Re: Edebug corrupting point in buffers.

[Eli Zaretskii <eliz@gnu.org>]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Cc: acm@muc.de,  emacs-devel@gnu.org
> Date: Thu, 03 Nov 2022 16:39:23 -0400
> 
> > I think before digging into the reasons, we should decide what kind of
> > behavior we would consider "correct" and/or "useful" in the relevant
> > use cases with Edebug.  The answer is not easy, because AFAIU Edebug
> > cannot easily know which window(s) and which buffer(s) are affected by
> > the program being debugged.
> 
> If we follow the model or "traditional debuggers" which run in
> a separate process, then it would make a lot of sense for Edebug to
> save&restore points in all the buffers that it (Edebug) touches, so as
> to better preserve the behavior we get when Edebug is not invoked.
> 
> For that Edebug doesn't need to know which buffers are affected by the
> program being debugged, it just needs to know the buffers that it
> (itself) affects, which doesn't sound impractically difficult.

What do you mean by "the buffers that Edebug touches", exactly?
"Touches" or "affects" in what sense?

The next question is "how would Edebug know which buffers it touches?"

And the next one after that would be "what about buffers Edebug
touches that are displayed in more than one window?"