Re: [Emacs-diffs] master 6a69725: Fix redisplay of frame title when current buffer is killed

Re: [Emacs-diffs] master 6a69725: Fix redisplay of frame title when current buffer is killed

[Stefan Monnier]

> +  /* Killing a buffer might have global effects which require
> +     redisplaying frames.  For example, if the buffer's name appears
> +     in the frame title.  */
> +  windows_or_buffers_changed = 11;

Hmm... this means that using with-temp-buffer will cause
a full redisplay.  That's bad.

Whats the scenario where this was needed?  I mean: if the buffer's name
appears in the frame, it's presumably because it's displayed in one of
its windows, so place to catch this is when that window's buffer is
changed, rather than when a buffer is killed.


        Stefan

Re: [Emacs-diffs] master 6a69725: Fix redisplay of frame title when current buffer is killed

[Eli Zaretskii]

> From: Stefan Monnier <monnier@IRO.UMontreal.CA>
> Date: Sat, 24 Oct 2015 09:46:07 -0400
> Cc: Eli Zaretskii <eliz@gnu.org>
> 
> > +  /* Killing a buffer might have global effects which require
> > +     redisplaying frames.  For example, if the buffer's name appears
> > +     in the frame title.  */
> > +  windows_or_buffers_changed = 11;
> 
> Hmm... this means that using with-temp-buffer will cause
> a full redisplay.  That's bad.

Yes, which is why I fixed it differently in bc0b340, which see.

> Whats the scenario where this was needed?  I mean: if the buffer's name
> appears in the frame, it's presumably because it's displayed in one of
> its windows, so place to catch this is when that window's buffer is
> changed, rather than when a buffer is killed.

I thought about that.  But changing the name of the frame's selected
buffer is just the tip of the iceberg: the frame title accepts all the
%-constructs that a mode line accepts, and so should be considered for
redisplay whenever we consider mode lines for redisplay.  So I
eventually set update_mode_lines there, as that is already handled in
x_consider_frame_title.