bug#10304: 24.0.92: display bug

[Lars Magne Ingebrigtsen]

Eli Zaretskii <eliz@gnu.org> writes:

>> Then the window fixed itself suddenly without me doing anything.
>
> Does the trace you posted include the time when the window
> fixed itself, and if so, can you try identifying in the trace the
> place where it happened?  Could it be here:
>
>   expose_frame (796, 696, 24, 484)
>   expose_window (796, 696, 24, 464)
>   expose_window (796, 1160, 24, 20)

I had switched to an xterm that was partially overshadowing the Emacs
frame, and typing away in that xterm.  And then suddenly the Emacs frame
updated itself and everything was OK.  So I'm not sure how that relates
to the blog, but it was after the glyph dump, at least.

>> Here's the output from the log file.
>
> I will study it, but it is important for me to understand how this log
> is related to whatever you did in Emacs.  Does this reflect only the
> "page back and forth, and the matrix dump", or does it include more
> stuff?

What I did was read the message (from the *Summary* buffer), hit SPC,
and the *Article* buffer scrolled forwards and had the wrong display.

Then I jumped to and xterm to read the instructions I had saved in a
file there.  Then I think I went back to the Emacs frame, typed `M-x
trace-redisplay', scrolled back and forth (with the bug still
displaying), then did a dump-glyph-matrix, then went back to the xterm,
and then Emacs spontaneously fixed itself, and then I went back to the
Emacs frame and switched the trace off.

I think.  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
  bloggy blog http://lars.ingebrigtsen.no/