Re: USPosition vs. PPosition with Virtual viewports
- From: Lubos Lunak <l lunak suse cz>
- To: wm-spec-list gnome org
- Subject: Re: USPosition vs. PPosition with Virtual viewports
- Date: Sat, 29 Oct 2005 13:32:37 +0200
Dne čt 27. října 2005 20:11 Sasha Vasko napsal(a):
> Lubos Lunak wrote:
> >>_NET_PLACEMENT, ATOM[]/32
> >>
> >>With possible atoms :
> >> _NET_PPOSITION
> >> _NET_USPOSITION
> >> _NET_USPOSITION_IS_DESKTOP_RELATIVE
> >
> > I don't understand what you actually mean with these hints. My idea was
> > basically a hint saying "you stupid app I'm a sane WM, don't try your
> > USPosition tricks on me that you need only for some BorkenWM-0.1". I.e.
> > if we now go to toolkit authors and ask them to remove all the UPosition
> > abuse (which I've already tried for Qt) they'll refuse with "but we need
> > that for CrapWM". With a hint they could check whether they need it. The
> > same way with a hint set by apps the WM could be less paranoid about
> > those apps.
>
> Hmm, maybe _NET_PPOSITION_RESPECTED, _NET_USPOSITION_RESPECTED and
> _NET_USPOSITION_IS_DESKTOP_RELATIVE then
>
> A property _NET_PLACEMENT will be set on root by a window manager
> indicating what placement flags it actually takes into consideration and
> how.
>
> If say such property contains _NET_PPOSITION_RESPECTED, then clients
> should know that they can use PPosition to achieve their goals.
> If _NET_USPOSITION_IS_DESKTOP_RELATIVE is set - and apps abuse
> USPosition by restoring some saved position from last run - they should
> at least adjust it by the current viewport.
> _NET_USPOSITION_RESPECTED is probably excessive, just stuck it in there
> for the sake of completeness.
Isn't all of this rather a workaround for a workaround then a real fix? I
think just a flag for WM I-do-it-properly and a flag for apps
I-do-it-properly is a simpler solution that would solve all these problems.
Instead of complicating the matters with _NET_USPOSITION_IS_DESKTOP_RELATIVE,
where the would have to be changed to use it anyway, those apps can be simply
fixed to do it properly with they detect a proper WM, then they set their
flag and the WM can also do it properly.
Which, if I'm getting your problem right, also means it would be only your
problem. The app would use USPosition only for -geometry, without caring
about viewports or whatever, only your WM would make necessary fixups if
needed everytime it sees USPosition set by an application with the proper
support flag set.
--
Lubos Lunak
KDE Developer
[
Date Prev][Date Next] [
Thread Prev][Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]