Transparent panel thoughts

[Brian J. Tarricone]

On 11/17/04 12:26, Randy Chung wrote:
> Brian J. Tarricone wrote:
> >On 11/17/04 02:35, Randy Chung wrote:
> >
> >>My understanding of the problem is that it's only problematic when the 
> >>panel is high enough in the desktop such that it'll block the window bar 
> >>when a window is maximized.  If this is the case, what about doing some 
> >>checking when maximizing to see if the panel is above a certain 
> >>threshold, and only then keeping the window from maximizing under the 
> >>panel?  Is this feasible? (I haven't had a chance to look at the 
> >>interaction between the wm and the panel lately)
> >
> >
> >the WM doesn't treat the panel in any "special" way, it just sees it as
> >another dock-type window.  i *think* what you want here is essentially
> >special-casing the panel in some way, which is a definite no-no.  i could
> >be wrong though.
> >
> What I had in mind was less special casing the panel specifically, and 
> more along the lines of special casing any window that's always on top. 
>  I'm not sure how well that would work, especially since I can imagine 
> a few scenarios where that would be undesired behavior.  What I suppose 
> it comes down to is whether it would make more people happy to have 
> always-on-top windows special cased or if the status quo makes more 
> people happy.

nope, the issue isn't always-on-top windows, it's dock windows.
always-on-top is fine; you can always max other windows under a-o-t windows.

i can't really give a concrete argument against this, but deciding
maximisation behavior around the positioning of dock-type windows (if any)
on the screen just sounds like a hack and a bad idea.

> Alternatively, if configure option bloat was acceptable, it could be 
> added as another option you could set (but this sort of thing seems to 
> be getting shot down pretty quickly lately)

i really hate the idea of a configure option for this, but of course it's
ultimately up to olivier.

	-brian