Implemented suspending of emacsclient frames.

[bpt/emacs.git] / README.multi-tty

                        -*- coding: utf-8; mode: text; -*-
GOAL
----

The goal of this branch is to implement support for opening multiple,
different tty devices and simultaneous X and tty frames from a single
Emacs session.

Some use cases:
Emacs is notoriously slow at startup, so most people use another
editor or emacsclient for quick editing jobs from the console.
Unfortunately, emacsclient was very awkward to use, because it did not
support opening a new Emacs frame on the current virtual console.
Now, with multi-tty support, it can do that.  (Emacsclient starts up
faster than vi!)

Some Gnus users (including me) run Gnus in an X frame in its own Emacs
instance, which they typically leave running for weeks.  It would be
nice if they could connect to this instance from a remote ssh session
and check their messages without opening a remote X frame or resorting
to gnus-slave.

WHO IS DOING IT
---------------

I'm Károly Lőrentey.  My address: lorentey@elte.hu.

Comments, bug reports, suggestions and patches are welcome!

Retrieving the latest version of the branch:

        tla register-archive lorentey@elte.hu--2004 http://lorentey.web.elte.hu/arch/2004/
        tla get lorentey@elte.hu--2004/emacs--multi-tty <directory>

(I use a recent arch development snapshot, but any of the released
versions of arch will do fine, I think.)

If you don't have arch, the branch has a homepage from which you can
download conventional patches against Emacs CVS HEAD:

        http://lorentey.web.elte.hu/project/emacs.html

If you want to be notified of new revisions, sign up to the
multi-tty@listbox.elte.hu mailing list by visiting the
below homepage and following its instructions.

        https://listbox.elte.hu/mailman/listinfo/multi-tty

Please note that the multi-tty mailing list is read-only, and is
reserved for automatic commit messages.  Discussion about the branch
and bug reports should be sent directly to me (lorentey@elte.hu), or
to the emacs-devel@gnu.org mailing list.  (I hope to merge my branch
into CVS HEAD reasonably soon, so I don't want to set up an elaborate
development infrastructure for the multi-tty branch.)

STATUS
------

The branch is now very stable and almost full-featured. I hope the
major problems were fixed.  (It still needs testing on other
architectures, though.)  Both multiple tty device support and
simultaneous X and tty frame support works fine.  Emacsclient has been
extended to support opening new tty and X frames.  It has been changed
open new Emacs frames by default.

Please let me know if you find any bugs in this branch.

HOW TO COMPILE AND TEST
-----------------------

To try out the multi-tty branch, compile and run the multi-tty branch
with the following commands:

        mkdir +build
        cd +build
        ../configure <your favourite options>
        make bootstrap
        src/emacs -nw           # You can also try without -nw
        M-x server-start

and then (from a shell prompt on another terminal) start emacsclient
with
        lib-src/emacsclient /optional/file/names...
or
        lib-src/emacsclient -t /optional/file/names...

You'll hopefully have two fully working, independent frames on
separate terminals. The new frame is closed automatically when you
have finished editing the specified files (C-x #), but delete-frame
(C-x 5 0) also works.  Of course, you can create frames on more than
two tty devices.

Creating new frames on the same tty with C-x 5 2 works, and they
behave the same way as in previous Emacs versions.  If you exit emacs,
all terminals should be restored to their previous states.

This is work in progress, and probably full of bugs.  You should
always run emacs from gdb, so that you'll have a live instance to
debug if something goes wrong.  Please send me your bug reports.

I think the best way to use the new Emacs is to have it running inside
a disconnected GNU screen session, and always use emacsclient for
normal work.  This way not a single keystroke of your work will be
lost if the display device that you are using crashes, or the network
connection times out, or whatever.  As long as the server that runs
your Emacs remains up, all your buffers remain exactly as they were
before the crash, and you can continue working as if nothing had
happened.  (I had an extremely unstable X server for some time while I
was developing these patches, and running Emacs this way has saved me
a number of M-x recover-session invocations.)

