Re: ruler-mode is broken

[Luc Teirlinck]

I did not follow this thread.  However, does ruler mode, directly or
indirectly, use any of the functions `window-at' or
`coordinates-in-window-p'?  To me, these functions seem broken on
GNU/Linux.  (They work fine on MS Windows, so if ruler-mode also has
problems on MS Windows, then this is unrelated.)

Again, I do not know whether there is any relation. I just point this
out "just in case".

I reported the problems with `window-at' and `coordinates-in-window-p'
in connection with a bug in windmove.  See the thread "windmove and
the minibuffer" from last May and June
(http://mail.gnu.org/archive/html/emacs-devel/).  Richard fixed the bug
in windmove by replacing `window-at' with a new function
`window-inside-edges' that works on GNU/Linux.  I do not know whether
this function is suitable for ruler mode, assuming the two problems
are related.

The behavior of `window-at' and `coordinates-in-window-p', described
in the thread "windmove and the minibuffer", remains unchanged.  (I
just reproduced the exact same results as described in that thread
with this morning's CVS.)  I do not know whether Richard somehow
eventually concluded that this behavior was _not_ a bug, but then why
do these functions yield different answers for GNU/Linux and MS
Windows in exactly the same situation?

Sincerely,

Luc.