Commit | Line | Data |
---|---|---|
3224dac1 | 1 | -*- coding: utf-8; mode: text; -*- |
28d440ab KL |
2 | GOAL |
3 | ---- | |
4 | ||
2246281f KL |
5 | The goal of this branch is to implement support for opening multiple, |
6 | different tty devices and simultaneous X and tty frames from a single | |
7 | Emacs session. | |
28d440ab | 8 | |
7b00d185 KL |
9 | Some use cases: |
10 | ||
11 | Emacs is notoriously slow at startup, so most people use another | |
12 | editor or emacsclient for quick editing jobs from the console. | |
13 | Unfortunately, emacsclient was very awkward to use, because it did not | |
14 | support opening a new Emacs frame on the current virtual console. | |
15 | Now, with multi-tty support, it can do that. (Emacsclient starts up | |
16 | faster than vi!) | |
17 | ||
18 | Some Gnus users (including me) run Gnus in an X frame in its own Emacs | |
19 | instance, which they typically leave running for weeks. It would be | |
20 | nice if they could connect to this instance from a remote ssh session | |
21 | and check their messages without opening a remote X frame or resorting | |
22 | to gnus-slave. | |
23 | ||
24 | ||
28d440ab KL |
25 | WHO IS DOING IT |
26 | --------------- | |
27 | ||
28 | I'm Károly Lőrentey. My address: lorentey@elte.hu. | |
29 | ||
2246281f | 30 | Comments, bug reports, suggestions and patches are welcome! |
28d440ab | 31 | |
6548cf00 | 32 | Retrieving the latest version of the branch: |
4f0359de KL |
33 | |
34 | tla register-archive lorentey@elte.hu--2004 http://lorentey.web.elte.hu/arch/2004/ | |
daf01701 | 35 | tla get lorentey@elte.hu--2004/emacs--multi-tty <directory> |
4f0359de | 36 | |
7b00d185 KL |
37 | (I use a recent arch development snapshot, but any of the released |
38 | versions of arch will do fine, I think.) | |
39 | ||
40 | If you don't have arch, the branch has a homepage from which you can | |
41 | download conventional patches against Emacs CVS HEAD: | |
4f0359de | 42 | |
7b00d185 | 43 | http://lorentey.web.elte.hu/project/emacs.html |
28d440ab | 44 | |
6c3d443b KL |
45 | If you want to be notified of new revisions, sign up to the |
46 | multi-tty@listbox.elte.hu mailing list by visiting the | |
47 | below homepage and following its instructions. | |
48 | ||
49 | https://listbox.elte.hu/mailman/listinfo/multi-tty | |
50 | ||
51 | Please note that the multi-tty mailing list is read-only, and is | |
52 | reserved for automatic commit messages. Discussion about the branch | |
53 | and bug reports should be sent directly to me (lorentey@elte.hu), or | |
54 | to the emacs-devel@gnu.org mailing list. | |
55 | ||
28d440ab KL |
56 | STATUS |
57 | ------ | |
58 | ||
7b00d185 | 59 | Multi-tty support is stable, I think most of the problems were fixed. |
2246281f KL |
60 | (It still needs testing on other architectures, though.) Simultaneous |
61 | X and tty frame support works fine, although there are some small | |
62 | issues. Emacsclient has been extended to support opening a new | |
63 | terminal frame. | |
64 | ||
65 | ||
66 | Please let me know if you find any bugs in this branch. | |
28d440ab | 67 | |
7b00d185 | 68 | To try it out, compile and run the multi-tty branch with the following |
04c3243c | 69 | commands: |
bc279d67 KL |
70 | |
71 | mkdir +build | |
72 | cd +build | |
daf01701 | 73 | ../configure |
bc279d67 | 74 | make bootstrap |
2246281f | 75 | src/emacs -nw # You can also try without -nw |
7b00d185 | 76 | M-x server-start |
bc279d67 | 77 | |
7b00d185 KL |
