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 Eric Hanchrow <offby1@blarg.net>
46 IRIE Tetsuya <irie@t.email.ne.jp>
47 Yoshiaki Kasahara <kasahara@nc.kyushu-u.ac.jp>
48 Bas Kok <nekkobassu@yahoo.com>
49 Jurej Kubelka <Juraj.Kubelka@email.cz>
50 David Lichteblau <david@lichteblau.com>
51 Richard Lewis <rtf@jabble.com>
52 mace <mace@kirjakaapeli.lib.hel.fi>
53 Suresh Madhu <madhu@cs.unm.edu>
54 Xavier Mallard <zedek@gnu-rox.org>
55 Istvan Marko <mi-mtty@kismala.com>
56 Ted Morse <morse@ciholas.com>
57 Gergely Nagy <algernon@debian.org>
58 Dan Nicolaescu <dann@ics.uci.edu>
59 Kalle Olavi Niemitalo <kon@iki.fi>
60 Mark Plaksin <happy@mcplaksin.org>
61 Frank Ruell <stoerte@dreamwarrior.net>
62 Tom Schutzer-Weissmann <trmsw@yahoo.co.uk>
63 Joakim Verona <joakim@verona.se>
64 Dan Waber <dwaber@logolalia.com>
67 Richard Stallman was kind enough to review an earlier version of my
74 The multi-tty mailing list (discussion & bug reports):
76 Address: multi-tty@lists.fnord.hu
77 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty
78 Archive: http://news.gmane.org/gmane.emacs.multi-tty/
80 Commit notifications (read-only):
82 Address: multi-tty-commits@lists.fnord.hu
83 Signup: http://lists.fnord.hu/mailman/listinfo/multi-tty-commits
89 The branch is now very stable and almost full-featured. All of the
90 major problems have been fixed, only a few minor issues remain. (It
91 still needs to be ported to Windows/Mac/DOS, though.) Both multiple
92 tty device support and simultaneous X and tty frame support works
93 fine. Emacsclient has been extended to support opening new tty and X
94 frames. It has been changed to open new Emacs frames by default.
96 The multi-tty branch has been scheduled for inclusion in the next
97 major release of Emacs (version 23). I expect the merge into the
98 development trunk to occur sometime during next year (2006), after the
99 merge of the Unicode branch.
101 Tested on GNU/Linux, Solaris 8, FreeBSD and OpenBSD. Please let me
102 know if you succeed or fail to use it on other platforms---I'll have a
103 few tricky test cases for you.
107 * GTK support. If you compile your Emacs with the GTK
108 toolkit, some functionality of multi-tty will be lost. In
109 particular, you will not be able to work on multiple X
110 displays at once. Current releases of GTK have limitations
111 and bugs that prevent full-blown multi-display support in
112 Emacs. (GTK crashes when Emacs tries to disconnect from an
113 X server.) Use the Lucid toolkit if you want to see a
114 complete feature set.
116 * The single-kboard mode.
118 If your multi-tty Emacs session seems to be frozen, you
119 probably have a recursive editing session or a pending
120 minibuffer prompt (which is a kind of recursive editing) on
121 another display. To unfreeze your session, switch to that
122 display and complete the recursive edit, for example by
123 pressing C-] (`abort-recursive-edit').
125 I am sorry to say that currently there is no way to break
126 out of this "single-kboard mode" from a frozen display. If
127 you are unable to switch to the display that locks the
128 others (for example because it is on a remote computer),
129 then you can use emacsclient to break out of all recursive
132 emacsclient -e '(top-level)'
134 Note that this (perhaps) unintuitive behaviour is by design.
135 Single-kboard mode is required because of an intrinsic Emacs
136 limitation that is very hard to eliminate. (This limitation
137 is related to the single-threaded nature of Emacs.)
139 I plan to implement better user notification and support for
140 breaking out of single-kboard mode from locked displays.
142 * Mac, Windows and DOS support is broken, doesn't even
143 compile. Multiple display support will probably not provide
144 new Emacs features on these systems, but the multi-tty
145 branch changed a few low-level interfaces, and the
146 system-dependent source files need to be adapted
147 accordingly. The changes are mostly trivial, so almost
148 anyone can help, if only by compiling the branch and
149 reporting the compiler errors.
152 HOW TO GET THE BRANCH
153 ---------------------
155 The branch uses Bazaar 1 (http://bazaar.canonical.com) for version control.
157 Retrieving the latest version of the branch:
159 baz register-archive -f http://aszt.inf.elte.hu/~lorentey/mirror/arch/2004
160 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
162 This incantation uses an archive mirror that is hosted on a
163 high-bandwidth site. Please note that on average there is a two-hour
164 delay for commits to arrive on this mirror. My primary mirror is on the
165 low-bandwidth http://lorentey.hu/ site:
167 baz register-archive -f http://lorentey.hu/arch/2004/
168 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
170 This is "instantly" updated, but very slow from outside Hungary.
171 (By "instantly" I mean as soon as I connect the notebook I work on to
172 a network. It could take days.)
174 The Arch supermirror provides mirroring services for all public Arch
175 repositories. We have a mirror there, too, if you prefer.
177 baz register-archive -f http://mirrors.sourcecontrol.net/lorentey%40elte.hu--2004
178 baz get lorentey@elte.hu--2004/emacs--multi-tty <directory>
180 My GPG key id is 0FB27A3F; it is available from
181 hkp://wwwkeys.eu.pgp.net/, or from my homepage at
182 http://lorentey.hu/rolam/gpg.html)
184 Don't worry if the above checkout takes a few minutes to complete;
185 once you have a source tree, updating it to the latest revision will
186 be _much_ faster. Use the following command for the update:
190 You can find more information about Bazaar on
191 http://bazaar.canonical.com/. It's a distributed source control
192 system that is somewhat less broken than competing projects.
194 If you don't have Bazaar, the branch has a homepage from which you can
195 download conventional patches against Emacs CVS HEAD:
197 http://lorentey.hu/project/emacs.html
199 I suggest you use Bazaar whenever feasible.
205 If you run Debian, or a distribution based on Debian, you are welcome
206 to use our binary packages; put these lines in your /etc/apt/sources.list:
209 deb http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
210 deb-src http://aszt.inf.elte.hu/~lorentey/mirror/apt unstable multi-tty
212 Note that these packages are intended solely to provide an easy way to
213 test the new multi-tty features. They are not to be taken as Emacs
214 releases, and it's a mistake to expect robust operation or any kind of
215 timely support for them. Do not install them, unless you'd like to
216 have your editor crash on you.
