[argyllcms] Re: XRandR 1.2 is faulty...

  • From: "David H. Vree" <david.h.vree@xxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Sun, 28 Dec 2008 00:40:02 -0500

Hal V. Engel wrote:
On Saturday 27 December 2008 00:34:21 Graeme Gill wrote:
David H. Vree wrote:
I am running Linux (Ubuntu 8.10) and get the warning "XRandR 1.2 is
faulty - falling back to older extensions" with almost every Argyll
command.

Can anyone tell me specifically what might be wrong with xrandr so i can
file a bug report upstream?
As I understand it, the "bug" is with the video drivers not fully
supporting XRandR 1.2, and with the X11 server not properly communicating
this through to the client software. Encouraging those responsible for the
various video drivers to fully support XRandR 1.2 (Good luck - people like
NVidia have been dragging their feet).

NVidia promised to do this over two years ago and have yet to deliver a driver that supports XRadR 1.2. Phoronix reported last week that this is now a top priority for nVidia and nVidia X11 driver developers have posted to the nV News forums that it is now a priority. But they have not given an estimated date so we may see this sometime in the next few months but maybe not. AMD's proprietary video drivers have supported this for a few months now so newer versions of their driver should work. So if you are using an older AMD/ATI driver you may want to upgrade (also the newer versions of the driver are reported to be of MUCH higher quality).
Most of the open source drivers have had XRandR 1.2 support for some time now.

Just in case they fixed it, I upgraded to the latest nvidia driver -- version 800. The problem still exists.

I'll add my lone voice to those calling for this to be addressed. I am guessing that when nVidia fixes this it will also fix the "Screen Resolution" applet in Ubuntu that has been busted forever for me.

Is it safe to assume this is not a big problem for Argyll however?
In the documentation for dispcal I read that the environment variable
ARGYLL_IGNORE_XRANDR1_2 can be set to true in order to automatically use
another method.  Should just do this and ignore this problem?
I think this will suppress the warning message.

Graeme Gill.


Other related posts: