-supports-face-attributes-p terminal-supports-face-attributes-p
578 display-visual-class terminal-visual-class
579 framep-on-display framep-on-terminal
580 frames-on-display-list frames-on-terminal-list
582 The following functions were introduced in the multi-tty branch, and
583 were renamed without aliases:
585 delete-display delete-terminal
586 display-controlling-tty-p controlling-tty-p
587 display-list terminal-list
588 display-live-p terminal-live-p
589 display-name terminal-name
590 display-tty-type tty-type
591 frame-display frame-terminal
592 selected-display selected-terminal
594 ** The single-keyboard mode of MULTI_KBOARD is extremely confusing
595 sometimes; Emacs does not respond to stimuli from other keyboards.
596 At least a beep or a message would be important, if the single-mode
597 is still required to prevent interference. (Reported by Dan
600 Update: selecting a region with the mouse enables single_kboard
601 under X. This is very confusing.
603 Update: After discussions with Richard Stallman, this will be
604 resolved by having locked displays warn the user to wait, and
605 introducing a complex protocol to remotely bail out of
606 single-kboard mode by pressing C-g.
608 Update: Warning the user is not trivial to implement, as Emacs has
609 only one echo area, shared by all frames. Ideally the warning
610 should not be displayed on the display that is locking the others.
611 Perhaps the high probability of user confusion caused by
612 single_kboard mode deserves a special case in the display code.
613 Alternatively, it might be good enough to signal single_kboard mode
614 by changing the modelines or some other frame-local display element
615 on the locked out displays.
617 Update: In fact struct kboard does have an echo_string slot.
619 ** The session management module is prone to crashes when the X
620 connection is closed and then later I try to connect to a new X
623 #0 0xb7ebc806 in SmcGetIceConnection () from /usr/X11R6/lib/libSM.so.6
624 #1 0x080e6641 in x_session_check_input (bufp=0xbf86c9c0) at xsmfns.c:144
625 #2 0x080d3bbc in XTread_socket (device=0xa722ff8, expected=1, hold_quit=0xbf86ca90) at xterm.c:7037
626 #3 0x080fa404 in read_avail_input (expected=1) at keyboard.c:6696
627 #4 0x080fa4ca in handle_async_input () at keyboard.c:6900
628 #5 0x080d51fa in x_term_init (display_name=162628899, xrm_option=0x0, resource_name=0x857068c "emacs") at xterm.c:10622
629 #6 0x080d920e in x_display_info_for_name (name=162628899) at xfns.c:3975
630 #7 0x080d92f9 in check_x_display_info (object=1) at xfns.c:274
631 #8 0x080d97b8 in Fx_create_frame (parms=151221485) at xfns.c:3016
632 #9 0x0815bf72 in Ffuncall (nargs=2, args=0xbf86ceec) at eval.c:2851
634 I installed a workaround to prevent this. The X session manager is
635 only contacted when the very first display in the Emacs session is
636 an X display. Also, x_delete_display() on this display aborts
637 session management, and XTread_socket only calls
638 x_session_check_input when it is called for the display that the
639 session was opened on. While this does not really fix the bug, it
640 makes it much less frequent, because session manager support will
641 not normally be enabled when Emacs can survive the shutdown of the
644 See if xsmfns.c should be updated.
646 ** Hunt down display-related functions in frame.el and extend them all
647 to accept display ids.
649 ** rif->flush_display_optional (NULL) calls should be replaced by a
652 ** The set-locale-environment hack (adding the DISPLAY option) should
653 be replaced with a clean design.
655 ** standard-display-table should be display-local.
656 standard-display-european should be display-local.
658 ** With iswitchb-default-method set to 'always-frame, only frames on
659 the current display should be considered. This might involve
660 extending `get-buffer-window'.
662 ** Have a look at Vlocale_coding_system. Seems like it would be a
663 tedious job to localize it, although most references use it for
664 interfacing with libc and are therefore OK with the global
667 Exceptions found so far: x-select-text and
668 x-cut-buffer-or-selection-value.
670 ** Have a look at fatal_error_hook.
672 ** Have a look at set_frame_matrix_frame.
674 ** Check if we got term-setup-hook right.
676 ** I think tip_frame should be display-local.
678 ** Check display reference count handling in x_create_tip_frame.
680 ** make-frame does not correctly handle extra parameters in its
683 (frame-parameter (make-frame (list (cons 'foobar 42))) 'foobar)
686 (This is likely an error in the CVS trunk.)
