Re: Initial window placement
- From: Sasha_Vasko osca state mo us
- To: Havoc Pennington <hp redhat com>
- Cc: <wm-spec-list gnome org>, Philipp Lohmann <Philipp Lohmann Sun COM>
- Subject: Re: Initial window placement
- Date: Thu, 16 Aug 2001 14:09:31 -0500
> Sasha_Vasko osca state mo us writes:
> > Good example is XMMS trying to restore its location.
> >
> > Not that I'm proponent of letting clients doing that sort of things.
> > You are right - that should solely be handled by window manager
> > session management policy, but so many of those clients attempt to
> > implement it on their own, causing wierd things of window's slowly
> > edging out of the screen. WE could at least provide them with
> > some hints, so they avoid blatant errors.
> >
>
> My WM simply ignores configure requests for windows with the NORMAL
> semantic type. ;-)
>
> I'm considering adding a dialog:
>
> if (strcmp (window_title, "XMMS") == 0)
> show_dialog ("XMMS detected. Would you like to XKillClient()?");
Very tempting is it not :) XV falls into the same category , as well as
Netscape. Maybe we should come up with the Hall of Shame for clients like
that, or even Black List ?
>
> Anyhow, I think if we provide a hint here it should be a way to
> request that the WM session manage the window, not a way for clients
> to do their own SM. For cross-session position saving, then the SM
> spec already allows it, and we don't need to add anything more. For
> saving across e.g. two launchings of XMMS in the same session, we
> could do something like the hint I mentioned earlier, we give XMMS a
> "save geometry ID" that it can restore.
I completely agree. save_ID is great idea and nicely solves all this
problems with attempting to match names of the multiple windows with the
same class/name/title
>
> Havoc
Cheers
Sasha Vasko
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]