GNU bug report logs -
#57241
Commit b87400c broke master
Previous Next
Full log
View this message in rfc822 format
On 8/16/2022 6:42 AM, Eli Zaretskii wrote:
>> From: Dario Gjorgjevski <dario.gjorgjevski <at> gmail.com>
>> Date: Tue, 16 Aug 2022 15:23:58 +0200
>> Cc: 57241 <at> debbugs.gnu.org, jporterbugs <at> gmail.com
>>
>> On Tue, Aug 16, 2022 at 3:17 PM Eli Zaretskii <eliz <at> gnu.org> wrote:>
>>> Can you (load "fringe") in such a build? If not, a simple 'require'
>>> won't help.
>>
>> Yes, I can.
>
> OK, I tried to fix this now.
Maybe it would make sense for fringe.el to be loaded in all builds, but
to have more things be stub implementations like 'define-fringe-bitmap'
is? Or maybe 'fringe-custom-set-bitmap' should just be autoloaded?
'require' works here, but if other places start using
'fringe-custom-set-bitmap' they'll break --without-x builds too. For
example, gdb-mi.el and edebug.el seem like good candidates for
eventually doing something like bookmark.el does.
> Jim, I wonder whether the default value should be nil on TTY frames.
It could make sense on --without-x builds, but I don't think it should
depend on the frame type. With the Emacs daemon, you might switch
between TTY and GUI frames, and it would be strange if you didn't see
the fringe indicator in a GUI frame if you had set the indicator in a
TTY frame.
That said, on --without-x builds, setting a fringe indicator on a
particular line of text should be safe, right? It just doesn't display
anything, I think. Previously, I believe the code would (try to) set the
fringe indicator on --without-x builds, and it seems to not be a
problem, or else someone would have filed a bug about that against Emacs 28.
This bug report was last modified 2 years and 264 days ago.
Previous Next
GNU bug tracking system
Copyright (C) 1999 Darren O. Benham,
1997,2003 nCipher Corporation Ltd,
1994-97 Ian Jackson.