688 ** Dan Nicolaescu suggests that -nw should be added as an alias for -t
689 in emacsclient. Good idea. (Alas, implementing this is not
690 trivial, getopt_long does not seem to support two-letter ``short''
691 options. Patches are welcome.)
693 ** Mark Plaksin suggests that emacsclient should accept the same
694 X-related command-line arguments as Emacs. Most of the X-related
695 argument-handling is done in Lisp, so this should be quite easy to
698 ** Gergely Nagy suggests that C-x # should only kill the current
699 frame, not any other emacsclient frame that may have the same file
700 opened for editing. I think I agree with him.
702 ** Very strange bug: visible-bell does not work on secondary
703 terminals in xterm and konsole. The screen does flicker a bit,
704 but it's so quick it isn't noticable.
706 (Update: This is probably some problem with padding or whatnot on
707 the secondary terminals.)
709 ** Move baud_rate to struct display.
711 ** Implement support for starting an interactive Emacs session without
712 an initial frame. (The user would connect to it and open frames
713 later, with emacsclient.)
715 ** Fix Mac support (I can't do this entirely myself). Note that the
716 current state of Mac-specific source files in the multi-tty tree
717 are not useful; before starting work on Mac support, revert to
718 pristine, pre-multi-tty versions.
720 ** Fix W32 support (I can't do this entirely myself). Note that the
721 current state of W32-specific source files in the multi-tty tree
722 are not useful; before starting work on W32 support, revert to
723 pristine, pre-multi-tty versions.
725 ** Fix DOS support (I can't do this entirely myself). Note that the
726 current state of DOS-specific source files in the multi-tty tree
727 are not useful; before starting work on DOS support, revert to
728 pristine, pre-multi-tty versions.
730 ** Do a grep on XXX and ?? for more issues.
732 ** flow-ctrl.el must be updated.
734 ** Fix stuff_char for multi-tty. Doesn't seem to be of high priority.
739 (ex-TODO items with explanations.)
741 -- Introduce a new struct for terminal devices.
743 (Done, see struct tty_output. The list of members is not yet
746 -- Change the bootstrap procedure to initialize tty_list.
748 (Done, but needs review.)
750 -- Change make-terminal-frame to support specifying another tty.
752 (Done, new frame parameters: `tty' and `tty-type'.)
754 -- Implement support for reading from multiple terminals.
756 (Done, read_avail_input tries to read from each terminal, until one
757 succeeds. MULTI_KBOARD is not used. Secondary terminals don't send
760 (Update: They do, now.)
762 (Update2: After enabling X, they don't.)
764 -- other-frame should cycle through the frames on the `current'
767 (Done, by trivially modifiying next_frame and prev_frame.)
769 -- Support different terminal sizes.
773 -- Make sure terminal resizes are handled gracefully. (Could be
776 (Done. We don't get automatic SIGWINCH for additional ttys,
779 -- Extend emacsclient to automatically open a new tty when it connects
782 (Done. It's an ugly hack, needs more work.)
784 -- Redisplay must refresh the topmost frame on *all* terminals, not
785 just the initial terminal.
787 (Done, but introduced an ugly redisplay problems. Ugh.)
789 -- Fix redisplay problems.
791 (Done; it turned out that the entire Wcm structure must be moved
792 inside tty_output. Why didn't I catch this earlier?)
794 -- Provide a way for emacsclient to tell Emacs that the tty has been
797 (Done, simply forward the SIGWINCH signal.)
799 -- Each keypress should automatically select the frame corresponding
800 to the terminal that it was coming from. This means that Emacs
801 must know from which terminal the last keyboard event came from.
803 (Done, it was quite simple, the input event system already
804 supported multiple frames.)
806 -- Fix SIGIO issue with secondary terminals.
808 (Done, emacsclient signals Emacs after writing to the proxy pseudo
809 terminal. Note that this means that multi-tty does not work with
812 (Update: This is bullshit. There is a read_input_waiting function,
813 extend that somehow.)
815 (Update of update: The first update was not right either, extending
816 read_input_waiting was not necessary. Secondary ttys do seem to
817 send signals on input.)
819 (Update^3: Not any more.)
821 -- Make make-terminal-frame look up the `tty' and `tty-type' frame
822 parameters from the currently selected terminal before the global
827 -- Put all cached terminal escape sequences into struct tty_output.
