bug#39977: 28.0.50; Unhelpful stack trace

[Eli Zaretskii]

> Cc: enometh@meer.net, 39977@debbugs.gnu.org
> From: martin rudalics <rudalics@gmx.at>
> Date: Sat, 21 Mar 2020 10:32:24 +0100
> 
>  >> I'm afraid that we already might mishandle some of those simple cases.
>  >
>  > That just makes my point stronger, doesn't it?
> 
> Not really.  It's easy for delete_frame to refuse deleting a frame right
> at the beginning.  But once it has accepted a deletion, it might become
> hard to deal with all the consequences.

I don't think I understand where you are going with this.

>  >>   > The problem is how to do this without breaking legitimate code.  For
>  >>   > example, changing the window configuration temporarily, then changing
>  >>   > it back is quite legitimate,
>  >>
>  >> Right in the middle of redisplay, while constructing the mode line or
>  >> the title format?
>  >
>  > Why not?  As long as things are back as they were by the time :eval
>  > returns, I see no reason to disallow such code.
> 
> Such a change in the window configuration would take place in a state
> where certain variables have temporary settings only.  Selected frame,
> selected window and current buffer have been set by redisplay in a fast,
> improvised manner.  I would never trust the outcome of save_window_save
> or 'set-window-configuration' in such a state.

This isn't about trust.  This is about letting users' Lisp do anything
they want as long as the results allow redisplay to continue after
that Lisp returns.  I don't think it's right to disallow certain
actions just because they _might_ cause problems.

>  > No, they are there in cases where we simply don't know how to
>  > continue.
> 
> If that's the reason, then SELECTED_FRAME can easily set selected_frame
> to some live frame and continue.

Something like that, yes.