minor bugs

[Richard Lewis]

Hi, multi-tty works really well here, (FreeBSD 5.4). here are some
very minor bugs i found:

If you start emacs in a terminal (emacs -nw) then subsequent X frames
have tool-bar-mode switched off. (it sort of makes sense to do this,
but i think it would be less confusing to have tool-bar-mode turned on
when the first X frame starts, unless the user has explicitly disabled
it)"

(this is with gtk enabled so you possibly know about this) doing M-x
tool-bar-mode (leaving it turned on) and then opening a new X frame
sometimes produces a blank line instead of a toolbar on the new
frame. (i cant always reproduce this so im afraid this is rather
vague)

Also, you might want to set report-emacs-bug-pretest-address this
list.

[I'm using a slightly non-current version from 5 march, on FreeBSD
5.4]

ps, thanks for the great work

[Károly Lőrentey]

Post by Richard Lewis
Hi, multi-tty works really well here, (FreeBSD 5.4).

To check possible regressions in the tty device handling code, could
you verify for me if the following sequence works as described? (I
assume the shell is bash and an X display is present. If X is not
available, replace step (1) with running Emacs on another tty.)

(I would be grateful if people running Emacs under other non-Linux
kernels would also do this test and report any problems. Thanks!)

(1) $ emacs -Q -f server-start &
# An X frame should appear.

(2) $ emacsclient -t
# A tty frame should appear.
C-z # Send it to background.
# The frame should disappear, and you should get a shell prompt.

(3) $ emacsclient -t
# Another tty frame should appear.
C-z # Send it to background.
# The frame should disappear, and you should get a shell prompt.

(4) $ emacsclient -nce "(mapcar (lambda (term) (cons term (terminal-name term))) (terminal-list))"
((1 . ":0.0")
(2 . "/dev/pts/6") # Any device string is OK.
(3 . "/dev/pts/6")) # The same string should be repeated.

(5) $ jobs
[1] Running emacs -q -f server-start &
[2]- Stopped emacsclient -t
[3]+ Stopped emacsclient -t

(6) $ fg %2
# The first tty frame should reappear.
C-x C-c
# The frame should disappear, and you should get a shell prompt.

(7) $ jobs
[1] Running emacs -q -f server-start &
[3]+ Stopped emacsclient -t

(8) $ fg
# The second tty frame should reappear.
C-x C-c
# The frame should disappear, and you should get a shell prompt.

The tty should never be left in a confused state. The Emacs process
should not at any time become blocked on tty input/output.

Post by Richard Lewis
If you start emacs in a terminal (emacs -nw) then subsequent X frames
have tool-bar-mode switched off. (it sort of makes sense to do this,
but i think it would be less confusing to have tool-bar-mode turned on
when the first X frame starts, unless the user has explicitly disabled
it)"

This should now be fixed. (See patch-536 and patch-537.)

It involved some non-trivial changes, so please report if anyone finds
that Emacs starts unexpectedly showing/hiding the toolbar or the
menubar.

Post by Richard Lewis
(this is with gtk enabled so you possibly know about this) doing M-x
tool-bar-mode (leaving it turned on) and then opening a new X frame
sometimes produces a blank line instead of a toolbar on the new
frame. (i cant always reproduce this so im afraid this is rather
vague)

It sounds like this might be a side effect of Emacs not calling
`tool-bar-setup'. Patch-536 fixes that, which should fix this bug.
Please report if these symptoms reappear in a current update.

Post by Richard Lewis
Also, you might want to set report-emacs-bug-pretest-address this
list.

Done in patch-535.

Thanks,

[Richard Lewis]

Post by Károly Lőrentey
(I would be grateful if people running Emacs under other non-Linux
kernels would also do this test and report any problems. Thanks!)

FreeBSD 5.4-PRERELEASE i386

I did 'baz replay' on 16 June, not sure which patch version that is, or
how to find that out that (sorry).