(I am thinking about introducing a --daemon option to get rid of the
above screen dependency.)

Problems:

        * Suspending Emacs is disabled if there are multiple tty
          devices.  Also, there is no way to suspend emacsclient. This
          will be fixed.

        * Mac, Windows and DOS support is broken, probably doesn't
          even compile -- this will be solved later.

        * Only tested on my GNU/Linux box and on Solaris 8.

NEWS
----

For the NEWS file:

** Support for multiple terminal devices has been added.

*** You can specify a terminal device (`tty' parameter) and a terminal
    type (`tty-type' parameter) to `make-terminal-frame'.

*** You can test for the presence of multiple terminal support by
    testing for the `multi-tty' feature.

*** Emacsclient has been extended to support opening a new terminal
    frame. Its behaviour has been changed to open a new Emacs frame by
    default.  Use the -c option to get the old behavior of opening
    files in the currently selected Emacs frame.

*** A make-frame-on-tty function has been added to make it easier to
    create frames on new terminals.

*** New functions: frame-tty-name, frame-tty-type for accessing
    terminal parameters, and delete-tty for closing the terminal
    device.

*** talk.el has been extended for multiple tty support.


** Support for simultaneous graphical and terminal frames has been
   added.

*** The function `make-frame-on-display' now works during a terminal
    session, and `make-frame-on-tty' works during a graphical session.

*** The `window-system' variable has been made frame-local.

*** The new `initial-window-system' variable contains the
    `window-system' value for the first frame.

THANKS
------

The following is an (incomplete) list of people who have contributed
to the project by testing, bug reports, and suggestions.  Thanks!

Robert J. Chassel <bob at rattlesnake dot com>
Romain Francoise <romain at orebokech dot com>
Ami Fischman <ami at fischman dot org>
Istvan Marko <mi-mtty ar kismala dot com>
Dan Nicolaescu <dann at ics dot uci dot edu>
Mark Plaksin <happy at mcplaksin dot org>

Richard Stallman was kind enough to review my patches.

CHANGELOG
---------

See arch logs.

THINGS TO DO
------------

** There is a flicker during the startup of `emacs -nw'; it's as if
   the terminal is initialized, reset and then initialialized again.
   Debug this.  (Hint: narrow_foreground_group is called twice during
   startup.)

** Dan Nicolaescu suggests that -nw should be added as an alias for -t
   in emacsclient.  Good idea.  (Alas, implementing this is not
   trivial, getopt_long does not seem to support two-letter ``short''
   options.)

** Mark Plaksin suggests that emacsclient should accept the same
   X-related command-line arguments as Emacs.  Most of the X-related
   argument-handling is done in Lisp, so this should be quite easy to
   implement.

** Make `struct display' accessible to Lisp programs.  Accessor functions:

        (displayp OBJECT):  Returns t if OBJECT is a display.

        (selected-display):  Returns the display object of the selected frame.

        (frame-display FRAME):  Returns the display object of FRAME.

        (display-frames DISPLAY):  Returns a list of frames on DISPLAY.

        (display-type DISPLAY):  Returns the type of DISPLAY, as a
                symbol.  (See `framep'.)

        (display-device DISPLAY): Returns the name of the device that
                DISPLAY uses, as a string.  (E.g: "/dev/pts/16", or
                ":0.0")

   See next issue why this is necessary.

** The following needs to be supported:

        $ emacsclient -t
                C-z
        $ bg
        $ emacsclient -t
                (This fails now.)

   The cleanest way to solve this is to allow multiple displays on the
   same terminal device; each new emacsclient process should create
   its own display.  As displays are currently identified by their
   device names, this is not possible until struct display becomes
   accessible as a Lisp-level object.

** Add an elaborate mechanism for display-local variables.  (There are
   already a few of these; search for `terminal-local' in the Elisp
   manual.)

** Very strange bug: visible-bell does not work on secondary
   terminals in xterm and konsole.  The screen does flicker a bit,
   but it's so quick it isn't noticable.

