Proposal and RFC: DAL, the Desktop Abstraction Layer

[Ikke]

> I would wait for the updated documentation/tutorial from Havoc instead 
> of diving into the code (IMHO you should only ever need to look at the 
> code if you intend to hack on it and thats why we have documentation in 
> the first place).
I read most of the glib and low-level testcases from CVS by now, and
start to understand *some* of the stuff. Will need to read some source
comments too, still.
Indeed, let's hope the DBUS guys are able to update docs soon :-)

> Anyhow the bulk of the work will be in specifying the various 
> freedesktop services so if you have time on your hands by all means 
> start there.
I defined a "MusicPlayer" interface (a very basic, simple one), started
writing some helper functions, and will write 2 (fake) implementations
(who just output stuff to stdout, don't play actual music) and one
"launcher" (which could in a real world case be Nautilus, or something
else).

> It would be nice if those specs were formal based (EG xml format)
That was what I intended to do.
> as they might come in handy if we ever have a GOB equivalent for Dbus (or 
> indeed if GOB were to support it for GObjects). Being able to select a 
> spec and get the boiler plate code autogenerated for it would be a real 
> boon!
I did some work on autogenerating code before, this was in PHP though
(generate php class files from LDAP ObjectClasses by querying the LDAP
server). Don't know too much about generating C code, only had a quick
look on GOB a week ago or so.
One question here: if GOB and GObjects are used, what are the
consequences regarding Qt/KDE? The whole thing should stay
desktop/toolkit neutral...

> jamie

Thanks, Ikke