78 | and then (from a shell prompt on another terminal) start emacsclient |
79 | with | |
28d440ab | 80 | |
7b00d185 | 81 | lib-src/emacsclient -t /optional/file/names... |
28d440ab | 82 | |
04c3243c | 83 | You'll hopefully have two fully working, independent frames on |
7b00d185 KL |
84 | separate terminals. The new frame is closed automatically when you |
85 | have finished editing the specified files (C-x #), but delete-frame | |
86 | (C-x 5 0) also works. Of course, you can create frames on more than | |
87 | two tty devices. | |
52c2ee2a | 88 | |
7b00d185 KL |
89 | Creating new frames on the same tty with C-x 5 2 works, and they |
90 | behave the same way as in previous Emacs versions. If you exit emacs, | |
91 | all terminals should be restored to their previous states. | |
daf01701 | 92 | |
7b00d185 KL |
93 | This is work in progress, and probably full of bugs. You should |
94 | always run emacs from gdb, so that you'll have a live instance to | |
2246281f | 95 | debug if something goes wrong. Please send me your bug reports. |
04c3243c | 96 | |
7b00d185 | 97 | Problems: |
6548cf00 | 98 | |
7b00d185 KL |
99 | * Suspending Emacs is disabled if there are multiple tty |
100 | devices. Also, there is no way to suspend emacsclient. This | |
101 | will be fixed. | |
102 | ||
7b00d185 KL |
103 | * Mac, Windows and DOS support is broken, probably doesn't |
104 | even compile -- this will be solved later. | |
105 | ||
2246281f | 106 | * Only tested on my GNU/Linux box and on Solaris 8. |
28d440ab KL |
107 | |
108 | NEWS | |
109 | ---- | |
110 | ||
111 | For the NEWS file: | |
112 | ||
2246281f | 113 | ** Support for multiple terminal devices has been added. |
6548cf00 | 114 | |
2246281f KL |
115 | *** You can specify a terminal device (`tty' parameter) and a terminal |
116 | type (`tty-type' parameter) to `make-terminal-frame'. | |
6548cf00 | 117 | |
2246281f KL |
118 | *** You can test for the presence of multiple terminal support by |
119 | testing for the `multi-tty' feature. | |
28d440ab | 120 | |
2246281f KL |
121 | *** Emacsclient has been extended to support opening a new terminal |
122 | frame (see -t option). | |
123 | ||
124 | *** A make-frame-on-tty function has been added to make it easier to | |
125 | create frames on new terminals. | |
126 | ||
127 | *** New functions: frame-tty-name, frame-tty-type for accessing | |
128 | terminal parameters, and delete-tty for closing the terminal | |
129 | device. | |
130 | ||
131 | *** talk.el has been extended for multiple tty support. | |
132 | ||
133 | ||
134 | ** Support for simultaneous graphical and terminal frames has been | |
135 | added. | |
136 | ||
137 | *** The function `make-frame-on-display' now works during a terminal | |
138 | session, and `make-frame-on-tty' works during a graphical session. | |
139 | ||
140 | *** The `window-system' variable has been made frame-local. | |
141 | ||
142 | *** The new `initial-window-system' variable contains the | |
143 | `window-system' value for the first frame. | |
819b8f00 | 144 | |
d448e982 KL |
145 | THANKS |
146 | ------ | |
147 | ||
148 | The following is an (incomplete) list of people who have contributed | |
149 | to the project by testing, bug reports, and suggestions. Thanks! | |
150 | ||
151 | Robert J. Chassel <bob@rattlesnake.com> | |
152 | Romain Francoise <romain@orekobech.com> | |
153 | Ami Fischman <ami@fischman.org> | |
154 | ||
28d440ab KL |
155 | |
156 | CHANGELOG | |
157 | --------- | |
158 | ||
159 | See arch logs. | |
