At 04:32 PM 5/3/00 -0400, John A. Turner wrote:
>>>>> "AP" == Andy Piper
<andy(a)xemacs.org>:
JAT> cvs-update'd and rebuilt yesterday, still crashing periodically...
AP> John,
AP>
AP> Can you try again with the latest cvs. I have fixed a whole bunch of GC
AP> and window/frame deletion bugs and I'm hoping this might be gone - or
AP> at least give you an assert instead.
unfortunately no - just got a spectacular crash when I clicked on a gutter
tab
*sigh*. Ok, I still have some extra error checking to put in on the X side.
running under gdb now to see if I can get a stack
also seeing the same display oddities I think I've mentioned before - here
are some snapshots
although I set the buffers-tab face to this:
'(buffers-tab ((t (:foreground "firebrick4" :background "gray60"
:size
"10" :family "helvetica" :italic t))) t))
when it starts I see this:
Attachment Converted: "c:\eudora\attach\snap4.jpg"
Ok, I'll try this. This is similar to a bug Stephen Turnbull has reported
which I'm still looking at.
then as soon as I do something, e.g. C-x, it initializes, and I see:
Attachment Converted: "c:\eudora\attach\snap21.jpg"
note the Motif-blue line beneath the gutter
This is expected. The face font has changed height but the gutter height
hasn't been recalculated. I convinced myself this was a hard problem to fix
but I can't remember why. If your test case is reproducible for me I'll see
what I can do. I just wonder if this is part of the problem. Can you try
running without setting the tab face at all and see if its any more stable.
then, after I finish the command, say loading my home dir into dired,
it
goes funny:
Attachment Converted: "c:\eudora\attach\snap31.jpg"
That's worrying.
then once it redisplays again, it's fine
but it still sometimes displays in the wrong color, as with the frame in
which I'm composing this msg:
Attachment Converted: "c:\eudora\attach\snap41.jpg"
You've reported this before, I'm still not sure why it happens.
but the crashes are much more important than this stuff, so I'll
send a
stack once I get one...
Yup. They have to be nailed. I will try and put together an X error
checking patch for you to try.
andy
--------------------------------------------------------------
Dr Andy Piper
Principal Consultant, BEA Systems Ltd