828 Currently, they are still stored in global variables, so we don't
829 really support multiple terminal types.
831 (Done. It was not fun.)
833 -- Implement sane error handling after initialization. (Currently
834 emacs exits if you specify a bad terminal type.) The helpful error
835 messages must still be provided when Emacs starts.
839 -- Implement terminal deletion, i.e., deleting local frames, closing
840 the tty device and restoring its previous state without exiting
843 (Done, but at the moment only called when an error happens during
844 initialization. There is a memory corruption error around this
845 somewhere.) (Update: now it is fully enabled.)
847 -- Implement automatic deletion of terminals when the last frame on
848 that terminal is closed.
852 -- Restore tty screen after closing the terminal.
854 (Done, we do the same as Emacs 21.2 for all terminals.)
856 -- 'TERM=dumb src/emacs' does not restore the terminal state.
860 -- C-g should work on secondary terminals.
862 (Done, but the binding is not configurable.)
864 -- Deal with SIGHUP in Emacs and in emacsclient. (After this, the
865 server-frames may be removed from server.el.)
867 (Done, nothing to do. It seems that Emacs does not receive SIGHUP
868 from secondary ttys, which is actually a good thing.) (Update: I
869 think it would be a bad idea to remove server-frames.)
871 -- Change emacsclient/server.el to support the -t argument better,
872 i.e. automatically close the socket when the frame is closed.
874 (Seems to be working OK.)
876 -- Fix mysterious memory corruption error with tty deletion. To
877 trigger it, try the following shell command:
879 while true; do TERM=no-such-terminal-definition emacsclient -h; done
881 Emacs usually dumps core after a few dozen iterations. (The bug
882 seems to be related to the xfreeing or bzeroing of
883 tty_output.Wcm. Maybe there are outside references to struct Wcm?
884 Why were these vars collected into a struct before multi-tty
887 (Done. Whew. It turned out that the problem had nothing to do
888 with hypothetical external references to Wcm, or any other
889 tty_output component; it was simply that delete_tty closed the
890 filehandles of secondary ttys twice, resulting in fclose doubly
891 freeing memory. Utterly trivial matter. I love the C's memory
892 management, it puts hair on your chest.)
894 -- Support raw secondary terminals. (Note that SIGIO works only on
895 the controlling terminal.) Hint: extend read_input_waiting for
896 multiple ttys and hopefully this will be fixed.
898 (Done, it seems to have been working already for some time. It
899 seems F_SETOWN does work, after all. Not sure what made it fail
900 earlier, but it seems to be fixed (there were several changes
901 around request_sigio, maybe one of them did it).
902 read_input_waiting is only used in sys_select, don't change
903 it.) (Update: After adding X support, it's broken again.)
904 (Update^2: No it isn't.) :-)
906 -- Find out why does Emacs abort when it wants to close its
907 controlling tty. Hint: chan_process[] array. Hey, maybe
908 noninterrupt-IO would work, too? Update: no, there is no process
911 (Done. Added add/delete_keyboard_wait_descriptor to
912 term_init/delete_tty. The hint was right, in a way.)
914 -- Issue with SIGIO: it needs to be disabled during redisplay. See if
915 fcntl kernel behaviour could be emulated by emacsclient.
917 (Done. Simply disabled the SIGIO emulation hack in emacsclient.)
918 (Update: it was added back.) (Update^2: and removed again.)
920 -- server.el: There are issues with saving files in buffers of closed
921 clients. Try editing a file with emacsclient -f, and (without
922 saving it) do a delete-frame. The frame is closed without
923 question, and a surprising confirmation prompt appears in another
926 (Done. delete-frame now asks for confirmation if it still has
927 pending buffers, and modified buffers don't seem to be deleted.)
929 -- emacsclient.el, server.el: Handle eval or file open errors when
934 -- Make parts of struct tty_output accessible from Lisp. The device
935 name and the type is sufficient.
937 (Done, see frame-tty-name and frame-tty-type.)
939 -- Export delete_tty to the Lisp environment, for emacsclient.
941 (Done, see delete-tty.)
943 -- Get rid of the accessor macros in termchar.h, or define macros for
948 -- Move device-specific parameters (like costs) commonly used by
949 device backends to a common, device-dependent structure.
951 (Done. See struct display_method in termhooks.h.)
955 (Done. Well, it seems to be working.)