160 | ||
7b00d185 KL |
161 | THINGS TO DO |
162 | ------------ | |
163 | ||
d448e982 | 164 | ** Robert J. Chassell reports: |
e9cda827 | 165 | |
d448e982 KL |
166 | > * After starting the frame in the VC, I saw this message in the |
167 | > *Message* buffer | |
168 | > | |
169 | > error in process filter: server-process-filter: \ | |
170 | > Wrong type argument: sequencep,\ | |
171 | > framep | |
172 | > error in process filter: Wrong type argument: sequencep, framep | |
173 | > | |
174 | > This also happens when I start a new frame in an xterm. | |
175 | ||
176 | ** emacs -nw --eval '(y-or-n-p "Foobar")' segfaults. | |
3224dac1 | 177 | |
114a8b8c KL |
178 | ** Fix interactive use of temacs. There are face-related SEGVs, most |
179 | likely because of changes in realize_default_face, realize_face. | |
180 | ||
4ca927b4 KL |
181 | ** Very strange bug: visible-bell does not work on secondary |
182 | terminals. This might be something xterm (konsole) specific. | |
183 | ||
7b00d185 KL |
184 | ** Find out the best way to support suspending Emacs with multiple |
185 | ttys. My guess: disable it on the controlling tty, but from other | |
186 | ttys pass it on to emacsclient somehow. (It is (I hope) trivial to | |
187 | extend emacsclient to handle suspend/resume. A `kill -STOP' almost | |
188 | works right now.) | |
189 | ||
428a555e | 190 | ** Move baud_rate to struct display. |
7b00d185 | 191 | |
7b00d185 KL |
192 | ** Implement support for starting an interactive Emacs session without |
193 | an initial frame. (The user would connect to it and open frames | |
194 | later, with emacsclient.) Not necessarily a good idea. | |
195 | ||
3224dac1 KL |
196 | ** Fix Mac support (I can't do this myself). Note that the current |
197 | state of Mac-specific source files in the multi-tty tree are not | |
198 | useful; before starting work on Mac support, revert to pristine, | |
199 | pre-multi-tty versions. | |
7b00d185 | 200 | |
3224dac1 KL |
201 | ** Fix W32 support (I can't do this myself). Note that the current |
202 | state of W32-specific source files in the multi-tty tree are not | |
203 | useful; before starting work on W32 support, revert to pristine, | |
204 | pre-multi-tty versions. | |
7b00d185 | 205 | |
3224dac1 KL |
206 | ** Fix DOS support (I can't do this myself). Note that the current |
207 | state of DOS-specific source files in the multi-tty tree are not | |
208 | useful; before starting work on DOS support, revert to pristine, | |
209 | pre-multi-tty versions. | |
7b00d185 KL |
210 | |
211 | ** Do a grep on XXX and ?? for more issues. | |
212 | ||
213 | ** Understand Emacs's low-level input system (it seems complicated) :-) | |
428a555e | 214 | What does interrupt_input do? I tried to disable it for raw |
7b00d185 KL |
215 | secondary tty support, but it does not seem to do anything useful. |
216 | (Update: Look again. X unconditionally enables this, maybe that's | |
217 | why raw terminal support is broken again. I really do need to | |
218 | understand input.) | |
219 | ||
7b00d185 KL |
220 | ** emacsclient -t from an Emacs term buffer does not work, complains |
221 | about face problems. This can even lock up Emacs (if the recursive | |
428a555e KL |
222 | frame sets single_kboard). Update: the face problems are caused by |
223 | bugs in term.el, not in multi-tty. The lockup is caused by | |
4ca927b4 KL |
224 | single_kboard mode, and is not easily solvable. The best thing to |
225 | do is to simply refuse to create a tty frame of type `eterm'. | |