** Clean up the frame-local variable system.  I think it's ugly and
   error-prone.  But maybe I just haven't yet fully understood it.

** Move baud_rate to struct display.

** Implement support for starting an interactive Emacs session without
   an initial frame.  (The user would connect to it and open frames
   later, with emacsclient.)

** Fix Mac support (I can't do this myself).  Note that the current
   state of Mac-specific source files in the multi-tty tree are not
   useful; before starting work on Mac support, revert to pristine,
   pre-multi-tty versions.

** Fix W32 support (I can't do this myself).  Note that the current
   state of W32-specific source files in the multi-tty tree are not
   useful; before starting work on W32 support, revert to pristine,
   pre-multi-tty versions.

** Fix DOS support (I can't do this myself).  Note that the current
   state of DOS-specific source files in the multi-tty tree are not
   useful; before starting work on DOS support, revert to pristine,
   pre-multi-tty versions.

** Do a grep on XXX and ?? for more issues.

** Understand Emacs's low-level input system (it seems complicated) :-)
   What does interrupt_input do?  I tried to disable it for raw
   secondary tty support, but it does not seem to do anything useful.
   (Update: Look again. X unconditionally enables this, maybe that's
   why raw terminal support is broken again.  I really do need to
   understand input.)

** Maybe standard-display-table should be display-local.

DIARY OF CHANGES
----------------

(ex-TODO items with explanations.)

-- Introduce a new struct for terminal devices.

   (Done, see struct tty_output.  The list of members is not yet
   complete.)

-- Change the bootstrap procedure to initialize tty_list.

   (Done, but needs review.)

-- Change make-terminal-frame to support specifying another tty.

   (Done, new frame parameters: `tty' and `tty-type'.)

-- Implement support for reading from multiple terminals.

   (Done, read_avail_input tries to read from each terminal, until one
   succeeds.  MULTI_KBOARD is not used.  Secondary terminals don't send
   SIGIO!)

   (Update: They do, now.)

   (Update2: After enabling X, they don't.)

-- other-frame should cycle through the frames on the `current'
   terminal only.

   (Done, by trivially modifiying next_frame and prev_frame.)

-- Support different terminal sizes.

   (Done, no problem.)

-- Make sure terminal resizes are handled gracefully.  (Could be
   problematic.)

   (Done.  We don't get automatic SIGWINCH for additional ttys,
   though.)

-- Extend emacsclient to automatically open a new tty when it connects
   to Emacs.

   (Done.  It's an ugly hack, needs more work.)

-- Redisplay must refresh the topmost frame on *all* terminals, not
   just the initial terminal.

   (Done, but introduced an ugly redisplay problems.  Ugh.)

-- Fix redisplay problems.

   (Done; it turned out that the entire Wcm structure must be moved
   inside tty_output.  Why didn't I catch this earlier?)

-- Provide a way for emacsclient to tell Emacs that the tty has been
   resized.

   (Done, simply forward the SIGWINCH signal.)

-- Each keypress should automatically select the frame corresponding
   to the terminal that it was coming from.  This means that Emacs
   must know from which terminal the last keyboard event came from.

   (Done, it was quite simple, the input event system already
   supported multiple frames.)

-- Fix SIGIO issue with secondary terminals.

   (Done, emacsclient signals Emacs after writing to the proxy pseudo
   terminal.  Note that this means that multi-tty does not work with
   raw ttys!)

   (Update: This is bullshit.  There is a read_input_waiting function,
   extend that somehow.)

   (Update of update: The first update was not right either, extending
   read_input_waiting was not necessary.  Secondary ttys do seem to
   send signals on input.)

   (Update^3: Not any more.)

-- Make make-terminal-frame look up the `tty' and `tty-type' frame
   parameters from the currently selected terminal before the global
   default.

   (Done.)

-- Put all cached terminal escape sequences into struct tty_output.
   Currently, they are still stored in global variables, so we don't
   really support multiple terminal types.

   (Done.  It was not fun.)