222 The multi-tty branch is compiled the same way as Emacs itself:
224 make maintainer-clean # (If you have compiled Emacs before)
226 ./configure --without-gtk <your favourite options>
230 If you have strange compilation errors, they may be caused by old
231 *.elc files that are left over from an earlier bootstrap. The `make
232 maintainer-clean' target deletes them, so it is a good idea to run
233 that before reporting a bug. (Emacs requires a clean recompilation
234 after certain kinds of source code changes.)
239 To test the multi-tty branch, start up the Emacs server with the
245 and then (from a shell prompt on another terminal) start emacsclient
247 emacsclient -t /optional/file/names... (for a tty frame)
248 emacsclient /optional/file/names... (for an X frame)
250 (Make sure both emacs and emacsclient are multi-tty versions.)
251 You'll hopefully have two fully working, independent frames on
252 separate terminals. The new frame is closed automatically when you
253 finish editing the specified files (C-x #), but delete-frame (C-x 5 0)
254 also works. Of course, you can create frames on more than two tty
257 Creating new frames on the same tty with C-x 5 2 (make-frame-command)
258 works, and behaves the same way as in previous Emacs versions. If you
259 exit emacs, all terminals should be restored to their previous states.
261 This is work in progress, and probably full of bugs. It is a good
262 idea to run emacs from gdb, so that you'll have a live instance to
263 debug if something goes wrong. Please send me your bug reports on our
264 mailing list: multi-tty@lists.fnord.hu
269 I think the best way to use the new Emacs is to have it running inside
270 a disconnected GNU screen session, and always use emacsclient for
271 normal work. One advantage of this is that not a single keystroke of
272 your work will be lost if the display device that you are using
273 crashes, or the network connection times out, or whatever. (I had an
274 extremely unstable X server for some time while I was developing these
275 patches, and running Emacs this way has saved me a number of M-x
276 recover-session invocations.)
278 I use the following two bash scripts to handle my Emacs sessions:
280 -------------------------------------------------------connect-emacs--
282 # Usage: connect-emacs <name> <args>...
284 # Connects to the Emacs instance called NAME. Starts up the instance
285 # if it is not already running. The rest of the arguments are passed
291 if [ -z "$name" ]; then
292 echo "Usage: connect_emacs <name> <args>..." >&2
295 preload-emacs "$name" wait
296 /usr/bin/emacsclient.emacs-multi-tty -s "$name" "$@"
297 ----------------------------------------------------------------------
299 -------------------------------------------------------preload-emacs--
301 # Usage: preload-emacs <name> [<waitp>]
303 # Preloads the Emacs instance called NAME in a detached screen
304 # session. Does nothing if the instance is already running. If WAITP
305 # is non-empty, the function waits until the server starts up and
306 # creates its socket; otherwise it returns immediately.
310 screendir="/var/run/screen/S-$USER"
311 serverdir="/tmp/emacs$UID"
312 emacs=/usr/bin/emacs-multi-tty # Or wherever you installed your multi-tty Emacs
314 if [ -z "$name" ]; then
315 echo "Usage: preload_emacs <name> [<waitp>]" >&2
319 if [ ! -e "$screendir"/*."$name" ]; then
320 if [ -e "$serverdir/$name" ]; then
321 # Delete leftover socket (for the wait option)
322 rm "$serverdir/$name"
324 screen -dmS "$name" "$emacs" -nw --eval "(setq server-name \"$name\")" -f server-start
326 if [ ! -z "$waitp" ]; then
327 while [ ! -e "$serverdir/$name" ]; do sleep 0.1; done
329 ----------------------------------------------------------------------
331 I have the following in my profile to have two instances automatically
332 preloaded for editing and email:
337 It is useful to set up short aliases for connect-emacs. I use the
340 alias edit="connect-emacs editor"
342 alias et="connect-emacs editor -t"
343 alias gnus="connect-emacs gnus"
355 For the NEWS file: (Needs much, much work)
357 ** Support for multiple terminal devices and simultaneous graphical
358 and tty frames has been added. You can test for the presence of
359 this feature in your Lisp code by testing for the `multi-tty'
362 *** The `window-system' variable has been made frame-local. The new
363 `initial-window-system' variable contains the `window-system'
364 value for the first frame.
366 *** You can specify a terminal device (`tty' parameter) and a terminal
367 type (`tty-type' parameter) to `make-terminal-frame'.
369 *** The new function `make-frame-on-tty' allows you to create a new
370 frame on another tty device interactively.
372 *** The function `make-frame-on-display' now works during a tty
373 session, and `make-frame-on-tty' works during a graphical session.
375 *** Emacsclient has been extended to support opening a new terminal
376 frame. Its behaviour has been changed to open a new Emacs frame by
377 default. Use the -c option to get the old behavior of opening
378 files in the currently selected Emacs frame.
380 *** C-z now invokes `suspend-frame', C-x C-c now invokes
381 `save-buffers-kill-terminal'.
383 *** New functions: frame-tty-name, frame-tty-type, delete-tty,
384 suspend-tty, resume-tty, terminal-id, terminal-parameters,
385 terminal-parameter, set-terminal-parameter,
386 modify-terminal-parameters, environment, let-environment
388 *** New variables: local-key-translation-map, local-function-key-map
390 *** The `keyboard-translate-table' variable and the terminal and
391 keyboard coding systems have been made terminal-local.
393 *** In addition to the global key-translation-map and
394 function-key-map, Emacs has terminal-local
395 local-key-translation-map and local-function-key-map variables,
396 and uses them instead of the global keymaps to set up translations
397 and function key sequences relevant to a specific terminal device.
399 *** talk.el has been extended for multiple tty support.
403 (The rest of this file consists of my development notes and as such it
404 is probably not very interesting for anyone else.)
409 ** See if `tty-defined-color-alist' needs to be terminal-local.
410 Dan says it should be, so convert it.
412 ** emacsclient -t on the console does not work after su. You have to
413 use non-root accounts or start as root to see this.
419 *ERROR*: Could not open file: /dev/tty1
421 The tty can be opened as /dev/tty by emacsclient, but not by Emacs.
422 This seems to be a serious problem. Currently my only idea is to
423 bring back the ugly pty proxy hack from the initial versions of
424 multi-tty. Suggestions would be appreciated.
426 Update: we could change emacsclient to pass its open file
427 descriptor to the Emacs process. Unfortunately, this requires a
428 new Lisp-level Emacs API, and as file descriptors are not otherwise
429 exported to Lisp, this approach seems at least as ugly as the pty
432 ** lisp/vc.el depends on the terminal type during load time.
433 `vc-annotate-color-map' is one example that needs to be fixed.
435 ** Understand how `quit_throw_to_read_char' works, and fix any bugs
438 ** See if getcjmp can be eliminated somehow. Why does Emacs allow
439 asynchronous input processing while it's reading input anyway?