4d553a13 | 226 | |
2246281f KL |
227 | ** Maybe standard-display-table should be display-local. |
228 | ||
6548cf00 KL |
229 | DIARY OF CHANGES |
230 | ---------------- | |
231 | ||
232 | (ex-TODO items with explanations.) | |
28d440ab | 233 | |
114a8b8c | 234 | -- Introduce a new struct for terminal devices. |
28d440ab | 235 | |
daf01701 | 236 | (Done, see struct tty_output. The list of members is not yet |
28d440ab KL |
237 | complete.) |
238 | ||
28d440ab KL |
239 | -- Change the bootstrap procedure to initialize tty_list. |
240 | ||
241 | (Done, but needs review.) | |
242 | ||
28d440ab KL |
243 | -- Change make-terminal-frame to support specifying another tty. |
244 | ||
245 | (Done, new frame parameters: `tty' and `tty-type'.) | |
246 | ||
9628b887 KL |
247 | -- Implement support for reading from multiple terminals. |
248 | ||
249 | (Done, read_avail_input tries to read from each terminal, until one | |
28d7d09f | 250 | succeeds. MULTI_KBOARD is not used. Secondary terminals don't send |
6548cf00 | 251 | SIGIO!) |
9628b887 | 252 | |
819b8f00 KL |
253 | (Update: They do, now.) |
254 | ||
daf01701 KL |
255 | (Update2: After enabling X, they don't.) |
256 | ||
9628b887 | 257 | -- other-frame should cycle through the frames on the `current' |
114a8b8c | 258 | terminal only. |
9628b887 | 259 | |
6548cf00 | 260 | (Done, by trivially modifiying next_frame and prev_frame.) |
9628b887 KL |
261 | |
262 | -- Support different terminal sizes. | |
114a8b8c | 263 | |
9628b887 KL |
264 | (Done, no problem.) |
265 | ||
266 | -- Make sure terminal resizes are handled gracefully. (Could be | |
267 | problematic.) | |
268 | ||
6548cf00 KL |
269 | (Done. We don't get automatic SIGWINCH for additional ttys, |
270 | though.) | |
9628b887 KL |
271 | |
272 | -- Extend emacsclient to automatically open a new tty when it connects | |
273 | to Emacs. | |
114a8b8c | 274 | |
9628b887 KL |
275 | (Done. It's an ugly hack, needs more work.) |
276 | ||
52c2ee2a KL |
277 | -- Redisplay must refresh the topmost frame on *all* terminals, not |
278 | just the initial terminal. | |
114a8b8c | 279 | |
52c2ee2a | 280 | (Done, but introduced an ugly redisplay problems. Ugh.) |
9628b887 | 281 | |
6548cf00 KL |
282 | -- Fix redisplay problems. |
283 | ||
52c2ee2a KL |
284 | (Done; it turned out that the entire Wcm structure must be moved |
285 | inside tty_output. Why didn't I catch this earlier?) | |
6548cf00 KL |
286 | |
287 | -- Provide a way for emacsclient to tell Emacs that the tty has been | |
288 | resized. | |
289 | ||
290 | (Done, simply forward the SIGWINCH signal.) | |
291 | ||
292 | -- Each keypress should automatically select the frame corresponding | |
293 | to the terminal that it was coming from. This means that Emacs | |
294 | must know from which terminal the last keyboard event came from. | |
6548cf00 | 295 | |
52c2ee2a KL |
296 | (Done, it was quite simple, the input event system already |
297 | supported multiple frames.) | |
6548cf00 KL |
298 | |
299 | -- Fix SIGIO issue with secondary terminals. | |
300 | ||
301 | (Done, emacsclient signals Emacs after writing to the proxy pseudo | |
52c2ee2a KL |
302 | terminal. Note that this means that multi-tty does not work with |
303 | raw ttys!) | |
28d440ab | 304 | |
0a125897 KL |
305 | (Update: This is bullshit. There is a read_input_waiting function, |
306 | extend that somehow.) | |
307 | ||
819b8f00 KL |
308 | (Update of update: The first update was not right either, extending |