957 -- Allow simultaneous X and tty frames. (Handling input could be
958 tricky. Or maybe not.)
960 (Done. Allowed, that is. It is currently extremely unstable, to
961 the point of being unusable. The rif variable causes constant
962 core dumps. Handling input is indeed tricky.)
964 -- Rewrite multi-tty input in terms of MULTI_KBOARD.
966 (Done. In fact, there was no need to rewrite anything, I just
967 added a kboard member to tty_display_info, and initialized the
968 frame's kboard from there.)
970 -- Fix rif issue with X-tty combo sessions. IMHO the best thing to do
971 is to get rid of that global variable (and use the value value in
972 display_method, which is guaranteed to be correct).
974 (Done, did exactly that. Core dumps during combo sessions became
975 much rarer. In fact, I have not yet met a single one.)
977 -- Add multi-tty support to talk.el.
981 -- Clean up the source of emacsclient. It is a mess.
983 (Done, eliminated stupid proxy-pty kludge.)
985 -- Fix faces on tty frames during X-tty combo sessions. There is an
986 init_frame_faces call in init_sys_modes, see if there is a problem
989 (Done, there was a stupid mistake in
990 Ftty_supports_face_attributes_p. Colors are broken, though.)
992 -- C-x 5 2, C-x 5 o, C-x 5 0 on an emacsclient frame unexpectedly
993 exits emacsclient. This is a result of trying to be clever with
994 delete-frame-functions.
996 (Fixed, added delete-tty-after-functions, and changed server.el to
999 -- Something with (maybe) multi-keyboard support broke function keys
1000 and arrows on ttys during X+tty combo sessions. Debug this.
1002 (I can't reproduce it, maybe the terminal type was wrong.)
1004 -- Fix input from raw ttys (again).
1006 (Now it seems to work all right.)
1008 -- During an X-tty combo session, a (message "Hello") from a tty frame
1009 goes to the X frame. Fix this.
1011 (Done. There was a safeguard against writing to the initial
1012 terminal frame during bootstrap which prevented echo_area_display
1013 from working correctly on a tty frame during a combo session.)
1015 -- If there are no frames on its controlling terminal, Emacs should
1016 exit if the user presses C-c there.
1018 (Done, as far as possible. See the SIGTERM comment in
1019 interrupt_signal on why this seems to be impossible to solve this
1022 -- During an X session, Emacs seems to read from stdin. Also, Emacs
1023 fails to start without a controlling tty.
1025 (Fixed by replacing the troublesome termcap display with a dummy
1026 bootstrap display during bootstrap.
1028 -- Do tty output through struct display, like graphical display
1033 -- Define an output_initial value for output_method for the initial
1034 frame that is dumped with Emacs. Checking for this frame (e.g. in
1035 cmd_error_internal) is ugly.
1037 (Done, breaking interactive temacs.)
1039 -- The command `emacsclient -t -e '(delete-frame)'' fails to exit.
1043 -- frame-creation-function should always create a frame that is on the
1044 same display as the selected frame. Maybe frame-creation-function
1045 should simply be removed and make-frame changed to do the right
1048 (Done, with a nice hack. frame-creation-function is now frame-local.)
1050 -- Fix C-g on raw ttys.
1052 (Done. I disabled the interrupt/quit keys on all secondary
1053 terminals, so Emacs sees C-g as normal input. This looks like an
1054 overkill, because emacsclient has extra code to pass SIGINT to
1055 Emacs, so C-g should remain the interrupt/quit key on emacsclient
1056 frames. See the next entry why implementing this distinction would
1059 -- Make sure C-g goes to the right frame with ttys. This is hard, as
1060 SIGINT doesn't have a tty parameter. :-(
1062 (Done, the previous change fixes this as a pleasant side effect.)
1064 -- I have seen a case when Emacs with multiple ttys fell into a loop
1065 eating 100% of CPU time. Strace showed this loop:
1068 kill(30284, SIGIO) = 0
1069 --- SIGIO (I/O possible) @ 0 (0) ---
1070 ioctl(6, FIONREAD, [0]) = -1 EIO (Input/output error)
1071 ioctl(5, FIONREAD, [0]) = -1 EIO (Input/output error)
1072 ioctl(0, FIONREAD, [0]) = 0
1073 sigreturn() = ? (mask now [])
1074 gettimeofday({1072842297, 747760}, NULL) = 0
1075 gettimeofday({1072842297, 747806}, NULL) = 0
1076 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1077 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1078 gettimeofday({1072842297, 748245}, NULL) = 0
1080 I have seen something similar with a single X frame, but have not
1081 been able to reproduce it for debugging.
