Re: [RP] latest cvs xemacs problems

[shawn]

Ben Leslie <address@hidden> writes:

> On Sun, 20 Jan 2002, shawn wrote:
> 
> > Ben Leslie <address@hidden> writes:
> > 
> > [snip]
> > 
> > > > What versions are we talking about here?
> > > 
> > > Mmm, an cvs version from a couple of months ago, to the current cvs 
> > > version :).
> > > Sorry thats not much help I guess :).
> > >  
> > > > > I will try and track down the error, but if anyone has seen this 
> > > > > before
> > > > > or might know why this is happening that would save me a lot
> > > > > trouble.
> > > > 
> > > > You'll want to play with the maximize_normal function in manage.c.
> > > 
> > > 
> > > Ok cool. That sounds like a good starting point, I'll see what I can do.
> > 
> > If the cvs version you had was a couple months old then it could be a
> > change I made back in august...Unfortunately, my Changelog entry is
> > almost totally useless:
> > 
> > 2001-08-23  Shawn  <address@hidden>
> >     * src/manage.c (maximize_normal): In the new height on windows
> >     with increment hints code, reorder the steps.
> > 
> > At some point windows with icrement hints weren't ALWAYS being
> > maximized fully. They were always maximized 1 increment unit too
> > small. When I run xemacs and remove the toolbar it looks like xemacs
> > wants the window size to be 1 increment smaller than what rp wants
> > it. xemacs could have been functioning fine because of the resize bug
> > in rp...
> > 
> > Could you tell me the file revision of manage.c and events.c that work
> > for xemacs?
> > 
> > Really this is xemacs' problem and not rp's problem. xemacs is icccm
> > incompliant. But, xemacs is everywhere and ratpoison is not...
> > 
> 
> 
> Ok found the problem. Don't know the cleanest fix, my solution has been
> to #if-ed out the line.
> 
> Around line 323 in events.c, (configure_request).
> 
>         if (win->state != NormalState)
>           {
>             /* The window isn't visible so grant it whatever it likes */
>             XConfigureWindow (dpy, win->w, e->value_mask & 
> (CWX|CWY|CWBorderWidth|CWWidth|CWHeight), 
>                               &changes);
>             send_configure (win);
>           }
>         else
>           {
> #if 0
>             /* Draw the hard line. Get back in line, you misbehaving
>                window! */
>             maximize (win);
>             send_configure (win);
> #endif
>           }
> 
> The #if-ed out lines are what I've added.

Are you running the latest out of CVS? those lines have already been
removed in the latest CVS...

> Hopefully this can help work out what the _real_ problem is :).

xemacs is the problem. Its not icccm compliant.