309 | read_input_waiting was not necessary. Secondary ttys do seem to | |
310 | send signals on input.) | |
311 | ||
daf01701 KL |
312 | (Update^3: Not any more.) |
313 | ||
fca177d4 KL |
314 | -- Make make-terminal-frame look up the `tty' and `tty-type' frame |
315 | parameters from the currently selected terminal before the global | |
316 | default. | |
317 | ||
318 | (Done.) | |
319 | ||
320 | -- Put all cached terminal escape sequences into struct tty_output. | |
321 | Currently, they are still stored in global variables, so we don't | |
322 | really support multiple terminal types. | |
323 | ||
52c2ee2a | 324 | (Done. It was not fun.) |
fca177d4 KL |
325 | |
326 | -- Implement sane error handling after initialization. (Currently | |
6548cf00 KL |
327 | emacs exits if you specify a bad terminal type.) The helpful error |
328 | messages must still be provided when Emacs starts. | |
329 | ||
fca177d4 KL |
330 | (Done.) |
331 | ||
332 | -- Implement terminal deletion, i.e., deleting local frames, closing | |
333 | the tty device and restoring its previous state without exiting | |
334 | Emacs. | |
9628b887 | 335 | |
fca177d4 KL |
336 | (Done, but at the moment only called when an error happens during |
337 | initialization. There is a memory corruption error around this | |
daf01701 | 338 | somewhere.) (Update: now it is fully enabled.) |
fca177d4 | 339 | |
8303ba32 KL |
340 | -- Implement automatic deletion of terminals when the last frame on |
341 | that terminal is closed. | |
342 | ||
343 | (Done.) | |
344 | ||
0a125897 KL |
345 | -- Restore tty screen after closing the terminal. |
346 | ||
347 | (Done, we do the same as Emacs 21.2 for all terminals.) | |
348 | ||
349 | -- 'TERM=dumb src/emacs' does not restore the terminal state. | |
114a8b8c | 350 | |
0a125897 | 351 | (Done.) |
fca177d4 | 352 | |
b2af72d2 KL |
353 | -- C-g should work on secondary terminals. |
354 | ||
355 | (Done, but the binding is not configurable.) | |
356 | ||
357 | -- Deal with SIGHUP in Emacs and in emacsclient. (After this, the | |
358 | server-frames may be removed from server.el.) | |
359 | ||
360 | (Done, nothing to do. It seems that Emacs does not receive SIGHUP | |
daf01701 | 361 | from secondary ttys, which is actually a good thing.) (Update: I |
7b00d185 | 362 | think it would be a bad idea to remove server-frames.) |
b2af72d2 | 363 | |
daf01701 | 364 | -- Change emacsclient/server.el to support the -t argument better, |
b2af72d2 KL |
365 | i.e. automatically close the socket when the frame is closed. |
366 | ||
367 | (Seems to be working OK.) | |
368 | ||
16c290d8 | 369 | -- Fix mysterious memory corruption error with tty deletion. To |
fca177d4 KL |
370 | trigger it, try the following shell command: |
371 | ||
372 | while true; do TERM=no-such-terminal-definition emacsclient -h; done | |
373 | ||
374 | Emacs usually dumps core after a few dozen iterations. (The bug | |
3224dac1 | 375 | seems to be related to the xfreeing or bzeroing of |
9d9f1812 KL |
376 | tty_output.Wcm. Maybe there are outside references to struct Wcm? |
377 | Why were these vars collected into a struct before multi-tty | |
378 | support?) | |
fca177d4 | 379 | |
16c290d8 KL |
380 | (Done. Whew. It turned out that the problem had nothing to do |
381 | with hypothetical external references to Wcm, or any other | |
382 | tty_output component; it was simply that delete_tty closed the | |
383 | filehandles of secondary ttys twice, resulting in fclose doubly | |