441 ** `delete-frame' events are handled by `special-event-map'
442 immediately when read by `read_char'. This is fine but it prevents
443 higher-level keymaps from binding that event to get notified of the
446 Sometimes it would be useful for Lisp code to be notified of frame
447 deletions after they have happened, usually because they want to
448 clean up after the deleted frame. Not all frame-local states can
449 be stored as a frame parameter. For example,
450 `display-splash-screen' uses `recursive-edit' with a special keymap
451 override to create its buffer---and it leads to all kinds of
452 nastiness if Emacs stays in this recursive edit mode after the
453 frame containing the splash screen is deleted. Basically, the
454 splash-screen implementation wants to throw out of the recursive
455 edit when the frame is deleted; however, it is not legal to throw
456 from `delete-frame-functions' because `delete-frame' must not fail.
457 (Introducing `delete-frame-after-functions' would not help either
458 because `delete-frame' may not fail at that time either.)
460 Currently `fancy-splash-screens' installs a
461 `delete-frame-functions' hook that sets up a timer to exit the
462 recursive edit. This is an adequate solution, but it would perhaps
463 be better to have something like a `frame-deleted' event that could
464 be bound in the normal way.
466 ** Trouble: `setenv' doesn't actually set environment variables in the
467 Emacs process. This defeats the purpose of the elaborate
468 `server-with-environment' magic around the `tgetent' call in
471 ** (Possibly) create hooks in struct device for creating frames on a
472 specific terminal, and eliminate the hackish terminal-related frame
473 parameters (display, tty, tty-type).
478 ** Decide whether to keep the C implementation of terminal parameters,
479 or revert to the previous, purely Lisp code. It turned out that
480 local environments do not need terminal parameters after all.
482 ** Move Fsend_string_to_terminal to term.c, and declare get_named_tty
483 as static, removing it from dispextern.h.
484 Move fatal to emacs.c and declare it somewhere.
486 ** Search for `suspend-emacs' references and replace them with
487 `suspend-frame', if necessary. Ditto for `save-buffers-kill-emacs'
488 vs. `save-buffers-kill-display'.
490 ** Emacs crashes when a tty frame is resized so that there is no space
491 for all its windows. (Tom Schutzer-Weissmann)
493 ** Report GTK multi-display problems to GTK maintainers. For extra
496 Currently you can not connect to new X displays when you compile
497 Emacs with GTK support. If you want to play around with GTK
498 multidisplay (and don't mind core dumps), you can edit src/config.h
499 and define HAVE_GTK_MULTIDISPLAY there by hand.
501 http://bugzilla.gnome.org/show_bug.cgi?id=85715
503 Update: Han reports that GTK+ version 2.8.9 almost gets display
504 disconnects right. GTK will probably be fully fixed by the time
505 multi-tty gets into the trunk.
507 Update: I am still having problems with GTK+ 2.8.10. I have the
508 impression that the various multidisplay fixes will only get
509 released in GTK+ 2.10.
511 ** Audit `face-valid-attribute-values' usage in customize and
512 elsewhere. Its return value depends on the current window system.
513 Replace static initializers using it with runtime functions. For
514 example, custom's buttons are broken on non-initial device types.
516 ** Possibly turn off the double C-g feature when there is an X frame.
517 C.f. (emacs)Emergency Escape.
519 ** frames-on-display-list should also accept frames.
521 ** Consider the `tty-type' frame parameter and the `display-tty-type'
522 function. They serve the exact same purpose. I think it may be
523 a good idea to eliminate one of them, preferably `tty-type'.
525 ** The handling of lisp/term/*.el, and frame creation in general, is a
526 big, big mess. How come the terminal-specific file is loaded by
527 tty-create-frame-with-faces? I don't think it is necessary to load
528 these files for each frame; once per terminal should be enough.
529 Update: lisp/term/*.el is not loaded repeatedly anymore, but
530 faces.el still needs to be cleaned up.
532 ** Fix frame-set-background-mode in this branch. It was recently
533 changed in CVS, and frame.el in multi-tty has not yet been adapted
534 for the changes. (It needs to look at
535 default-frame-background-mode.) (Update: maybe it is fixed now;
536 needs testing.) (Note that the byte compiler has this to say about
539 term/rxvt.el:309:17:Warning: assignment to free variable
540 `default-frame-background-mode'
542 ** I think `(set-)terminal-local-value' and the terminal parameter
543 mechanism should be integrated into a single framework.
545 (Update: `(set-)terminal-local-value' is now eliminated, but the
546 terminal-local variables should still be accessible as terminal
547 parameters. This also applies to `display-name' and similar
550 ** Add the following hooks: after-delete-frame-hook (for server.el,
551 instead of delete-frame-functions),
552 after-delete-terminal-functions, after-create-terminal-functions.
554 ** BULK RENAME: The `display-' prefix of new Lisp-level functions
555 conflicts with stuff like `display-time-mode'. Use `device-'
556 or `terminal-' instead. I think I prefer `terminal-'.
558 It turns out that most of the offending Lisp functions were defined
559 in the trunk. Therefore, compatibility aliases should be defined
560 for the following names:
562 display-color-cells terminal-color-cells
563 display-color-p terminal-color-p
564 display-graphic-p terminal-graphic-p
565 display-grayscale-p terminal-grayscale-p
566 display-images-p terminal-images-p
567 display-mm-height terminal-mm-height
568 display-mm-width terminal-mm-width
569 display-mouse-p terminal-mouse-p
570 display-multi-font-p terminal-multi-font-p
571 display-multi-frame-p terminal-multi-frame-p
572 display-pixel-height terminal-pixel-height
573 display-pixel-width terminal-pixel-width
574 display-pixels-per-inch terminal-pixels-per-inch
575 display-planes terminal-planes
576 display-popup-menus-p terminal-popup-menus-p
577 display-save-under terminal-save-under
578 display-screens terminal-screens
579 display-supports-face-attributes-p terminal-supports-face-attributes-p
580 display-visual-class terminal-visual-class
581 framep-on-display framep-on-terminal
582 frames-on-display-list frames-on-terminal-list
584 The following functions were introduced in the multi-tty branch, and
585 were renamed without aliases:
587 delete-display delete-terminal
588 display-controlling-tty-p controlling-tty-p
589 display-list terminal-list
590 display-live-p terminal-live-p
591 display-name terminal-name
592 display-tty-type tty-type
593 frame-display frame-terminal
594 selected-display selected-terminal
596 ** The single-keyboard mode of MULTI_KBOARD is extremely confusing
597 sometimes; Emacs does not respond to stimuli from other keyboards.
598 At least a beep or a message would be important, if the single-mode
599 is still required to prevent interference. (Reported by Dan
602 Update: selecting a region with the mouse enables single_kboard
603 under X. This is very confusing.