1083 Update: This may have been caused by checking for nread != 0
1084 instead of nread > 0 after calling read_socket_hook in
1087 (Fixed. This was caused by unconditionally including stdin in
1088 input_wait_mask in init_process. The select call in
1089 wait_reading_process_input always returned immediately, indicating
1090 that there is pending input from stdin, which nobody read.
1092 Note that the above strace output seems to be an unrelated but
1093 similar bug. I think that is now fixed.)
1095 -- Exiting Emacs while there are emacsclient frames doesn't restore the
1096 ttys to their default states.
1098 (This seems to be fixed by some previous change.)
1100 -- Allow opening an X session after -nw.
1104 -- Fix color handling during tty+X combo sessions. (It seems that tty
1105 sessions automatically convert the face colors to terminal colors
1106 when the face is loaded. This conversion must happen instead on
1107 the fly in write_glyphs, which might be problematic, as color
1108 approximation is currently done in lisp (term/tty-colors.el).)
1109 (Update: hm, colors seem to work fine if I start emacs with -nw and
1110 then create an X frame. Maybe it's just a small buglet somewhere.)
1112 (Seems to be fixed. The problem was in startup.el, it did not
1113 initialize tty colors when the initial window system was
1116 -- emacs -nw --eval '(y-or-n-p "Foobar")' segfaults. (Reported by
1119 (Fixed, there was a keyboard initialization problem.)
1121 -- Fix interactive use of temacs. There are face-related SEGVs, most
1122 likely because of changes in realize_default_face, realize_face.
1126 -- Don't exit Emacs when the last X connection fails during a
1127 multi-display session.
1131 -- Dan Nicolaescu noticed that starting emacsclient on the same
1132 terminal device that is the controlling tty of the Emacs process
1133 gives unexpected results.
1137 -- Istvan Marko reported that Emacs hang on ttys if it was started
1138 from a shell script.
1140 (Fixed. There was a bug in the multi-tty version of
1141 narrow_foreground_group. tcsetpgrp blocks if it is called from a
1142 process that is not in the same process group as the tty.)
1144 -- emacsclient -t from an Emacs term buffer does not work, complains
1145 about face problems. This can even lock up Emacs (if the recursive
1146 frame sets single_kboard). Update: the face problems are caused by
1147 bugs in term.el, not in multi-tty. The lockup is caused by
1148 single_kboard mode, and is not easily resolvable. The best thing to
1149 do is to simply refuse to create a tty frame of type `eterm'.
1151 (Fixed, changed emacsclient to check for TERM=eterm. The face
1152 complaints seem to be caused by bugs in term.el; they are not
1153 related to multi-tty.)
1155 -- Find out the best way to support suspending Emacs with multiple
1156 ttys. My guess: disable it on the controlling tty, but from other
1157 ttys pass it on to emacsclient somehow. (It is (I hope) trivial to
1158 extend emacsclient to handle suspend/resume. A `kill -STOP' almost
1161 (Done. I needed to play with signal handling and the server
1162 protocol a bit to make emacsclient behave as a normal UNIX program
1163 wrt foreground/background process groups.)
1165 -- There is a flicker during the startup of `emacs -nw'; it's as if
1166 the terminal is initialized, reset and then initialialized again.
1167 Debug this. (Hint: narrow_foreground_group is called twice during
1172 -- Robert Chassell has found serious copy-paste bugs with the
1173 multi-tty branch. There seem to be redisplay bugs while copying
1174 from X to a terminal frame. Copying accented characters do not
1177 (Patch-124 should fix this, by changing the interprogram-*-function
1178 variables to be frame-local, as suggested by Mark Plaksin
1179 (thanks!). I think that the redisplay bugs are in fact not bugs,
1180 but delays caused by single_kboard --> perhaps MULTI_KBOARD should
1183 -- frame-creation-function was removed, which might be a bad idea.
1184 Think up a compatible solution.
1186 (It was an internal interface that may be changed when necessary.)
1188 -- Change Lisp code not to (getenv "TERM"); use the `tty-type' frame
1189 parameter or the frame-tty-type function instead. (M-x tags-search
1190 "TERM" helps with this.) Update: Actually, all getenv invocations
1191 should be checked for multi-tty compatibility, and an interface
1192 must be implemented to get the remote client's environment.
