but the problem still persists in texinfo-4.11, see:
https://bugzilla.redhat.com/show_bug.cgi?id=243971
Thanks very much for the report, testing, and patch.
I was unable to figure out exactly why the code in
window_new_screen_size (in window.c) was producing negative dimensions;
it's exceedingly painful to debug, since the window has to be resized
and that screws up the debugging output. Not to mention it doesn't
reliably crash, even with the same resize.
However, I installed a workaround in window.c to force the final window
width/height to zero if they do become negative. I also added the test
in display_update_one_window so it doesn't try to display a window
unless it has both width and height >= 1. Now I can't make it crash.
Please try out the version from CVS if you can, and/or add a note to
that bug report.
Thanks again,
Karl