605 Update: After discussions with Richard Stallman, this will be
606 resolved by having locked displays warn the user to wait, and
607 introducing a complex protocol to remotely bail out of
608 single-kboard mode by pressing C-g.
610 Update: Warning the user is not trivial to implement, as Emacs has
611 only one echo area, shared by all frames. Ideally the warning
612 should not be displayed on the display that is locking the others.
613 Perhaps the high probability of user confusion caused by
614 single_kboard mode deserves a special case in the display code.
615 Alternatively, it might be good enough to signal single_kboard mode
616 by changing the modelines or some other frame-local display element
617 on the locked out displays.
619 Update: In fact struct kboard does have an echo_string slot.
621 ** The session management module is prone to crashes when the X
622 connection is closed and then later I try to connect to a new X
625 #0 0xb7ebc806 in SmcGetIceConnection () from /usr/X11R6/lib/libSM.so.6
626 #1 0x080e6641 in x_session_check_input (bufp=0xbf86c9c0) at xsmfns.c:144
627 #2 0x080d3bbc in XTread_socket (device=0xa722ff8, expected=1, hold_quit=0xbf86ca90) at xterm.c:7037
628 #3 0x080fa404 in read_avail_input (expected=1) at keyboard.c:6696
629 #4 0x080fa4ca in handle_async_input () at keyboard.c:6900
630 #5 0x080d51fa in x_term_init (display_name=162628899, xrm_option=0x0, resource_name=0x857068c "emacs") at xterm.c:10622
631 #6 0x080d920e in x_display_info_for_name (name=162628899) at xfns.c:3975
632 #7 0x080d92f9 in check_x_display_info (object=1) at xfns.c:274
633 #8 0x080d97b8 in Fx_create_frame (parms=151221485) at xfns.c:3016
634 #9 0x0815bf72 in Ffuncall (nargs=2, args=0xbf86ceec) at eval.c:2851
636 I installed a workaround to prevent this. The X session manager is
637 only contacted when the very first display in the Emacs session is
638 an X display. Also, x_delete_display() on this display aborts
639 session management, and XTread_socket only calls
640 x_session_check_input when it is called for the display that the
641 session was opened on. While this does not really fix the bug, it
642 makes it much less frequent, because session manager support will
643 not normally be enabled when Emacs can survive the shutdown of the
646 See if xsmfns.c should be updated.
648 ** Hunt down display-related functions in frame.el and extend them all
649 to accept display ids.
651 ** rif->flush_display_optional (NULL) calls should be replaced by a
654 ** The set-locale-environment hack (adding the DISPLAY option) should
655 be replaced with a clean design.
657 ** standard-display-table should be display-local.
658 standard-display-european should be display-local.
660 ** With iswitchb-default-method set to 'always-frame, only frames on
661 the current display should be considered. This might involve
662 extending `get-buffer-window'.
664 ** Have a look at Vlocale_coding_system. Seems like it would be a
665 tedious job to localize it, although most references use it for
666 interfacing with libc and are therefore OK with the global
669 Exceptions found so far: x-select-text and
670 x-cut-buffer-or-selection-value.
672 ** Have a look at fatal_error_hook.
674 ** Have a look at set_frame_matrix_frame.
676 ** Check if we got term-setup-hook right.
678 ** I think tip_frame should be display-local.
680 ** Check display reference count handling in x_create_tip_frame.
682 ** make-frame does not correctly handle extra parameters in its
685 (frame-parameter (make-frame (list (cons 'foobar 42))) 'foobar)
688 (This is likely an error in the CVS trunk.)
690 ** Dan Nicolaescu suggests that -nw should be added as an alias for -t
691 in emacsclient. Good idea. (Alas, implementing this is not
692 trivial, getopt_long does not seem to support two-letter ``short''
693 options. Patches are welcome.)
695 ** Mark Plaksin suggests that emacsclient should accept the same
696 X-related command-line arguments as Emacs. Most of the X-related
697 argument-handling is done in Lisp, so this should be quite easy to
700 ** Gergely Nagy suggests that C-x # should only kill the current
701 frame, not any other emacsclient frame that may have the same file
702 opened for editing. I think I agree with him.
704 ** Very strange bug: visible-bell does not work on secondary
705 terminals in xterm and konsole. The screen does flicker a bit,
706 but it's so quick it isn't noticable.
708 (Update: This is probably some problem with padding or whatnot on
709 the secondary terminals.)
711 ** Move baud_rate to struct display.
713 ** Implement support for starting an interactive Emacs session without
714 an initial frame. (The user would connect to it and open frames
715 later, with emacsclient.)
717 ** Fix Mac support (I can't do this entirely myself). Note that the
718 current state of Mac-specific source files in the multi-tty tree
719 are not useful; before starting work on Mac support, revert to
720 pristine, pre-multi-tty versions.
722 ** Fix W32 support (I can't do this entirely myself). Note that the
723 current state of W32-specific source files in the multi-tty tree
724 are not useful; before starting work on W32 support, revert to
725 pristine, pre-multi-tty versions.
727 ** Fix DOS support (I can't do this entirely myself). Note that the
728 current state of DOS-specific source files in the multi-tty tree
729 are not useful; before starting work on DOS support, revert to
730 pristine, pre-multi-tty versions.
732 ** Do a grep on XXX and ?? for more issues.
734 ** flow-ctrl.el must be updated.
736 ** Fix stuff_char for multi-tty. Doesn't seem to be of high priority.
741 (ex-TODO items with explanations.)
743 -- Introduce a new struct for terminal devices.
745 (Done, see struct tty_output. The list of members is not yet
748 -- Change the bootstrap procedure to initialize tty_list.
750 (Done, but needs review.)
752 -- Change make-terminal-frame to support specifying another tty.
754 (Done, new frame parameters: `tty' and `tty-type'.)
756 -- Implement support for reading from multiple terminals.
758 (Done, read_avail_input tries to read from each terminal, until one
759 succeeds. MULTI_KBOARD is not used. Secondary terminals don't send
762 (Update: They do, now.)
764 (Update2: After enabling X, they don't.)
766 -- other-frame should cycle through the frames on the `current'
769 (Done, by trivially modifiying next_frame and prev_frame.)
771 -- Support different terminal sizes.
775 -- Make sure terminal resizes are handled gracefully. (Could be
778 (Done. We don't get automatic SIGWINCH for additional ttys,
781 -- Extend emacsclient to automatically open a new tty when it connects
784 (Done. It's an ugly hack, needs more work.)
786 -- Redisplay must refresh the topmost frame on *all* terminals, not
787 just the initial terminal.
789 (Done, but introduced an ugly redisplay problems. Ugh.)
791 -- Fix redisplay problems.
793 (Done; it turned out that the entire Wcm structure must be moved
794 inside tty_output. Why didn't I catch this earlier?)