3224dac1 | 384 | freeing memory. Utterly trivial matter. I love the C's memory |
16c290d8 KL |
385 | management, it puts hair on your chest.) |
386 | ||
819b8f00 | 387 | -- Support raw secondary terminals. (Note that SIGIO works only on |
3224dac1 | 388 | the controlling terminal.) Hint: extend read_input_waiting for |
819b8f00 | 389 | multiple ttys and hopefully this will be fixed. |
fca177d4 | 390 | |
819b8f00 KL |
391 | (Done, it seems to have been working already for some time. It |
392 | seems F_SETOWN does work, after all. Not sure what made it fail | |
393 | earlier, but it seems to be fixed (there were several changes | |
394 | around request_sigio, maybe one of them did it). | |
3224dac1 | 395 | read_input_waiting is only used in sys_select, don't change |
daf01701 | 396 | it.) (Update: After adding X support, it's broken again.) |
c1c63edb | 397 | |
819b8f00 | 398 | -- Find out why does Emacs abort when it wants to close its |
0a125897 | 399 | controlling tty. Hint: chan_process[] array. Hey, maybe |
16c290d8 KL |
400 | noninterrupt-IO would work, too? Update: no, there is no process |
401 | for stdin/out. | |
c33d2d42 | 402 | |
819b8f00 KL |
403 | (Done. Added add/delete_keyboard_wait_descriptor to |
404 | term_init/delete_tty. The hint was right, in a way.) | |
6548cf00 | 405 | |
819b8f00 | 406 | -- Issue with SIGIO: it needs to be disabled during redisplay. See if |
3224dac1 | 407 | fcntl kernel behaviour could be emulated by emacsclient. |
0a125897 | 408 | |
819b8f00 | 409 | (Done. Simply disabled the SIGIO emulation hack in emacsclient.) |
daf01701 | 410 | (Update: it was added back.) |
22de1e79 | 411 | |
819b8f00 | 412 | -- server.el: There are issues with saving files in buffers of closed |
16c290d8 KL |
413 | clients. Try editing a file with emacsclient -f, and (without |
414 | saving it) do a delete-frame. The frame is closed without | |
415 | question, and a surprising confirmation prompt appears in another | |
416 | frame. | |
417 | ||
819b8f00 KL |
418 | (Done. delete-frame now asks for confirmation if it still has |
419 | pending buffers, and modified buffers don't seem to be deleted.) | |
420 | ||
421 | -- emacsclient.el, server.el: Handle eval or file open errors when | |
daf01701 | 422 | doing -t. |
16c290d8 | 423 | |
819b8f00 | 424 | (Done.) |
22de1e79 | 425 | |
819b8f00 KL |
426 | -- Make parts of struct tty_output accessible from Lisp. The device |
427 | name and the type is sufficient. | |
22de1e79 | 428 | |
819b8f00 KL |
429 | (Done, see frame-tty-name and frame-tty-type.) |
430 | ||
431 | -- Export delete_tty to the Lisp environment, for emacsclient. | |
22de1e79 | 432 | |
819b8f00 KL |
433 | (Done, see delete-tty.) |
434 | ||
daf01701 KL |
435 | -- Get rid of the accessor macros in termchar.h, or define macros for |
436 | all members. | |
437 | ||
438 | (Done.) | |
439 | ||
440 | -- Move device-specific parameters (like costs) commonly used by | |
441 | device backends to a common, device-dependent structure. | |
442 | ||
443 | (Done. See struct display_method in termhooks.h.) | |
444 | ||
445 | -- Fix X support. | |
446 | ||
447 | (Done. Well, it seems to be working.) | |
448 | ||
449 | -- Allow simultaneous X and tty frames. (Handling input could be | |
450 | tricky. Or maybe not.) | |
451 | ||
452 | (Done. Allowed, that is. It is currently extremely unstable, to | |
453 | the point of being unusable. The rif variable causes constant | |
454 | core dumps. Handling input is indeed tricky.) | |
819b8f00 | 455 | |