-- Implement sane error handling after initialization.  (Currently
   emacs exits if you specify a bad terminal type.)  The helpful error
   messages must still be provided when Emacs starts.

   (Done.)

-- Implement terminal deletion, i.e., deleting local frames, closing
   the tty device and restoring its previous state without exiting
   Emacs.

   (Done, but at the moment only called when an error happens during
   initialization.  There is a memory corruption error around this
   somewhere.)  (Update: now it is fully enabled.)

-- Implement automatic deletion of terminals when the last frame on
   that terminal is closed.

   (Done.)

-- Restore tty screen after closing the terminal.

   (Done, we do the same as Emacs 21.2 for all terminals.)

-- 'TERM=dumb src/emacs' does not restore the terminal state.

   (Done.)

-- C-g should work on secondary terminals.

   (Done, but the binding is not configurable.)

-- Deal with SIGHUP in Emacs and in emacsclient.  (After this, the
   server-frames may be removed from server.el.)

   (Done, nothing to do.  It seems that Emacs does not receive SIGHUP
   from secondary ttys, which is actually a good thing.)  (Update: I
   think it would be a bad idea to remove server-frames.)

-- Change emacsclient/server.el to support the -t argument better,
   i.e. automatically close the socket when the frame is closed.

   (Seems to be working OK.)

-- Fix mysterious memory corruption error with tty deletion.  To
   trigger it, try the following shell command:

        while true; do TERM=no-such-terminal-definition emacsclient -h; done

   Emacs usually dumps core after a few dozen iterations.  (The bug
   seems to be related to the xfreeing or bzeroing of
   tty_output.Wcm.  Maybe there are outside references to struct Wcm?
   Why were these vars collected into a struct before multi-tty
   support?)

   (Done.  Whew.  It turned out that the problem had nothing to do
   with hypothetical external references to Wcm, or any other
   tty_output component; it was simply that delete_tty closed the
   filehandles of secondary ttys twice, resulting in fclose doubly
   freeing memory.  Utterly trivial matter.  I love the C's memory
   management, it puts hair on your chest.)

-- Support raw secondary terminals.  (Note that SIGIO works only on
   the controlling terminal.) Hint: extend read_input_waiting for
   multiple ttys and hopefully this will be fixed.

   (Done, it seems to have been working already for some time.  It
   seems F_SETOWN does work, after all.  Not sure what made it fail
   earlier, but it seems to be fixed (there were several changes
   around request_sigio, maybe one of them did it).
   read_input_waiting is only used in sys_select, don't change
   it.)  (Update: After adding X support, it's broken again.)

-- Find out why does Emacs abort when it wants to close its
   controlling tty.  Hint: chan_process[] array.  Hey, maybe
   noninterrupt-IO would work, too?  Update: no, there is no process
   for stdin/out.

   (Done.  Added add/delete_keyboard_wait_descriptor to
   term_init/delete_tty.  The hint was right, in a way.)

-- Issue with SIGIO: it needs to be disabled during redisplay.  See if
   fcntl kernel behaviour could be emulated by emacsclient.

   (Done.  Simply disabled the SIGIO emulation hack in emacsclient.)
   (Update: it was added back.)

-- server.el: There are issues with saving files in buffers of closed
   clients.  Try editing a file with emacsclient -f, and (without
   saving it) do a delete-frame.  The frame is closed without
   question, and a surprising confirmation prompt appears in another
   frame.

   (Done.  delete-frame now asks for confirmation if it still has
   pending buffers, and modified buffers don't seem to be deleted.)

-- emacsclient.el, server.el: Handle eval or file open errors when
   doing -t.

   (Done.)

-- Make parts of struct tty_output accessible from Lisp.  The device
   name and the type is sufficient.

   (Done, see frame-tty-name and frame-tty-type.)

-- Export delete_tty to the Lisp environment, for emacsclient.

   (Done, see delete-tty.)

-- Get rid of the accessor macros in termchar.h, or define macros for
   all members.

   (Done.)

