* Comint Completion Broken
@ 2006-05-16 10:48 Herbert Euler
2006-05-19 2:39 ` Herbert Euler
0 siblings, 1 reply; 6+ messages in thread
From: Herbert Euler @ 2006-05-16 10:48 UTC (permalink / raw)
Hello,
Comint in the newest Unicode 2 CVS Emacs is broken. In the past, if
comint shows a completion window, this window will be automatically
deleted after the user types a complement supplied in it or just types
space. Now it won't be deleted even if the user types SPC directly,
contradicts to what is said in echo area: "Type space to flush".
Comparing the current source code and the latest working source code,
I found the following change for `comint-dynamic-list-completions' in
lisp/comint.el:
2945,2948c2980,2982
< (unless (eq first ?\s)
< (setq unread-command-events (listify-key-sequence key)))
< (unless (eq first ?\t)
< (set-window-configuration
comint-dynamic-list-completions-config))))))
---
> (if (eq first ?\s)
> (set-window-configuration
>comint-dynamic-list-completions-config)
> (setq unread-command-events (listify-key-sequence key)))))))
The above one is the correct code, the below one is the wrong code.
Note that these two code aren't equal. If the current source is
substituted with the above one, comint will act correctly.
There's no such entry in changelog shows
`comint-dynamic-list-completions' is changed recently. Why do someone
alter the behavior of `comint-dynamic-list-completions'?
Regards,
Guanpeng Xu
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
^ permalink raw reply [flat|nested] 6+ messages in thread
* RE: Comint Completion Broken
2006-05-16 10:48 Comint Completion Broken Herbert Euler
@ 2006-05-19 2:39 ` Herbert Euler
2006-05-19 2:55 ` Miles Bader
0 siblings, 1 reply; 6+ messages in thread
From: Herbert Euler @ 2006-05-19 2:39 UTC (permalink / raw)
>Comint in the newest Unicode 2 CVS Emacs is broken. In the past, if
>comint shows a completion window, this window will be automatically
>deleted after the user types a complement supplied in it or just types
>space. Now it won't be deleted even if the user types SPC directly,
>contradicts to what is said in echo area: "Type space to flush".
>
>Comparing the current source code and the latest working source code,
>I found the following change for `comint-dynamic-list-completions' in
>lisp/comint.el:
>
>2945,2948c2980,2982
>< (unless (eq first ?\s)
>< (setq unread-command-events (listify-key-sequence key)))
>< (unless (eq first ?\t)
>< (set-window-configuration
>comint-dynamic-list-completions-config))))))
>
>---
>> (if (eq first ?\s)
>> (set-window-configuration
>>comint-dynamic-list-completions-config)
>> (setq unread-command-events (listify-key-sequence key)))))))
>
>The above one is the correct code, the below one is the wrong code.
>Note that these two code aren't equal. If the current source is
>substituted with the above one, comint will act correctly.
Excuse me, but is there somebody reading this? The main
branch and Unicode 2 branch share the same `comint.el',
I think the behavior of comint is same.
Typing a whole word and closing completion window manually
is not as convenient as typing a character and closing
completion window automatically, isn't it?
Regards,
Guanpeng Xu
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Comint Completion Broken
2006-05-19 2:39 ` Herbert Euler
@ 2006-05-19 2:55 ` Miles Bader
2006-05-19 6:03 ` Romain Francoise
0 siblings, 1 reply; 6+ messages in thread
From: Miles Bader @ 2006-05-19 2:55 UTC (permalink / raw)
Cc: emacs-devel
"Herbert Euler" <herberteuler@hotmail.com> writes:
> Excuse me, but is there somebody reading this? The main
> branch and Unicode 2 branch share the same `comint.el',
> I think the behavior of comint is same.
...
> Typing a whole word and closing completion window manually
> is not as convenient as typing a character and closing
> completion window automatically, isn't it?
I'm confused...
As far as I can see:
* comint.el is exactly the same in the trunk / emacs-unicode-2 branch
* the behavior of comint completion is exactly the same in both:
Typing a SPC deletes the completion window, typing anything else
leaves the completion window in place. [This seems like reasonable
behavior -- fairly often I _want_ to leave the window in place, so I
can select it and do more advanced searching or whatever.]
This is using up-to-date CVS checkouts of both the trunk and branch.
So there seems to be no problem. What are you asking about???
-miles
--
P.S. All information contained in the above letter is false,
for reasons of military security.
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: Comint Completion Broken
2006-05-19 2:55 ` Miles Bader
@ 2006-05-19 6:03 ` Romain Francoise
2006-05-19 6:18 ` Miles Bader
0 siblings, 1 reply; 6+ messages in thread
From: Romain Francoise @ 2006-05-19 6:03 UTC (permalink / raw)
Cc: Herbert Euler, emacs-devel
Miles Bader <miles.bader@necel.com> writes:
> * the behavior of comint completion is exactly the same in both:
> Typing a SPC deletes the completion window, typing anything else
> leaves the completion window in place. [This seems like reasonable
> behavior -- fairly often I _want_ to leave the window in place, so I
> can select it and do more advanced searching or whatever.]
