[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: RFC: flicker-free double-buffered Emacs under X11

From: Daniel Colascione
Subject: Re: RFC: flicker-free double-buffered Emacs under X11
Date: Thu, 27 Oct 2016 16:10:21 -0700
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/26.0.50 (gnu/linux)

Paul Eggert <address@hidden> writes:

> Thanks for working on this.
> I tried building with that patch, and had problems because your email
> client munged the patch. I cleaned it up by hand and arrived at the
> attached patch instead. (It needs a better commit message of course.)
> Maybe next time you could attach the output of git-format-patch.

I need to get Gnus working on that machine. I'd hoped Squirrelmail would
have been too simplistic to munge the patch, but I was wrong.

> Although the resulting system builds under Fedora 24 x86-64 with GTK,
> I can't tell any difference in flickering. I don't see much flicker
> without the patch, or with it. (I am using X, not Wayland.) Maybe my
> eyes are not good enough....

I've found the flickering while editing to be very machine-dependent. It
was driving me nuts on mine, which is why I wrote the patch. Flickering
while resizing seems to be a more general problem though.

I've found I can reliably reproduce the flickering by isearching for a
string that appears in a buffer only once, then leaning on C-s.

> I don't see the need to expose double-buffering to Lisp via
> x-double-buffered-p or inhibit-double-buffering. Unless there's a
> demonstrated need, I'd remove these from the Lisp interface, to
> simplify it.

I'll feel more comfortable about removing the lisp interface (and so any
way for users to this the thing off) once we're sure it's not
causing problems.

> The configure.ac change should use AC_CACHE_CHECK.

None of the other X extension checks do. I'm sure AC_CACHE_CHECK is
great, but it feels out of scope.

> set_up_x_back_buffer should be a static function, since it's not used
> outside its module.

Thanks; it wasn't always internal to that file. I'll fix the function.

> In C code we typically use comment /* like this */, not // like
> this. Also, the spacing for pointer decls is typically 'type *id', not
> 'type * id' or 'type* id'.

Thanks. I thought I'd gotten rid of all of those. Speaking of comments
though: if we're C99-only anyway now, we might as well allow
C99 comments.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]