DM communication standard

[Oswald Buddenhagen]

On Sun, Aug 08, 2004 at 11:50:10PM -0400, Havoc Pennington wrote:
> On Sun, 2004-08-08 at 16:13, Oswald Buddenhagen wrote:
> > 
> > as for d-bus ... yeah, i wanted to do this originally as well. but havoc
> > didn't seem to be overly-excited about the idea and suggested staying
> > with a separate control protocol - dunno if he still stands to this
> > view.
> 
> Hmm, I don't remember the conversation
>
at kastle. :)

> or what I meant at that time. ;-)
>
dunno, either. ;)

> I don't see offhand a reason dbus wouldn't work - you would use the
> _systemwide_ bus daemon for this to export a DisplayManager service.
> 
yup. sure it would work, i never expected anything else. :)

> > and the second reason being, that kde did not arrive at the d-bus
> > revolution, yet ...
> 
> Right, neither has GNOME, but in this case we're talking about the
> systemwide daemon (what HAL etc. are using) rather than the DCOP
> replacement (a separate issue, the per-session daemon)
> 
i know, i know. i thought in the direction: when we use dbus everywhere
else, it would be logical to use it here as well. but as long as it
would be more or less an alien, we can use something else just as well.

> Not that I'd encourage you to rewrite a ton of code to use D-BUS if
> it's already working.
>
it's not a ton; the entire file is 18 kb right now. a lot of it is
creation and handling of the sockets. another big part is parsing the
commands and constructing replies, i.e. stuff related to the protocol
being pure ascii; otoh, the kdm backend is plain c without any toolkit,
so i suppose i couldn't use an idl compiler anyway.

> Especially since D-BUS remains somewhat API unstable.
> 
that's certainly a killer, as i need the stuff right now. kde 4 maybe
... :)

-- 
Hi! I'm a .signature virus! Copy me into your ~/.signature, please!
--
Chaos, panic, and disorder - my work here is done.