796 -- Provide a way for emacsclient to tell Emacs that the tty has been
799 (Done, simply forward the SIGWINCH signal.)
801 -- Each keypress should automatically select the frame corresponding
802 to the terminal that it was coming from. This means that Emacs
803 must know from which terminal the last keyboard event came from.
805 (Done, it was quite simple, the input event system already
806 supported multiple frames.)
808 -- Fix SIGIO issue with secondary terminals.
810 (Done, emacsclient signals Emacs after writing to the proxy pseudo
811 terminal. Note that this means that multi-tty does not work with
814 (Update: This is bullshit. There is a read_input_waiting function,
815 extend that somehow.)
817 (Update of update: The first update was not right either, extending
818 read_input_waiting was not necessary. Secondary ttys do seem to
819 send signals on input.)
821 (Update^3: Not any more.)
823 -- Make make-terminal-frame look up the `tty' and `tty-type' frame
824 parameters from the currently selected terminal before the global
829 -- Put all cached terminal escape sequences into struct tty_output.
830 Currently, they are still stored in global variables, so we don't
831 really support multiple terminal types.
833 (Done. It was not fun.)
835 -- Implement sane error handling after initialization. (Currently
836 emacs exits if you specify a bad terminal type.) The helpful error
837 messages must still be provided when Emacs starts.
841 -- Implement terminal deletion, i.e., deleting local frames, closing
842 the tty device and restoring its previous state without exiting
845 (Done, but at the moment only called when an error happens during
846 initialization. There is a memory corruption error around this
847 somewhere.) (Update: now it is fully enabled.)
849 -- Implement automatic deletion of terminals when the last frame on
850 that terminal is closed.
854 -- Restore tty screen after closing the terminal.
856 (Done, we do the same as Emacs 21.2 for all terminals.)
858 -- 'TERM=dumb src/emacs' does not restore the terminal state.
862 -- C-g should work on secondary terminals.
864 (Done, but the binding is not configurable.)
866 -- Deal with SIGHUP in Emacs and in emacsclient. (After this, the
867 server-frames may be removed from server.el.)
869 (Done, nothing to do. It seems that Emacs does not receive SIGHUP
870 from secondary ttys, which is actually a good thing.) (Update: I
871 think it would be a bad idea to remove server-frames.)
873 -- Change emacsclient/server.el to support the -t argument better,
874 i.e. automatically close the socket when the frame is closed.
876 (Seems to be working OK.)
878 -- Fix mysterious memory corruption error with tty deletion. To
879 trigger it, try the following shell command:
881 while true; do TERM=no-such-terminal-definition emacsclient -h; done
883 Emacs usually dumps core after a few dozen iterations. (The bug
884 seems to be related to the xfreeing or bzeroing of
885 tty_output.Wcm. Maybe there are outside references to struct Wcm?
886 Why were these vars collected into a struct before multi-tty
889 (Done. Whew. It turned out that the problem had nothing to do
890 with hypothetical external references to Wcm, or any other
891 tty_output component; it was simply that delete_tty closed the
892 filehandles of secondary ttys twice, resulting in fclose doubly
893 freeing memory. Utterly trivial matter. I love the C's memory
894 management, it puts hair on your chest.)
896 -- Support raw secondary terminals. (Note that SIGIO works only on
897 the controlling terminal.) Hint: extend read_input_waiting for
898 multiple ttys and hopefully this will be fixed.
900 (Done, it seems to have been working already for some time. It
901 seems F_SETOWN does work, after all. Not sure what made it fail
902 earlier, but it seems to be fixed (there were several changes
903 around request_sigio, maybe one of them did it).
904 read_input_waiting is only used in sys_select, don't change
905 it.) (Update: After adding X support, it's broken again.)
906 (Update^2: No it isn't.) :-)
908 -- Find out why does Emacs abort when it wants to close its
909 controlling tty. Hint: chan_process[] array. Hey, maybe
910 noninterrupt-IO would work, too? Update: no, there is no process
913 (Done. Added add/delete_keyboard_wait_descriptor to
914 term_init/delete_tty. The hint was right, in a way.)
916 -- Issue with SIGIO: it needs to be disabled during redisplay. See if
917 fcntl kernel behaviour could be emulated by emacsclient.
919 (Done. Simply disabled the SIGIO emulation hack in emacsclient.)
920 (Update: it was added back.) (Update^2: and removed again.)
922 -- server.el: There are issues with saving files in buffers of closed
923 clients. Try editing a file with emacsclient -f, and (without
924 saving it) do a delete-frame. The frame is closed without
925 question, and a surprising confirmation prompt appears in another
928 (Done. delete-frame now asks for confirmation if it still has
929 pending buffers, and modified buffers don't seem to be deleted.)
931 -- emacsclient.el, server.el: Handle eval or file open errors when
936 -- Make parts of struct tty_output accessible from Lisp. The device
937 name and the type is sufficient.
939 (Done, see frame-tty-name and frame-tty-type.)
941 -- Export delete_tty to the Lisp environment, for emacsclient.
943 (Done, see delete-tty.)
945 -- Get rid of the accessor macros in termchar.h, or define macros for
950 -- Move device-specific parameters (like costs) commonly used by
951 device backends to a common, device-dependent structure.
953 (Done. See struct display_method in termhooks.h.)
957 (Done. Well, it seems to be working.)
959 -- Allow simultaneous X and tty frames. (Handling input could be
960 tricky. Or maybe not.)
962 (Done. Allowed, that is. It is currently extremely unstable, to
963 the point of being unusable. The rif variable causes constant
964 core dumps. Handling input is indeed tricky.)
966 -- Rewrite multi-tty input in terms of MULTI_KBOARD.
968 (Done. In fact, there was no need to rewrite anything, I just
969 added a kboard member to tty_display_info, and initialized the
970 frame's kboard from there.)
972 -- Fix rif issue with X-tty combo sessions. IMHO the best thing to do
973 is to get rid of that global variable (and use the value value in
974 display_method, which is guaranteed to be correct).
976 (Done, did exactly that. Core dumps during combo sessions became
977 much rarer. In fact, I have not yet met a single one.)
979 -- Add multi-tty support to talk.el.
983 -- Clean up the source of emacsclient. It is a mess.
985 (Done, eliminated stupid proxy-pty kludge.)
987 -- Fix faces on tty frames during X-tty combo sessions. There is an
988 init_frame_faces call in init_sys_modes, see if there is a problem
991 (Done, there was a stupid mistake in
992 Ftty_supports_face_attributes_p. Colors are broken, though.)
994 -- C-x 5 2, C-x 5 o, C-x 5 0 on an emacsclient frame unexpectedly
995 exits emacsclient. This is a result of trying to be clever with
996 delete-frame-functions.
