Re: Proposal: _NET_WM_STATE_MINIMIZED

[Sasha Vasko <sasha aftercode net>]

Giles Atkinson wrote:

> Elijah,
>
> What is the semantic difference between setting _NET_WM_STATE_MINIMIZED and
> setting the ICCCM property WM_STATE to IconicState?  If there is none,
> this does not look like a good idea.

There is none.
In fact it may be informative to consult this thread in ML archives :

http://mail.gnome.org/archives/wm-spec-list/2001-December/msg00007.html

for a background on _NET_WM_STATE_HIDDED vs. _MINIMIZED

> Also, is there any semantic difference between "minimised" and "iconic"?
> The former looks like poor terminology for X as the window size is 
> unchanged.
> I believe MS Windows really does shrink windows as well as hiding them when
> they are "minimised".  Occasionally the resulting runt windows are visible.

In X, Iconic means that application should animate its icon window, 
instead of its main window. You cannot simply shrink main window in 
Iconic state. You actually have to unmap it ( including all its parents 
if any ).  ICCCM provides very detailed description as to how Iconic 
state should be tracked using Map/Unmap messages and WM_STATE property. 
The only reason _HIDDEN was introduced is to help distinguish between 
unmaps that happen when virtual desktop is switched, and iconification 
unmaps.

Now its concievable that someone will want to introduce yet another 
state where main window is simply resized to be smaller ( similar to 
shaded state ) In that case such window manager is free to introduce new 
_NET_WM_STATE atoms to identify it, but I don't think it should be made 
part of the spec. After all that's why we went with atoms, and not 
bitflags, so that window managers can introduce whatever wicked states 
they want, without disrupting everyone else.

> Thanks,
>
> Giles

Thanks
Sasha