1194 (Done. Only getenv calls in lisp/term/*.el were changed; other
1195 calls should be mostly left as they are.)
1197 -- Add an elaborate mechanism for display-local variables. (There are
1198 already a few of these; search for `terminal-local' in the Elisp
1201 (Not needed. Display-local variables could be emulated by
1202 frame-local variables.)
1204 -- Emacs assumes that all terminal frames have the same locale
1205 settings as Emacs itself. This may lead to bogus results in a
1206 multi-locale setup. (E.g., while logging in from a remote client
1207 with a different locale.)
1208 (Update after new bugreport by Friedrich Delgado Friedrichs:
1209 (at least) the structs terminal_coding and keyboard_coding in
1210 coding.c must be moved to struct display, and the Lisp interface
1211 [set-]keyboard-coding-system must be adapted for the change.)
1213 (Fixed. Emacs now uses the locale settings as seen by the
1214 emacsclient process for server tty frames.)
1215 (Update: Not really; Vlocale_coding_system is still global.)
1217 -- Make `struct display' accessible to Lisp programs. Accessor functions:
1219 (displayp OBJECT): Returns t if OBJECT is a display.
1220 => Implemented as display-live-p.
1222 (display-list): Returns list of currently active displays.
1225 (selected-display): Returns the display object of the selected frame.
1226 => Not strictly necessary, but implemented anyway.
1228 (frame-display FRAME): Returns the display object of FRAME.
1231 (display-frames DISPLAY): Returns a list of frames on DISPLAY.
1232 => Already implemented, see frames-on-display-list.
1234 (display-type DISPLAY): Returns the type of DISPLAY, as a
1235 symbol. (See `framep'.)
1236 => Implemented as display-live-p.
1238 (display-device DISPLAY): Returns the name of the device that
1239 DISPLAY uses, as a string. (E.g: "/dev/pts/16", or
1241 => Implemented as display-name.
1245 See next issue why this is necessary.
1247 (Update: The consensus on emacs-devel seems to be to do this via
1248 integer identifiers. That's fine by me.)
1252 -- The following needs to be supported:
1259 The cleanest way to solve this is to allow multiple displays on the
1260 same terminal device; each new emacsclient process should create
1261 its own display. As displays are currently identified by their
1262 device names, this is not possible until struct display becomes
1263 accessible as a Lisp-level object.
1267 -- Miles Bader suggests that C-x C-c on an emacsclient frame should
1268 only close the frame, not exit the entire Emacs session. Update:
1269 see above for a function that does this. Maybe this should be the
1272 (Done. This is the new default. No complaints so far.)
1274 -- Clean up the frame-local variable system. I think it's ugly and
1275 error-prone. But maybe I just haven't yet fully understood it.
1277 (Nothing to do. It doesn't seem ugly any more. It's rather clever.)
1279 -- Support multiple character locales. A version of
1280 `set-locale-environment' needs to be written for setting up
1281 display-local settings on ttys. I think calling
1282 set-display-table-and-terminal-coding-system and
1283 set-keyboard-coding-system would be enough. The language
1284 environment itself should remain a global setting.
1286 (Done, by an ugly hack.)
1288 -- The terminal customization files in term/*.el tend to change global
1289 parameters, which may confuse Emacs with multiple displays. Change
1290 them to tweak only frame-local settings, if possible. (They tend
1291 to call define-key to set function key sequences a lot.)
1293 (Done, by making `function-key-map' terminal-local (i.e., part of
1294 struct kboard). This has probably covered all the remaining problems.)
1296 -- Make `function-key-map' and `key-translation-map' terminal-local.
1300 -- Implement `terminal-local-value' and `set-terminal-local-value' to
1301 allow deterministic access to terminal local bindings. The
1302 encode-kb package can not set up `key-translation-map' without
1303 these. The terminal-local bindings seem to be independent of what
1308 -- xt-mouse.el needs to be adapted for multi-tty. It currently
1309 signals an error on kill-emacs under X, which prevents the user
1310 from exiting Emacs. (Reported by Mnemonikk on freenode.)
1315 -- Having {reset,init}_all_sys_modes in set-input-mode breaks arrow
1316 keys on non-selected terminals under screen, and sometimes on other
1317 terminal types as well. The other function keys continue to work
1318 fine. Sometimes faces on these screens become garbled.
