Re: 27.0.50: How can I test a buffer-local window-configuration-change-hook in batch mode?

[Stefan Monnier]

> Thanks.  However, unless these emulators are components of a
> garden-variety system out there, using them for Emacs testing would
> mean that only a few of us could ever run these tests.

Yes, it's not a real solution, especially since we don't even have
a "sample" to point to for other people to write&run such tests.

Maybe a better approach would be to allow Emacs to run the redisplay
in batch mode (i.e. build glyph matrices and just do nothing for the
actual redraw).

> Also, using a display emulator for such testing would most probably
> mean we cannot verify the correctness of the resulting display in some
> automated way, as we do with ert facilities.

Indeed, it doesn't solve all display-testing problems.  It would be good
enough for the window-configuration-change-hook case and a few others
(e.g. testing window-start and window-end positions, or pixel position
of particular chunks of text), but to do anything more we'd probably
have to provide ways to examine the glyph matrix.

> Bottom line: I think using these emulators is only a very partial
> solution to being able to test our display code.

I only meant it for people who want to start writing such tests, in the
absence of actual support for that.


        Stefan