Re: Application launch detection

[Havoc Pennington <hp redhat com>]

Peter Astrand <astrand lysator liu se> writes: 
> The problem with extending Xt, GTK or QT is that there will always be
> thousand and thousands of applications and toolkits without this
> extension. Eg, this feedback mechanism will only work with perhaps 90% of
> the applications. This means that the desktop environment much decide
> if to use feedback or not - not an easy task. Including some flag in the
> .desktop file is probably the only way to do it. Not a clean solution. 
> 

I think two things are clear:

 a) this won't get added to Xlib (for the reasons Owen mentioned) 
 b) LD_PRELOAD sucks ;-)

Therefore we need to plan on requiring toolkit support. We can do an
LD_PRELOAD hack on some platforms for some apps, but this sure
shouldn't be the default case or the long-term solution.

Peter, could you write down a more formal description of how the
XSendEvent() works, so we can get a spec that explains that mechanism?
We need to document exactly what toolkits should do (what properties
to set, events to listen for, name of the env variable to use).  Then
GTK and Qt can implement it, and maybe we can even get Xt/Motif to do
so.

Also, are you interested in working with George on this issue for
GNOME 1.4 - I guess modifying Xalf so the panel and Nautilus can use
it when launching apps. We'd need to support a GTK 1.2.9 that natively
implements the protocol we decide on, and the panel would provide Xalf
with a flag indicating whether the app natively supports the protocol
or not.

Havoc