oddity w/ current ediff package and multiple frames
Howdy all.
I was wondering if anyone else has seen the following behavior (this is with
21.1.14 under Solaris 2.6, non mule, latest official packages from
xemacs.org). Basically ediff w/ the "plain" setup messes with other frames if
multiple frames are open when toggling the 'quick help' feature.
Here's the formula to reproduce:
xemacs -q
eval this in *scratch*:
(defun my-ediff-mode-hook ()
(interactive)
(setq-default ediff-window-setup-function 'ediff-setup-windows-plain)
(setq-default ediff-split-window-function 'split-window-horizontally)
(setq ediff-use-toolbar-p nil)
)
(add-hook 'ediff-mode-hook 'my-ediff-mode-hook)
open a new frame, go back to original one.
M-x ediff-files (pick any 2 files)
in the ediff control panel hit '?' repeatedly and watch what happens.
For me the other frame becomes the ediff setup and the help window toggles
there, in the original frame where I expected things to work, it will
alternately show the quick help text without resizing the window and with
resizing the window, and doing next/prev diff doesn't move the display
properly. It also sends the focus over to that other frame (w/ fvwm2 & click
to focus) but that comes back when the alternating of the resizing the control
buffer.
The ediff that comes w/ vanilla Emacs 20.7 doesn't exhibit this (in Emacs
20.7), but if I load the current version (I pointed the load path down a
directory with the diff files symlinked from the xemacs-packages tree to test
this) it exhibits similar behavior (affects other frame, but the resizing
behavior doesn't alternate, it always resizes the other frames control panel,
and the focus doesn't shift).
Then I tested it w/ 21.4.1, and even stranger things occurred...new frames
ended up being opened for me, including one that contained the horizontally
split ediff buffers but not the control panel!
So it sounded to me like this was tickling some XEmacs bugs as well...
Unfortunately I can't investigate this too closely right now, but thought I'd
send an email and perhaps Michael might know what's going on without too much
thought/time involved in the fixes for the ediff code and figured that quite
likely somone else on the list will have an idea about the potential XEmacs
bugs being shown.
Thanks,
Chuck
--
Charles K. Hines <ckh(a)requesttech.com> <hines(a)gderome.com>
Principal Scientist at ReQuest Technologies Inc (http://www.ReQuestTech.com/)
Martial Arts Instructor [Modern Arnis and Balintawak Escrima]
"Go back to sleep, Chuck. You're just havin' a nightmare
-- of course, we ARE still in Hell." (Gary Larson)