On Mon, Jun 21, 2010 at 10:45:38PM +0100, Thomas Adam wrote:
> On Mon, Jun 21, 2010 at 09:25:14PM +0000, halfdog wrote:
> > Thomas Adam wrote:
> > > On Mon, Jun 21, 2010 at 07:54:46PM +0000, halfdog wrote:
> > >> I did not manage to make a movie, but produced 2 screenshots to
> > >> illustrate the
> > >> problem, 3rd one is montage of old behavior:
> > >>
> > >> 1) PrePage.png: Cursor approaches bottom edge of page 0, cursor y~100%
> > >> at bottom
> > >> of page
> > >> 2) PostPage.png: Paging has occured, fvwm displays page 1, cursor y~100%
> > >> at
> > >> bottom of page
> > >> 3) PostPageOld.png: Behavior with 2.5.26 fvwm: paged to page 1, cursor
> > >> y~0% at
> > >> top of page.
> > >
> > > I *think* I understand what you're saying now, but I can't reproduce it,
> > > can
> > > see nothing obvious as to why this would happen, especially since the code
> > > changes between those two versions doesn't infinge on any of the code that
> > > handles with panframes or paging.
> >
> > Which version did you test? What did your desktop look like after paging?
> > PostPage or PostPageOld?
>
> PostPage. Both times, and this was using 2.5.26, 2.5.27, 2.5.28, and 2.5.31
> (CVS).
>
> > Could it be that some X-server properties could interfere with the
> > calculations
> > and that they cause this different behavior?
>
> No, because the panframes are created around the edge of the screen, which
> FVWM reads from X11. You'd really notice a difference if that weren't the
> case, since you'd be paging in the middle of the screen.
>
> > Does fvwm read some other properties beside .fvwm/config?
>
> Yes, but nothing which would interfere or cause problems with the example
> config you supplied.
>
> > What could be of use? X-socket dumps and analysis? Compile fvwm with debug
> > options? Try to use 2.5.28 on old xserver or 2.5.26 on new server?
>
> Try a different X server, but this is not likely to be much problem.
>
> What you could do, if you're really sure that the problem exists between
> FVWM 2.5.26 and 2.5.28 (and thereafter) is to to use git-bisect to find the
> commit which introduced this problem.
>
> This assumes you know git. See:
>
> http://github.com/ThomasAdam/fvwm
You could try reverting this commit:
292aeb4c39bba181267910f1e2eea3f4441f206c
-- Thomas Adam
--
"Deep in my heart I wish I was wrong. But deep in my heart I know I am
not." -- Morrissey ("Girl Least Likely To" -- off of Viva Hate.)
