On Tue, Jan 17, 2006 at 11:25:31AM +0000, Daniel Stone wrote:
> On Tue, Jan 17, 2006 at 09:22:06AM -0200, Otavio Salvador wrote:
> > Sven Luther <[EMAIL PROTECTED]> writes:
> > >> Um, neither does xserver-xorg.{config,postinst,templates}. So there's
> > >> not much point until those get fixed. I've been meaning to do it, but
> > >> never got around to it. (Don't count on me doing it, seriously.)
> > >
> > > I may do it on the purely xresprobe level, as time permits. not sure if i
> > > want
> > > to handle the x scripts though.
> >
> > Looks like xresprove should be the harder part. After have the need
> > information available, it need to be used. I might be wrong.
>
> I don't expect xresprobe to be hard. There are two parts which will be
> hard:
> a) making the xserver-xorg scripts understand multiple input and
> output devices, and
> b) getting DDC on multiple devices.
>
> Right now we generally do DDC via VBE, which is kind of difficult when
Ah, i was mostly interested in the PPC case, where you get the EDID from
either the OF device-tree or from radeonfb /sys node.
> you have multiple cards. Doable, but really really nasty. So what you
> really want is the drivers to be split out into Xorg drivers, but also
> libraries where you can initialise the card, look at its outputs, and
> figure things out for yourself. But that's, uhm, a bit of a pipe dream.
Well, so let's do it for non-braindead hardware :)
> Also, one of the more difficult things to do properly is multiple
> outputs on single cards, especially when it comes to laptops with
> external displays.
Mmm, i believe again that both heads will provide the the above information on
powerpc.
> It's not at all trivial, but fixing xresprobe is definitely the easiest
> part. Working out what you'll do about DDC, and fixing xserver-xorg;
> now that's the hard part.
I think the most difficult part will be to come up with a nice user interface
for xresprobe and co to handle this.
Friendly,
Sven Luther
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