998 (Fixed, added delete-tty-after-functions, and changed server.el to
1001 -- Something with (maybe) multi-keyboard support broke function keys
1002 and arrows on ttys during X+tty combo sessions. Debug this.
1004 (I can't reproduce it, maybe the terminal type was wrong.)
1006 -- Fix input from raw ttys (again).
1008 (Now it seems to work all right.)
1010 -- During an X-tty combo session, a (message "Hello") from a tty frame
1011 goes to the X frame. Fix this.
1013 (Done. There was a safeguard against writing to the initial
1014 terminal frame during bootstrap which prevented echo_area_display
1015 from working correctly on a tty frame during a combo session.)
1017 -- If there are no frames on its controlling terminal, Emacs should
1018 exit if the user presses C-c there.
1020 (Done, as far as possible. See the SIGTERM comment in
1021 interrupt_signal on why this seems to be impossible to solve this
1024 -- During an X session, Emacs seems to read from stdin. Also, Emacs
1025 fails to start without a controlling tty.
1027 (Fixed by replacing the troublesome termcap display with a dummy
1028 bootstrap display during bootstrap.
1030 -- Do tty output through struct display, like graphical display
1035 -- Define an output_initial value for output_method for the initial
1036 frame that is dumped with Emacs. Checking for this frame (e.g. in
1037 cmd_error_internal) is ugly.
1039 (Done, breaking interactive temacs.)
1041 -- The command `emacsclient -t -e '(delete-frame)'' fails to exit.
1045 -- frame-creation-function should always create a frame that is on the
1046 same display as the selected frame. Maybe frame-creation-function
1047 should simply be removed and make-frame changed to do the right
1050 (Done, with a nice hack. frame-creation-function is now frame-local.)
1052 -- Fix C-g on raw ttys.
1054 (Done. I disabled the interrupt/quit keys on all secondary
1055 terminals, so Emacs sees C-g as normal input. This looks like an
1056 overkill, because emacsclient has extra code to pass SIGINT to
1057 Emacs, so C-g should remain the interrupt/quit key on emacsclient
1058 frames. See the next entry why implementing this distinction would
1061 -- Make sure C-g goes to the right frame with ttys. This is hard, as
1062 SIGINT doesn't have a tty parameter. :-(
1064 (Done, the previous change fixes this as a pleasant side effect.)
1066 -- I have seen a case when Emacs with multiple ttys fell into a loop
1067 eating 100% of CPU time. Strace showed this loop:
1070 kill(30284, SIGIO) = 0
1071 --- SIGIO (I/O possible) @ 0 (0) ---
1072 ioctl(6, FIONREAD, [0]) = -1 EIO (Input/output error)
1073 ioctl(5, FIONREAD, [0]) = -1 EIO (Input/output error)
1074 ioctl(0, FIONREAD, [0]) = 0
1075 sigreturn() = ? (mask now [])
1076 gettimeofday({1072842297, 747760}, NULL) = 0
1077 gettimeofday({1072842297, 747806}, NULL) = 0
1078 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1079 select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
1080 gettimeofday({1072842297, 748245}, NULL) = 0
1082 I have seen something similar with a single X frame, but have not
1083 been able to reproduce it for debugging.
1085 Update: This may have been caused by checking for nread != 0
1086 instead of nread > 0 after calling read_socket_hook in
1089 (Fixed. This was caused by unconditionally including stdin in
1090 input_wait_mask in init_process. The select call in
1091 wait_reading_process_input always returned immediately, indicating
1092 that there is pending input from stdin, which nobody read.
1094 Note that the above strace output seems to be an unrelated but
1095 similar bug. I think that is now fixed.)
1097 -- Exiting Emacs while there are emacsclient frames doesn't restore the
1098 ttys to their default states.
1100 (This seems to be fixed by some previous change.)
1102 -- Allow opening an X session after -nw.
1106 -- Fix color handling during tty+X combo sessions. (It seems that tty
1107 sessions automatically convert the face colors to terminal colors
1108 when the face is loaded. This conversion must happen instead on
1109 the fly in write_glyphs, which might be problematic, as color
1110 approximation is currently done in lisp (term/tty-colors.el).)
1111 (Update: hm, colors seem to work fine if I start emacs with -nw and
1112 then create an X frame. Maybe it's just a small buglet somewhere.)
1114 (Seems to be fixed. The problem was in startup.el, it did not
1115 initialize tty colors when the initial window system was
1118 -- emacs -nw --eval '(y-or-n-p "Foobar")' segfaults. (Reported by
1121 (Fixed, there was a keyboard initialization problem.)
1123 -- Fix interactive use of temacs. There are face-related SEGVs, most
1124 likely because of changes in realize_default_face, realize_face.
1128 -- Don't exit Emacs when the last X connection fails during a
1129 multi-display session.
1133 -- Dan Nicolaescu noticed that starting emacsclient on the same
1134 terminal device that is the controlling tty of the Emacs process
1135 gives unexpected results.
1139 -- Istvan Marko reported that Emacs hang on ttys if it was started
1140 from a shell script.
1142 (Fixed. There was a bug in the multi-tty version of
1143 narrow_foreground_group. tcsetpgrp blocks if it is called from a
1144 process that is not in the same process group as the tty.)
1146 -- emacsclient -t from an Emacs term buffer does not work, complains
1147 about face problems. This can even lock up Emacs (if the recursive
1148 frame sets single_kboard). Update: the face problems are caused by
1149 bugs in term.el, not in multi-tty. The lockup is caused by
1150 single_kboard mode, and is not easily resolvable. The best thing to
1151 do is to simply refuse to create a tty frame of type `eterm'.
1153 (Fixed, changed emacsclient to check for TERM=eterm. The face
1154 complaints seem to be caused by bugs in term.el; they are not
1155 related to multi-tty.)
1157 -- Find out the best way to support suspending Emacs with multiple
1158 ttys. My guess: disable it on the controlling tty, but from other
1159 ttys pass it on to emacsclient somehow. (It is (I hope) trivial to
1160 extend emacsclient to handle suspend/resume. A `kill -STOP' almost
1163 (Done. I needed to play with signal handling and the server
1164 protocol a bit to make emacsclient behave as a normal UNIX program
1165 wrt foreground/background process groups.)
1167 -- There is a flicker during the startup of `emacs -nw'; it's as if
1168 the terminal is initialized, reset and then initialialized again.
1169 Debug this. (Hint: narrow_foreground_group is called twice during
1174 -- Robert Chassell has found serious copy-paste bugs with the
1175 multi-tty branch. There seem to be redisplay bugs while copying
1176 from X to a terminal frame. Copying accented characters do not
1179 (Patch-124 should fix this, by changing the interprogram-*-function
1180 variables to be frame-local, as suggested by Mark Plaksin
1181 (thanks!). I think that the redisplay bugs are in fact not bugs,
1182 but delays caused by single_kboard --> perhaps MULTI_KBOARD should
1185 -- frame-creation-function was removed, which might be a bad idea.
1186 Think up a compatible solution.
1188 (It was an internal interface that may be changed when necessary.)
1190 -- Change Lisp code not to (getenv "TERM"); use the `tty-type' frame
1191 parameter or the frame-tty-type function instead. (M-x tags-search
1192 "TERM" helps with this.) Update: Actually, all getenv invocations
1193 should be checked for multi-tty compatibility, and an interface
1194 must be implemented to get the remote client's environment.
1196 (Done. Only getenv calls in lisp/term/*.el were changed; other
1197 calls should be mostly left as they are.)
1199 -- Add an elaborate mechanism for display-local variables. (There are
1200 already a few of these; search for `terminal-local' in the Elisp
1203 (Not needed. Display-local variables could be emulated by
1204 frame-local variables.)
1206 -- Emacs assumes that all terminal frames have the same locale
1207 settings as Emacs itself. This may lead to bogus results in a
1208 multi-locale setup. (E.g., while logging in from a remote client
1209 with a different locale.)
1210 (Update after new bugreport by Friedrich Delgado Friedrichs:
1211 (at least) the structs terminal_coding and keyboard_coding in
1212 coding.c must be moved to struct display, and the Lisp interface
1213 [set-]keyboard-coding-system must be adapted for the change.)
1215 (Fixed. Emacs now uses the locale settings as seen by the
1216 emacsclient process for server tty frames.)
1217 (Update: Not really; Vlocale_coding_system is still global.)
1219 -- Make `struct display' accessible to Lisp programs. Accessor functions:
1221 (displayp OBJECT): Returns t if OBJECT is a display.
1222 => Implemented as display-live-p.
1224 (display-list): Returns list of currently active displays.
1227 (selected-display): Returns the display object of the selected frame.
1228 => Not strictly necessary, but implemented anyway.
1230 (frame-display FRAME): Returns the display object of FRAME.
1233 (display-frames DISPLAY): Returns a list of frames on DISPLAY.
1234 => Already implemented, see frames-on-display-list.
1236 (display-type DISPLAY): Returns the type of DISPLAY, as a
1237 symbol. (See `framep'.)
1238 => Implemented as display-live-p.
1240 (display-device DISPLAY): Returns the name of the device that
1241 DISPLAY uses, as a string. (E.g: "/dev/pts/16", or
1243 => Implemented as display-name.
1247 See next issue why this is necessary.
1249 (Update: The consensus on emacs-devel seems to be to do this via
1250 integer identifiers. That's fine by me.)
1254 -- The following needs to be supported:
1261 The cleanest way to solve this is to allow multiple displays on the
1262 same terminal device; each new emacsclient process should create
1263 its own display. As displays are currently identified by their
1264 device names, this is not possible until struct display becomes
1265 accessible as a Lisp-level object.
1269 -- Miles Bader suggests that C-x C-c on an emacsclient frame should
1270 only close the frame, not exit the entire Emacs session. Update:
1271 see above for a function that does this. Maybe this should be the
1274 (Done. This is the new default. No complaints so far.)
1276 -- Clean up the frame-local variable system. I think it's ugly and
1277 error-prone. But maybe I just haven't yet fully understood it.
1279 (Nothing to do. It doesn't seem ugly any more. It's rather clever.)
1281 -- Support multiple character locales. A version of
1282 `set-locale-environment' needs to be written for setting up
1283 display-local settings on ttys. I think calling
1284 set-display-table-and-terminal-coding-system and
1285 set-keyboard-coding-system would be enough. The language
1286 environment itself should remain a global setting.
1288 (Done, by an ugly hack.)
1290 -- The terminal customization files in term/*.el tend to change global
1291 parameters, which may confuse Emacs with multiple displays. Change
1292 them to tweak only frame-local settings, if possible. (They tend
1293 to call define-key to set function key sequences a lot.)
1295 (Done, by making `function-key-map' terminal-local (i.e., part of
1296 struct kboard). This has probably covered all the remaining problems.)
1298 -- Make `function-key-map' and `key-translation-map' terminal-local.
1302 -- Implement `terminal-local-value' and `set-terminal-local-value' to
1303 allow deterministic access to terminal local bindings. The
1304 encode-kb package can not set up `key-translation-map' without
1305 these. The terminal-local bindings seem to be independent of what
1310 -- xt-mouse.el needs to be adapted for multi-tty. It currently
1311 signals an error on kill-emacs under X, which prevents the user
1312 from exiting Emacs. (Reported by Mnemonikk on freenode.)
1317 -- Having {reset,init}_all_sys_modes in set-input-mode breaks arrow
1318 keys on non-selected terminals under screen, and sometimes on other
1319 terminal types as well. The other function keys continue to work
1320 fine. Sometimes faces on these screens become garbled.
1322 This only seems to affect displays that are of the same terminfo
1323 type as the selected one. Interestingly, in screen Emacs normally
1324 reports the up arrow key as `M-o A', but after the above SNAFU, it
1325 complains about `M-[ a'. UNIX ttys are a complete mystery to me,
1326 but it seems the reset-reinitialize cycle somehow leaves the
1327 non-selected terminals in a different state than usual. I have no
1328 idea how this could happen.
1330 Currently set-input-mode resets only the currently selected
1331 terminal, which seems to somehow work around the problem.
1335 Dan Nicolaescu <dann@ics.uci.edu> writes:
1336 > Some terminals have 2 modes for cursor keys: Application Mode where
1337 > the cursor keys transmit the codes defined in the terminfo entry, and
1338 > Cursor mode. Applications have to send the smkx and rmkx terminfo
1339 > strings to switch between the 2 modes. So Emacs (and emacsclient) have
1340 > to send smkx when initializing and rmkx when quitting (or on
1343 (I think patch-370 fixed this.)
1345 -- This long-standing bug (first reported by Han Boetes) seems to come
1346 and go all the time. It is time to track it down and fix it.
1351 # From another xterm:
1352 emacsclient -e '(y-or-n-p "Do you want me to crash? ")'
1353 # Notice how the answer ends up in the *scratch* buffer
1357 (Fixed in patch-414 after detailed analysis by Kalle Olavi Niemitalo.)
1359 -- normal-erase-is-backspace-mode in simple.el needs to be updated for
1360 multi-tty (rep. by Dan Waber). (The Delete key is broken on X
1363 (Fixed in patch-427.)
1365 -- I think keyboard-translate-table should be made terminal-local.
1367 (Done in patch-431.)