-- Move device-specific parameters (like costs) commonly used by
   device backends to a common, device-dependent structure.

   (Done.  See struct display_method in termhooks.h.)

-- Fix X support.

   (Done.  Well, it seems to be working.)

-- Allow simultaneous X and tty frames.  (Handling input could be
   tricky.  Or maybe not.)

   (Done.  Allowed, that is.  It is currently extremely unstable, to
   the point of being unusable.  The rif variable causes constant
   core dumps.  Handling input is indeed tricky.)

-- Rewrite multi-tty input in terms of MULTI_KBOARD.

   (Done.  In fact, there was no need to rewrite anything, I just
   added a kboard member to tty_display_info, and initialized the
   frame's kboard from there.)

-- Fix rif issue with X-tty combo sessions.  IMHO the best thing to do
   is to get rid of that global variable (and use the value value in
   display_method, which is guaranteed to be correct).

   (Done, did exactly that.  Core dumps during combo sessions became
   much rarer.  In fact, I have not yet met a single one.)

-- Add multi-tty support to talk.el.

   (Done.)

-- Clean up the source of emacsclient.  It is a mess.

   (Done, eliminated stupid proxy-pty kludge.)

-- Fix faces on tty frames during X-tty combo sessions.  There is an
   init_frame_faces call in init_sys_modes, see if there is a problem
   with it.

   (Done, there was a stupid mistake in
   Ftty_supports_face_attributes_p. Colors are broken, though.)

-- C-x 5 2, C-x 5 o, C-x 5 0 on an emacsclient frame unexpectedly
   exits emacsclient.  This is a result of trying to be clever with
   delete-frame-functions.

   (Fixed, added delete-tty-after-functions, and changed server.el to
   use it.)

-- Something with (maybe) multi-keyboard support broke function keys
   and arrows on ttys during X+tty combo sessions.  Debug this.

   (I can't reproduce it, maybe the terminal type was wrong.)

-- Fix input from raw ttys (again).

   (Now it seems to work all right.)

-- During an X-tty combo session, a (message "Hello") from a tty frame
   goes to the X frame.  Fix this.

   (Done.  There was a safeguard against writing to the initial
   terminal frame during bootstrap which prevented echo_area_display
   from working correctly on a tty frame during a combo session.)

-- If there are no frames on its controlling terminal, Emacs should
   exit if the uses presses C-c there.

   (Done, as far as possible.  See the SIGTERM comment in
   interrupt_signal on why this seems to be impossible to solve this
   in general.)

-- During an X session, Emacs seems to read from stdin.  Also, Emacs
   fails to start without a controlling tty.

   (Fixed by replacing the troublesome termcap display with a dummy
   bootstrap display during bootstrap.

-- Do tty output through struct display, like graphical display
   backends.

   (Done.)

-- Define an output_initial value for output_method for the initial
   frame that is dumped with Emacs.  Checking for this frame (e.g. in
   cmd_error_internal) is ugly.

   (Done, broking interactive temacs.)

-- The command `emacsclient -t -e '(delete-frame)'' fails to exit.

   (Fixed.)

-- frame-creation-function should always create a frame that is on the
   same display as the selected frame.  Maybe frame-creation-function
   should simply be removed and make-frame changed to do the right
   thing.

   (Done, with a nice hack.  frame-creation-function is now frame-local.)

-- Fix C-g on raw ttys.

   (Done.  I disabled the interrupt/quit keys on all secondary
   terminals, so Emacs sees C-g as normal input.  This looks like an
   overkill, because emacsclient has extra code to pass SIGINT to
   Emacs, so C-g should remain the interrupt/quit key on emacsclient
   frames.  See the next entry why implementing this distinction would
   be a bad idea.)

-- Make sure C-g goes to the right frame with ttys.  This is hard, as
   SIGINT doesn't have a tty parameter. :-(

   (Done, the previous change fixes this as a pleasant side effect.)