1320 This only seems to affect displays that are of the same terminfo
1321 type as the selected one. Interestingly, in screen Emacs normally
1322 reports the up arrow key as `M-o A', but after the above SNAFU, it
1323 complains about `M-[ a'. UNIX ttys are a complete mystery to me,
1324 but it seems the reset-reinitialize cycle somehow leaves the
1325 non-selected terminals in a different state than usual. I have no
1326 idea how this could happen.
1328 Currently set-input-mode resets only the currently selected
1329 terminal, which seems to somehow work around the problem.
1333 Dan Nicolaescu <dann@ics.uci.edu> writes:
1334 > Some terminals have 2 modes for cursor keys: Application Mode where
1335 > the cursor keys transmit the codes defined in the terminfo entry, and
1336 > Cursor mode. Applications have to send the smkx and rmkx terminfo
1337 > strings to switch between the 2 modes. So Emacs (and emacsclient) have
1338 > to send smkx when initializing and rmkx when quitting (or on
1341 (I think patch-370 fixed this.)
1343 -- This long-standing bug (first reported by Han Boetes) seems to come
1344 and go all the time. It is time to track it down and fix it.
1349 # From another xterm:
1350 emacsclient -e '(y-or-n-p "Do you want me to crash? ")'
1351 # Notice how the answer ends up in the *scratch* buffer
1355 (Fixed in patch-414 after detailed analysis by Kalle Olavi Niemitalo.)
1357 -- normal-erase-is-backspace-mode in simple.el needs to be updated for
1358 multi-tty (rep. by Dan Waber). (The Delete key is broken on X
1361 (Fixed in patch-427.)
1363 -- I think keyboard-translate-table should be made terminal-local.
1365 (Done in patch-431.)
1367 -- The semantics of terminal-local variables are confusing; it is not
1368 clear what binding is in effect in any given time. See if
1369 current_kboard (or at least the terminal-local bindings exported to
1370 Lisp) might be changed to be tied to the selected frame instead.
1371 Currently, `function-key-map' and `key-translation-map' may be
1372 accessed reliably only using the hackish
1373 `(set-)terminal-local-value' functions.
1375 Perhaps there should be a difference between `last-command' &co.
1376 and these more conventional configuration variables.
1377 (E.g. `symbol-value' would use current_kboard to access
1378 `last-command', but SELECTED_FRAME()->display->kboard to get the
1379 value of `function-key-map'.
1381 (Fixed in patch-434.)
1383 -- If the first key pressed on a new tty terminal is a function key,
1384 it is not recognized correctly. May be related to the bug below.
1386 (Seems to have been fixed as a side effect of patch-434. "The bug
1387 below" was the set-input-mode madness.)
1389 (Update: this bug was fixed for good in patch-449. It was tracked
1390 down to a bug in `read_key_sequence': it failed to reinitialize its
1391 local function-key-map/key-translation-map references when it
1392 switched keyboards. I don't understand why did this bug only
1393 appear on brand new frames, though!)
1395 -- Disable connecting to a new X display when we use the GTK toolkit.
1397 (Disabled in patch-450.)
1399 -- Implement automatic forwarding of client environment variables to
1400 forked processes, as discussed on the multi-tty list. Terminal
1401 parameters are now accessible in C code, so the biggest obstacle is
1402 gone. The `getenv_internal' and `child_setup' functions in
1403 callproc.c must be changed to support the following variable:
1405 terminal-local-environment-variables is a variable defined in ...
1407 Enable or disable terminal-local environment variables.
1409 If set to t, `getenv', `setenv' and subprocess creation
1410 functions use the environment variables of the emacsclient
1411 process that created the selected frame, ignoring
1412 `process-environment'.
1414 If set to nil, Emacs uses `process-environment' and ignores
1415 the client environment.
1417 Otherwise, `terminal-local-environment-variables' should be a
1418 list of variable names (represented by Lisp strings) to look
1419 up in the client environment. The rest will come from
1420 `process-environment'.
