Re: set-window-configuration vs error-checking sub windows
At 10:02 AM 5/26/00 +0100, Jan Vroonhof wrote:
Andy Piper <andy(a)xemacs.org> writes:
> I suppose it might be worth rewriting your sledgehammer check to validate
> that for every instance in the frame cache its domain is (a) a window and
> (b) in that window's cache.
I just did this (code at end)
I stuck it in Feval, Ffuncall, Fset_window_configuration and
before/after garbage_collect[1].
It hits in Feval all the time while starting up Gnus. I can sometimes
get it too when I make buffers with C-x b and play a lot with windows.
You mean it asserts? I'll try your code and see if I can reproduce both
under cygwin and under linux.
Still no clue what could cause it. Could you review my check?
Looks sane to me. I guess the other thing to check is that instances are
not occuring twice.
andy
--------------------------------------------------------------
Dr Andy Piper
Principal Consultant, BEA Systems Ltd