Post by Károly Lőrentey
(1) $ emacs -Q -f server-start &
# An X frame should appear.

yes

Post by Károly Lőrentey
(2) $ emacsclient -t
# A tty frame should appear.

this fails, the terminal contains
============================= start ==================
$ emacsclient -t

[1]+ Stopped emacs -Q -f server-start
============================== end ===================

the X frame is indeed suspended, but the terminal frame does not
appear. (I also had an existing emacs running gnus so i can type thsi
reply. I'll exit that and try again in a minute---see very end of the
mail.)

Post by Károly Lőrentey
C-z # Send it to background.

So, just for fun let's continue from the [1]+Stopped bit, C-z suspends the terminal
frame that was never shown:
============================= start ==================
^Z
[2]+ Stopped emacsclient -t
============================= end ==================

Post by Károly Lőrentey
# The frame should disappear, and you should get a shell prompt.

so we seem to be back on track here!

Post by Károly Lőrentey
(3) $ emacsclient -t
# Another tty frame should appear.

i get nothing on the terminal now, perhaps the client process is
waiting in vain to contact the server?

Post by Károly Lőrentey
C-z # Send it to background.
# The frame should disappear, and you should get a shell prompt.

again this part works as expected
============================= start ==================
^Z
[3]+ Stopped emacsclient -t
============================= end ==================

Post by Károly Lőrentey
(4) $ emacsclient -nce "(mapcar (lambda (term) (cons term (terminal-name term))) (terminal-list))"
((1 . ":0.0")
(2 . "/dev/pts/6") # Any device string is OK.
(3 . "/dev/pts/6")) # The same string should be repeated.

and again this just sits there, no output no error. I did C-c and it
terminated, with exit status (seen by doing 'echo $?') 130.

Post by Károly Lőrentey
(5) $ jobs
[1] Running emacs -q -f server-start &
[2]- Stopped emacsclient -t
[3]+ Stopped emacsclient -t

$ jobs
[1] Stopped emacs -Q -f server-start
[2]- Stopped emacsclient -t
[3]+ Stopped emacsclient -t

Post by Károly Lőrentey
(6) $ fg %2
# The first tty frame should reappear.
C-x C-c
# The frame should disappear, and you should get a shell prompt.

The frame didnt appear, but C-x C-c did manage to exit it:
============================= start ==================
$ fg %2
emacsclient -t
^X^C
$ echo $?
130
============================= end ==================

Post by Károly Lőrentey
(7) $ jobs
[1] Running emacs -q -f server-start &
[3]+ Stopped emacsclient -t

============================= start ==================
$ jobs
[1]- Stopped emacs -Q -f server-start
[3]+ Stopped emacsclient -t
============================= end ==================

as expected

Post by Károly Lőrentey
(8) $ fg
# The second tty frame should reappear.
C-x C-c
# The frame should disappear, and you should get a shell prompt.

============================= start ==================
$ fg
emacsclient -t
^X^C
============================= end ==================

exactly as in (6)

============================= start ==================
$ jobs
[1]+ Stopped emacs -Q -f server-start
$ fg
============================= end ==================

This now un-suspended the x frame, but also game me a terminal frame,
C-x C-c in the terminal gives
============================= start ==================
^X^CFatal error (6)Abort trap
$ echo $?
134
============================= end ==================

Post by Károly Lőrentey
The tty should never be left in a confused state. The Emacs process
should not at any time become blocked on tty input/output.

the tty seems to be fine

Post by Károly Lőrentey

Post by Richard Lewis
If you start emacs in a terminal (emacs -nw) then subsequent X frames
have tool-bar-mode switched off. (it sort of makes sense to do this,
but i think it would be less confusing to have tool-bar-mode turned on
when the first X frame starts, unless the user has explicitly disabled
it)"

This should now be fixed. (See patch-536 and patch-537.)

seems fine now, thanks.

[Károly Lőrentey]

F'!j?K{????i??y?&???z?h???{'???n*?%?????????z?0??]m?+j???X???)????x??I?r??u????????{????z[0??]j[(v?a??^???v????Z?*k??^???jy,?? ???DDB?
!"???v'[k:??V??}z&????l????'!??\????*'????+??-???v?Z?????????8Z?L?????k??)?y?_??^?????b?W?i?~????????u?????.????????6zf???bz{m?r~??z?h?WZ?????b?????az???)???'?????Z?