7b00d185 | 456 | -- Rewrite multi-tty input in terms of MULTI_KBOARD. |
28d440ab | 457 | |
7b00d185 KL |
458 | (Done. In fact, there was no need to rewrite anything, I just |
459 | added a kboard member to tty_display_info, and initialized the | |
460 | frame's kboard from there.) | |
819b8f00 | 461 | |
fa971ac3 KL |
462 | -- Fix rif issue with X-tty combo sessions. IMHO the best thing to do |
463 | is to get rid of that global variable (and use the value value in | |
464 | display_method, which is guaranteed to be correct). | |
465 | ||
466 | (Done, did exactly that. Core dumps during combo sessions became | |
467 | much rarer. In fact, I have not yet met a single one.) | |
468 | ||
17086732 KL |
469 | -- Add multi-tty support to talk.el. |
470 | ||
471 | (Done.) | |
472 | ||
428a555e KL |
473 | -- Clean up the source of emacsclient. It is a mess. |
474 | ||
475 | (Done, eliminated stupid proxy-pty kludge.) | |
476 | ||
477 | -- Fix faces on tty frames during X-tty combo sessions. There is an | |
478 | init_frame_faces call in init_sys_modes, see if there is a problem | |
479 | with it. | |
480 | ||
481 | (Done, there was a stupid mistake in | |
482 | Ftty_supports_face_attributes_p. Colors are broken, though.) | |
483 | ||
484 | -- C-x 5 2, C-x 5 o, C-x 5 0 on an emacsclient frame unexpectedly | |
485 | exits emacsclient. This is a result of trying to be clever with | |
486 | delete-frame-functions. | |
487 | ||
488 | (Fixed, added delete-tty-after-functions, and changed server.el to | |
489 | use it.) | |
490 | ||
491 | -- Something with (maybe) multi-keyboard support broke function keys | |
492 | and arrows on ttys during X+tty combo sessions. Debug this. | |
493 | ||
494 | (I can't reproduce it, maybe the terminal type was wrong.) | |
495 | ||
496 | -- Fix input from raw ttys (again). | |
497 | ||
498 | (Now it seems to work all right.) | |
499 | ||
500 | -- During an X-tty combo session, a (message "Hello") from a tty frame | |
501 | goes to the X frame. Fix this. | |
502 | ||
503 | (Done. There was a safeguard against writing to the initial | |
504 | terminal frame during bootstrap which prevented echo_area_display | |
505 | from working correctly on a tty frame during a combo session.) | |
506 | ||
114a8b8c KL |
507 | -- If there are no frames on its controlling terminal, Emacs should |
508 | exit if the uses presses C-c there. | |
509 | ||
510 | (Done, as far as possible. See the SIGTERM comment in | |
511 | interrupt_signal on why this seems to be impossible to solve this | |
512 | in general.) | |
513 | ||
514 | -- During an X session, Emacs seems to read from stdin. Also, Emacs | |
515 | fails to start without a controlling tty. | |
516 | ||
517 | (Fixed by replacing the troublesome termcap display with a dummy | |
518 | bootstrap display during bootstrap. | |
519 | ||
520 | -- Do tty output through struct display, like graphical display | |
521 | backends. | |
522 | ||
523 | (Done.) | |
524 | ||
525 | -- Define an output_initial value for output_method for the initial | |
526 | frame that is dumped with Emacs. Checking for this frame (e.g. in | |
527 | cmd_error_internal) is ugly. | |
528 | ||
529 | (Done, broking interactive temacs.) | |
530 | ||
531 | -- The command `emacsclient -t -e '(delete-frame)'' fails to exit. | |
532 | ||
533 | (Fixed.) | |
534 | ||
4ca927b4 KL |
535 | -- frame-creation-function should always create a frame that is on the |
536 | same display as the selected frame. Maybe frame-creation-function | |
