Re: _NET_ACTIVE_WINDOW, revisited

[Matthias Clasen <mclasen redhat com>]

On Thu, 2005-07-21 at 10:42 -0600, Elijah Newren wrote:
> On 7/21/05, Billy Biggs <vektor dumbterm net> wrote:
> >   Isn't the only application-class where you want this
> > jump-to-current-desktop behaviour 'gaim' or other apps which have a task
> > bar item, you click on it, and the gaim window appears on your current
> > desktop?
> 
> Personally, no.
> 
> >   This seems like the special case to me.  You never want KMail or
> > Mozilla to jump desktops unless you do something really crazy, and
> > lots of apps have multiple windows they want to move focus between.
> 
> As far as I can tell, you've introduced a third possible policy that
> could be implemented; I don't think it matches what either Lubos or I
> had in mind.  I think that helps my case, though.  I'm not trying to
> argue what the policy of the WM should be, I'm just arguing that there
> are more valid policies and that the policy shouldn't be built into
> the spec because that results in apps/pagers/WMs working around it
> when they don't like it.  I hate having a desktop where all apps
> behave totally differently.  This kind of thing was the whole reason
> _NET_WM_WINDOW_TYPE was introduced and the motif hints were
> deprecated.
> 
> 
> Incidentally, I just realized that our window selector is buggy.  We'd
> like an activation-from-pager-that-is-workspace-aware kind of hint,
> but there isn't any and we currently get the behavior _we_ want by
> doing _net_current_desktop + _net_active_window.  That's wrong because
> it's an attempt to enforce policy when the WM should decide it.  In
> particular, the WM should be allowed to treat such activation requests
> by moving the window to the workspace where the user is.  So, I need
> to pull that code out of gnome-panel/libwnck and instead create a
> message type (_GNOME_ACTIVATION?) and use it where available
> (otherwise using the _net_active_window as it currently stands) until
> an activation-from-pager-that-is-workspace-aware hint is added to the
> EWMH.

Yes, I always had the feeling that it might be good to have a more
intelligent policy for move-activated-window vs
change-desktop-to-activated-window, which could depend on the type of
activated window and on the activator. That also explains why the
introductory section on activation is so vague about the actions of the
wm (I wrote it intentionally vague)

Matthias