The window used to get deleted automatically on key events, e.g.
u p TAB (window appears) t (window disappears) TAB
=> uptime
But that was apparently an accidental feature, and it caused a bug when
selecting completion candidates with mouse-1 in the window:
2006-03-27 Nick Roberts <nickrob@snap.net.nz>
* comint.el (comint-dynamic-list-completions): Allow user to
select *Completions* buffer.
--
Romain Francoise <romain@orebokech.com> | The sea! the sea! the open
it's a miracle -- http://orebokech.com/ | sea! The blue, the fresh, the
| ever free! --Bryan W. Procter
^ permalink raw reply [flat|nested] 6+ messages in thread* Re: Comint Completion Broken
2006-05-19 6:03 ` Romain Francoise
@ 2006-05-19 6:18 ` Miles Bader
2006-05-19 7:03 ` Herbert Euler
0 siblings, 1 reply; 6+ messages in thread
From: Miles Bader @ 2006-05-19 6:18 UTC (permalink / raw)
Cc: Herbert Euler, emacs-devel
Romain Francoise <romain@orebokech.com> writes:
>> Typing a SPC deletes the completion window, typing anything else
>> leaves the completion window in place. [This seems like reasonable
>> behavior -- fairly often I _want_ to leave the window in place, so I
>> can select it and do more advanced searching or whatever.]
>
> The window used to get deleted automatically on key events, e.g.
>
> u p TAB (window appears) t (window disappears) TAB
> => uptime
>
> But that was apparently an accidental feature, and it caused a bug when
> selecting completion candidates with mouse-1 in the window:
My memory is a bit fuzzy, but as I recall, the "really old" behavior
(what was in Emacs 21) was more or less the same the current behavior.
I'm not entirely sure of the whole ugly history after that, but for a
while, I think pcomplete was used for comint completion, which had the
"window always gets deleted" behavior. [but I can't find any record of
this in the ChangeLog, so maybe it's just an illusion I had???]
Whether or not it's due to pcomplete, until fairly recently comint
seemed to be doing the "window always gets deleted" thing, which was
extremely annoying; I'm assuming that's what Nick fixed.
Anyway, the current behavior seems both user-friendly and compatible
with Emacs 21. So I'd say things are in a good state.
-Miles
--
The secret to creativity is knowing how to hide your sources.
--Albert Einstein
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Comint Completion Broken
2006-05-19 6:18 ` Miles Bader
@ 2006-05-19 7:03 ` Herbert Euler
0 siblings, 0 replies; 6+ messages in thread
From: Herbert Euler @ 2006-05-19 7:03 UTC (permalink / raw)
Cc: romain, emacs-devel
>From: Miles Bader <miles.bader@necel.com>
>Reply-To: Miles Bader <miles@gnu.org>
>To: Romain Francoise <romain@orebokech.com>
>CC: Herbert Euler <herberteuler@hotmail.com>, emacs-devel@gnu.org
>Subject: Re: Comint Completion Broken
>Date: Fri, 19 May 2006 15:18:17 +0900
>
>Romain Francoise <romain@orebokech.com> writes:
> >> Typing a SPC deletes the completion window, typing anything else
> >> leaves the completion window in place. [This seems like reasonable
> >> behavior -- fairly often I _want_ to leave the window in place, so I
> >> can select it and do more advanced searching or whatever.]
> >
> > The window used to get deleted automatically on key events, e.g.
> >
> > u p TAB (window appears) t (window disappears) TAB
> > => uptime
> >
> > But that was apparently an accidental feature, and it caused a bug when
> > selecting completion candidates with mouse-1 in the window:
>
>My memory is a bit fuzzy, but as I recall, the "really old" behavior
>(what was in Emacs 21) was more or less the same the current behavior.
>
>I'm not entirely sure of the whole ugly history after that, but for a
>while, I think pcomplete was used for comint completion, which had the
>"window always gets deleted" behavior. [but I can't find any record of
>this in the ChangeLog, so maybe it's just an illusion I had???]
>
>Whether or not it's due to pcomplete, until fairly recently comint
>seemed to be doing the "window always gets deleted" thing, which was
>extremely annoying; I'm assuming that's what Nick fixed.
>
>Anyway, the current behavior seems both user-friendly and compatible
>with Emacs 21. So I'd say things are in a good state.
Well, in my opinion, deleting completion window automatically is
more user-friend. Perhaps one wants completion always at hand,
but if one has already finished inputing (indicated by inserting
a unique string in completion window or typing SPC), the
completion window should be deleted automatically. Otherwise,
one have to type C-x 1 to delete the completion window manually,
which is a bit boring.
And the most user-friend method seems to be, the completion
window gets updated and finally deleted as user inputing.
Regards,
Guanpeng Xu
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2006-05-19 7:03 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-05-16 10:48 Comint Completion Broken Herbert Euler
2006-05-19 2:39 ` Herbert Euler
2006-05-19 2:55 ` Miles Bader
2006-05-19 6:03 ` Romain Francoise
2006-05-19 6:18 ` Miles Bader
2006-05-19 7:03 ` Herbert Euler
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).