1422 (Implemented in patch-461; `terminal-getenv', `terminal-setenv' and
1423 `with-terminal-environment' are now replaced by extensions to
1424 `getenv' and `setenv', and the new `local-environment-variables'
1427 (Updated in patch-465 to fix the semantics of let-binding
1428 `process-environment'. `process-environment' was changed to
1429 override all local/global environment variables, and a new variable
1430 `global-environment' was introduced to have `process-environment's
1433 (Updated in patch-466 to fix the case when two emacsclient sessions
1434 share the same terminal, but have different environment. The local
1435 environment lists are now stored as frame parameters, so the
1436 C-level terminal parameters are not strictly necessary any more.)
1438 -- `Fdelete_frame' is called from various critical places where it is
1439 not acceptable for the frame deletion to fail, e.g. from
1440 x_connection_closed after an X error. `Fdelete_frame' now protects
1441 against `delete-frame-functions' throwing an error and preventing a
1442 frame delete. (patch-475)
1444 -- Fix set-input-mode for multi-tty. It's a truly horrible interface;
1445 what if we'd blow it up into several separate functions (with a
1446 compatibility definition)?
1448 (Done. See `set-input-interrupt-mode', `set-output-flow-control',
1449 `set-input-meta-mode' and `set-quit-char'.) (patch-457)
1451 -- Let-binding `overriding-terminal-local-map' on a brand new frame
1452 does not seem to work correctly. (See `fancy-splash-screens'.)
1453 The keymap seems to be set up right, but events go to another
1454 terminal. Or is it `unread-command-events' that gets Emacs
1455 confused? Investigate.
1457 (Emacs was confused because a process filter entered
1458 `recursive-edit' while Emacs was reading input. I added support
1459 for this in the input system.) (patch-489)
1461 -- I smell something funny around pop_kboard's "deleted kboard" case.
1462 Determine what are the circumstances of this case, and fix any
1463 bug that comes to light.
1465 (It happens simply because single_kboard's terminal is sometimes
1466 deleted while executing a command on it, for example the one that
1467 kills the terminal. There was no bug here, but I rewrote the whole
1468 single_kboard mess anyway.) (patch-489)
1470 -- Understand Emacs's low-level input system (it's black magic) :-)
1471 What exactly does interrupt_input do? I tried to disable it for
1472 raw secondary tty support, but it does not seem to do anything
1473 useful. (Update: Look again. X unconditionally enables this, maybe
1474 that's why raw terminal support is broken again. I really do need
1475 to understand input.)
1476 (Update: I am starting to understand the read_key_sequence->read-char
1477 ->kbd_buffer_get_event->read_avail_input->read_socket_hook path. Yay!)
1479 (Update: OK, it all seems so easy now (NOT). Input could be done
1480 synchronously (with wait_reading_process_input), or asynchronously
1481 by SIGIO or polling (SIGALRM). C-g either sets the Vquit_flag,
1482 signals a 'quit condition (when immediate_quit), or throws to
1483 `getcjmp' when Emacs was waiting for input when the C-g event
1486 -- Replace wrong_kboard_jmpbuf with a special return value of
1487 read_char. It is absurd that we use setjmp/longjmp just to return
1488 to the immediate caller.
1490 (Done in patch-500.)
1492 -- `tool-bar-mode', `scroll-bar-mode', `menu-bar-mode' and
1493 'fringe-mode' are modes global to the entire Emacs session, not
1494 just a single frame or a single terminal. This means that their
1495 status sometimes differs from what's actually displayed on the
1496 screen. As a consequence, the Options | Show/Hide menu sometimes
1497 shows incorrect status, and you have to select an option twice for
1498 it to have any visible effect on the current frame.
1500 Change Emacs so that the status of the items in the Options |
1501 Show/Hide menu correspond to the current frame.
1503 (Done in patch-537.)
1505 -- The `default-directory' variable should somehow be set to the
1506 cwd of the emacsclient process when the user runs emacsclient
1507 without file arguments. Perhaps it is OK to just override the
1508 directory of the *scratch* buffer.
1510 (Done in patch-539.)
1512 -- The borders on tooltip frames on X are messed up. More
1513 specifically, the frame's internal border (internal-border-width
1514 frame parameter) is not filled with the correct background color.
1516 It seems the border contents are drawn onto by the
1517 update_single_window call in `x-show-tip'. After some debugging, I
1518 think the window's background color is not set up
1519 correctly---calling `x_clear_area' fills the specified area with
1520 black, not light yellow.
1522 (Done in patch-544. A background_pixel field was defined both in
1523 struct frame and struct x_output, and Emacs got confused between
1526 ;;; arch-tag: 8da1619e-2e79-41a8-9ac9-a0485daad17d