537 | should simply be removed and make-frame changed to do the right | |
538 | thing. | |
539 | ||
540 | (Done, with a nice hack. frame-creation-function is now frame-local.) | |
541 | ||
542 | -- Fix C-g on raw ttys. | |
543 | ||
544 | (Done. I disabled the interrupt/quit keys on all secondary | |
545 | terminals, so Emacs sees C-g as normal input. This looks like an | |
546 | overkill, because emacsclient has extra code to pass SIGINT to | |
547 | Emacs, so C-g should remain the interrupt/quit key on emacsclient | |
548 | frames. See the next entry why implementing this distinction would | |
549 | be a bad idea.) | |
550 | ||
551 | -- Make sure C-g goes to the right frame with ttys. This is hard, as | |
552 | SIGINT doesn't have a tty parameter. :-( | |
553 | ||
554 | (Done, the previous change fixes this as a pleasant side effect.) | |
555 | ||
c0707982 KL |
556 | -- I have seen a case when Emacs with multiple ttys fell into a loop |
557 | eating 100% of CPU time. Strace showed this loop: | |
558 | ||
559 | getpid() = 30284 | |
560 | kill(30284, SIGIO) = 0 | |
561 | --- SIGIO (I/O possible) @ 0 (0) --- | |
562 | ioctl(6, FIONREAD, [0]) = -1 EIO (Input/output error) | |
563 | ioctl(5, FIONREAD, [0]) = -1 EIO (Input/output error) | |
564 | ioctl(0, FIONREAD, [0]) = 0 | |
565 | sigreturn() = ? (mask now []) | |
566 | gettimeofday({1072842297, 747760}, NULL) = 0 | |
567 | gettimeofday({1072842297, 747806}, NULL) = 0 | |
568 | select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0}) | |
569 | select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0}) | |
570 | gettimeofday({1072842297, 748245}, NULL) = 0 | |
571 | ||
572 | I have seen something similar with a single X frame, but have not | |
573 | been able to reproduce it for debugging. | |
574 | ||
575 | Update: This may have been caused by checking for nread != 0 | |
576 | instead of nread > 0 after calling read_socket_hook in | |
577 | read_avail_input. | |
578 | ||
579 | (Fixed. This was caused by unconditionally including stdin in | |
580 | input_wait_mask in init_process. The select call in | |
581 | wait_reading_process_input always returned immediately, indicating | |
582 | that there is pending input from stdin, which nobody read. | |
583 | ||
584 | Note that the above strace output seems to be an unrelated but | |
585 | similar bug. I think that is now fixed.) | |
4ca927b4 | 586 | |
2246281f KL |
587 | -- Exiting Emacs while there are emacsclient frames doesn't restore the |
588 | ttys to their default states. | |
589 | ||
590 | (This seems to be fixed by some previous change.) | |
591 | ||
592 | -- Allow opening an X session after -nw. | |
593 | ||
594 | (Done.) | |
595 | ||
d448e982 KL |
596 | -- Fix color handling during tty+X combo sessions. (It seems that tty |
597 | sessions automatically convert the face colors to terminal colors | |
598 | when the face is loaded. This conversion must happen instead on | |
599 | the fly in write_glyphs, which might be problematic, as color | |
600 | approximation is currently done in lisp (term/tty-colors.el).) | |
601 | (Update: hm, colors seem to work fine if I start emacs with -nw and | |
602 | then create an X frame. Maybe it's just a small buglet somewhere.) | |
603 | ||
604 | (Seems to be fixed. The problem was in startup.el, it did not | |
605 | initialize tty colors when the initial window system was | |
606 | graphical.) | |
2246281f | 607 | |
28d440ab | 608 | ;;; arch-tag: 8da1619e-2e79-41a8-9ac9-a0485daad17d |