-- I have seen a case when Emacs with multiple ttys fell into a loop
   eating 100% of CPU time.  Strace showed this loop:

        getpid()                                = 30284
        kill(30284, SIGIO)                      = 0
        --- SIGIO (I/O possible) @ 0 (0) ---
        ioctl(6, FIONREAD, [0])                 = -1 EIO (Input/output error)
        ioctl(5, FIONREAD, [0])                 = -1 EIO (Input/output error)
        ioctl(0, FIONREAD, [0])                 = 0
        sigreturn()                             = ? (mask now [])
        gettimeofday({1072842297, 747760}, NULL) = 0
        gettimeofday({1072842297, 747806}, NULL) = 0
        select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
        select(9, [0 3 5 6], NULL, NULL, {0, 0}) = 2 (in [5 6], left {0, 0})
        gettimeofday({1072842297, 748245}, NULL) = 0

   I have seen something similar with a single X frame, but have not
   been able to reproduce it for debugging.

   Update: This may have been caused by checking for nread != 0
   instead of nread > 0 after calling read_socket_hook in
   read_avail_input.

   (Fixed.  This was caused by unconditionally including stdin in
   input_wait_mask in init_process.  The select call in
   wait_reading_process_input always returned immediately, indicating
   that there is pending input from stdin, which nobody read.

   Note that the above strace output seems to be an unrelated but
   similar bug.  I think that is now fixed.)

-- Exiting Emacs while there are emacsclient frames doesn't restore the
   ttys to their default states.

   (This seems to be fixed by some previous change.)

-- Allow opening an X session after -nw.

   (Done.)

-- Fix color handling during tty+X combo sessions.  (It seems that tty
   sessions automatically convert the face colors to terminal colors
   when the face is loaded.  This conversion must happen instead on
   the fly in write_glyphs, which might be problematic, as color
   approximation is currently done in lisp (term/tty-colors.el).)
   (Update: hm, colors seem to work fine if I start emacs with -nw and
   then create an X frame.  Maybe it's just a small buglet somewhere.)

   (Seems to be fixed.  The problem was in startup.el, it did not
   initialize tty colors when the initial window system was
   graphical.)

-- emacs -nw --eval '(y-or-n-p "Foobar")' segfaults.  (Reported by
   Romain Francoise)

   (Fixed, there was a keyboard initialization problem.)

-- Fix interactive use of temacs.  There are face-related SEGVs, most
   likely because of changes in realize_default_face, realize_face.

   (Fixed.)

-- Don't exit Emacs when the last X connection fails during a
   multi-display session.

   (Fixed.)


-- Dan Nicolaescu noticed that starting emacsclient on the same
   terminal device that is the controlling tty of the Emacs process
   gives unexpected results.

   (Fixed.)

-- Istvan Marko reported that Emacs hang on ttys if it was started
   from a shell script.

   (Fixed.  There was a bug in the multi-tty version of
   narrow_foreground_group.  tcsetpgrp blocks if it is called from a
   process that is not in the same process group as the tty.)

-- emacsclient -t from an Emacs term buffer does not work, complains
   about face problems.  This can even lock up Emacs (if the recursive
   frame sets single_kboard).  Update: the face problems are caused by
   bugs in term.el, not in multi-tty.  The lockup is caused by
   single_kboard mode, and is not easily solvable.  The best thing to
   do is to simply refuse to create a tty frame of type `eterm'.

   (Fixed, changed emacsclient to check for TERM=eterm.  The face
   complaints seem to be caused by bugs in term.el; they are not
   related to multi-tty.)

-- Find out the best way to support suspending Emacs with multiple
   ttys.  My guess: disable it on the controlling tty, but from other
   ttys pass it on to emacsclient somehow.  (It is (I hope) trivial to
   extend emacsclient to handle suspend/resume.  A `kill -STOP' almost
   works right now.)

   (Done.  I needed to play with signal handling and the server
   protocol a bit to make emacsclient behave as a normal UNIX program
   wrt foreground/background process groups.)


;;; arch-tag: 8da1619e-2e79-41a8-9ac9-a0485daad17d