1369 -- The semantics of terminal-local variables are confusing; it is not
1370 clear what binding is in effect in any given time. See if
1371 current_kboard (or at least the terminal-local bindings exported to
1372 Lisp) might be changed to be tied to the selected frame instead.
1373 Currently, `function-key-map' and `key-translation-map' may be
1374 accessed reliably only using the hackish
1375 `(set-)terminal-local-value' functions.
1377 Perhaps there should be a difference between `last-command' &co.
1378 and these more conventional configuration variables.
1379 (E.g. `symbol-value' would use current_kboard to access
1380 `last-command', but SELECTED_FRAME()->display->kboard to get the
1381 value of `function-key-map'.
1383 (Fixed in patch-434.)
1385 -- If the first key pressed on a new tty terminal is a function key,
1386 it is not recognized correctly. May be related to the bug below.
1388 (Seems to have been fixed as a side effect of patch-434. "The bug
1389 below" was the set-input-mode madness.)
1391 (Update: this bug was fixed for good in patch-449. It was tracked
1392 down to a bug in `read_key_sequence': it failed to reinitialize its
1393 local function-key-map/key-translation-map references when it
1394 switched keyboards. I don't understand why did this bug only
1395 appear on brand new frames, though!)
1397 -- Disable connecting to a new X display when we use the GTK toolkit.
1399 (Disabled in patch-450.)
1401 -- Implement automatic forwarding of client environment variables to
1402 forked processes, as discussed on the multi-tty list. Terminal
1403 parameters are now accessible in C code, so the biggest obstacle is
1404 gone. The `getenv_internal' and `child_setup' functions in
1405 callproc.c must be changed to support the following variable:
1407 terminal-local-environment-variables is a variable defined in ...
1409 Enable or disable terminal-local environment variables.
1411 If set to t, `getenv', `setenv' and subprocess creation
1412 functions use the environment variables of the emacsclient
1413 process that created the selected frame, ignoring
1414 `process-environment'.
1416 If set to nil, Emacs uses `process-environment' and ignores
1417 the client environment.
1419 Otherwise, `terminal-local-environment-variables' should be a
1420 list of variable names (represented by Lisp strings) to look
1421 up in the client environment. The rest will come from
1422 `process-environment'.
1424 (Implemented in patch-461; `terminal-getenv', `terminal-setenv' and
1425 `with-terminal-environment' are now replaced by extensions to
1426 `getenv' and `setenv', and the new `local-environment-variables'
1429 (Updated in patch-465 to fix the semantics of let-binding
1430 `process-environment'. `process-environment' was changed to
1431 override all local/global environment variables, and a new variable
1432 `global-environment' was introduced to have `process-environment's
1435 (Updated in patch-466 to fix the case when two emacsclient sessions
1436 share the same terminal, but have different environment. The local
1437 environment lists are now stored as frame parameters, so the
1438 C-level terminal parameters are not strictly necessary any more.)
1440 -- `Fdelete_frame' is called from various critical places where it is
1441 not acceptable for the frame deletion to fail, e.g. from
1442 x_connection_closed after an X error. `Fdelete_frame' now protects
1443 against `delete-frame-functions' throwing an error and preventing a
1444 frame delete. (patch-475)
1446 -- Fix set-input-mode for multi-tty. It's a truly horrible interface;
1447 what if we'd blow it up into several separate functions (with a
1448 compatibility definition)?
1450 (Done. See `set-input-interrupt-mode', `set-output-flow-control',
1451 `set-input-meta-mode' and `set-quit-char'.) (patch-457)
1453 -- Let-binding `overriding-terminal-local-map' on a brand new frame
1454 does not seem to work correctly. (See `fancy-splash-screens'.)
1455 The keymap seems to be set up right, but events go to another
1456 terminal. Or is it `unread-command-events' that gets Emacs
1457 confused? Investigate.
1459 (Emacs was confused because a process filter entered
1460 `recursive-edit' while Emacs was reading input. I added support
1461 for this in the input system.) (patch-489)
1463 -- I smell something funny around pop_kboard's "deleted kboard" case.
1464 Determine what are the circumstances of this case, and fix any
1465 bug that comes to light.
1467 (It happens simply because single_kboard's terminal is sometimes
1468 deleted while executing a command on it, for example the one that
1469 kills the terminal. There was no bug here, but I rewrote the whole
1470 single_kboard mess anyway.) (patch-489)
1472 -- Understand Emacs's low-level input system (it's black magic) :-)
1473 What exactly does interrupt_input do? I tried to disable it for
1474 raw secondary tty support, but it does not seem to do anything
1475 useful. (Update: Look again. X unconditionally enables this, maybe
1476 that's why raw terminal support is broken again. I really do need
1477 to understand input.)
1478 (Update: I am starting to understand the read_key_sequence->read-char
1479 ->kbd_buffer_get_event->read_avail_input->read_socket_hook path. Yay!)
1481 (Update: OK, it all seems so easy now (NOT). Input could be done
1482 synchronously (with wait_reading_process_input), or asynchronously
1483 by SIGIO or polling (SIGALRM). C-g either sets the Vquit_flag,
1484 signals a 'quit condition (when immediate_quit), or throws to
1485 `getcjmp' when Emacs was waiting for input when the C-g event
1488 -- Replace wrong_kboard_jmpbuf with a special return value of
1489 read_char. It is absurd that we use setjmp/longjmp just to return
1490 to the immediate caller.
1492 (Done in patch-500.)
1494 -- `tool-bar-mode', `scroll-bar-mode', `menu-bar-mode' and
1495 'fringe-mode' are modes global to the entire Emacs session, not
1496 just a single frame or a single terminal. This means that their
1497 status sometimes differs from what's actually displayed on the
1498 screen. As a consequence, the Options | Show/Hide menu sometimes
1499 shows incorrect status, and you have to select an option twice for
1500 it to have any visible effect on the current frame.
1502 Change Emacs so that the status of the items in the Options |
1503 Show/Hide menu correspond to the current frame.
1505 (Done in patch-537.)
1507 -- The `default-directory' variable should somehow be set to the
1508 cwd of the emacsclient process when the user runs emacsclient
1509 without file arguments. Perhaps it is OK to just override the
1510 directory of the *scratch* buffer.
1512 (Done in patch-539.)
1514 -- The borders on tooltip frames on X are messed up. More
1515 specifically, the frame's internal border (internal-border-width
1516 frame parameter) is not filled with the correct background color.
1518 It seems the border contents are drawn onto by the
1519 update_single_window call in `x-show-tip'. After some debugging, I
1520 think the window's background color is not set up
1521 correctly---calling `x_clear_area' fills the specified area with
1522 black, not light yellow.
1524 (Done in patch-544. A background_pixel field was defined both in
1525 struct frame and struct x_output, and Emacs got confused between
1528 ;;; arch-tag: 8da1619e-2e79-41a8-9ac9-a0485daad17d