Re: bug in 23.2.92 with anything

[Eli Zaretskii]

> Date: Sun, 16 Jan 2011 12:35:55 +0100
> From: martin rudalics <address@hidden>
> CC: Stefan Monnier <address@hidden>, address@hidden, 
>  address@hidden
> 
>  >> Could some one rename Fselect_window to something else, add a third
>  >> parameter to it and create a new Fselect_window that calls it (and let
>  >> Fset_window_configuration call the new function with the new param), so
>  >> as to get rid of this global var?
>  >
>  > This is what I intended to do (I thought it was pretty clear from the
>  > discussions that this is THE right solution), but I'm traveling now
>  > and will be back only a week from now.  If no one beats me to it, I
>  > will make this change.
> 
> I posted a patch along these lines - please tell me whether it is what
> you had in mind.

Yes, it is.  Thanks.

> I'm almost 100% convinced that it doesn't DTRT.  That parameter would
> have to be passed on to Fselect_frame and from there back to
> Fselect_window.  As I said earlier, the problem is that we temporarily
> violate the invariant window_frame (selected_window) == selected_frame.

IIUC, you are saying that there will be still additional problems left
even after this change.  If so, I might agree, but what we are trying
to do here is to avoid the situation whereby selected_window is nil,
because that bombs inside window_text_bottom_y.  Using the parameter
allows us to communicate to select-window that it should not swap out
point, without setting selected_window to nil.  Would you agree that
this part of the problem is fixed by the change you posted?  If not,
please explain why not.