[Openicc] xrandr output properties / icc profiles

[Frederic Crozat]

On Thu, Jul 24, 2008 at 5:08 PM, Kai-Uwe Behrmann <ku.b at gmx.de> wrote:
> Am 23.07.08, 16:55 +0200 schrieb Frederic Crozat:
>> On Thu, Jul 24, 2008 at 4:47 PM, Kai-Uwe Behrmann <ku.b at gmx.de> wrote:
>> > Am 23.07.08, 16:26 +0200 schrieb Frederic Crozat:
>> >> On Thu, Jul 24, 2008 at 4:08 PM, Kai-Uwe Behrmann <ku.b at gmx.de> wrote:
>> >> > Am 18.07.08, 07:06 +0200 schrieb Frederic Crozat:
>> >> >> On Wed, Jul 16, 2008 at 4:28 PM, Graeme Gill <graeme at argyllcms.com> wrote:
>> >> >> > Tomas Carnecky wrote:
>> >> >> >> should != does
>> >> >> >>
>> >> >> >> See the server-side implementation of XRRQueryVersion():
>> >> >> >> http://gitweb.freedesktop.org/?p=xorg/xserver.git;a=blob;h=5525427f667dcfb0e5b8b3cb8a2a99c4623c6df8;hb=55644d2f1cddf1777587f213ed581b6087d9fdad;f=randr/rrdispatch.c#l37
>> >> >> >
>> >> >> > It could obviously be much better implemented then, and without a great
>> >> >> > deal of trouble - all it needs is a "return XRandR capabilities" API
>> >> >> > in the driver.
>> >> >>
>> >> >> The way to detect if driver is really supporting RandR 1.2 or with
>> >> >> compat-RandR 1.2 is used is to check if output name is "default" and
>> >> >> there is no CRTC attached. In that case, compat-RandR 1.2 is being
>> >> >> "generated" by Xserver.
>> >> >
>> >> > I can not see any guarantee in this behaviour.
>> >> > http://gitweb.freedesktop.org/?p=xorg/proto/randrproto.git;f=randrproto.txt;a=blob
>> >> > does not even mention "default".
>> >>
>> >> Check Xorg code ;)
>> >>
>> >> (thanks Matthias Classen for the hint)
>> >
>> > I checked here and my XRRScreenResources has one output and one crtc.
>> > The output has one crtc attached.
>> > 2 monitors over twinview + Xinerama
>> >
>> > Does you talk about XRandR 1.3?
>>
>> No, I'm talking about XRandR 1.2 : if a driver doesn't not support
>> 1.2, output name will be "default" (and not CRTC will be detected).
>
> Then there must be a bug in my server (xorg 7.2). I see the output
> "default" and one crtc is attached to the output. So the above
> described behaviour is not fullfilled here with a non supporting driver
> (nvidia).

Hmm, my bad, CRTC might be filled. Checking on "default" should be
enough (that is how I fixed gtk+).

PS : please, don't cc me
-- 
Frederic Crozat