Re: Debouncing slow mode line constructs (was: Emacs design and architecture. How about copy-on-write?)

[Eli Zaretskii]

> From: Ihor Radchenko <yantar92@posteo.net>
> Cc: dmitry@gutov.dev, luangruo@yahoo.com, acm@muc.de, incal@dataswamp.org,
>  emacs-devel@gnu.org
> Date: Fri, 22 Sep 2023 12:50:39 +0000
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> > Though if we just do this silently, it can hide performance problems, 
> >> > both discouraging the mode-line authors from fixing them, and creating 
> >> > odd behaviors (from the user's POV) when something which should change, 
> >> > doesn't.
> >> 
> >> A warning may be displayed when this "debouncing" is activated.
> >
> > You cannot easily display any warnings from redisplay.
> 
> Can't the code inside :eval display warnings?

No, because displaying a warning requires redisplay, and we don't
support recursive redisplays.

> In any case, it is not a big problem to arrange the warning to be
> displayed after redisplay finishes.

I don't see a reason to do that with this particular kind of problems.
We currently don't display anything by default even if Lisp called by
redisplay signals an error, we just log that in *Messages*.  Why
should these :eval problems we treated differently?