GSoC CM collaboration
Hal V. Engel
hvengel at astound.net
Wed Mar 5 15:22:25 PST 2008
On Wednesday 05 March 2008 14:57:41 Harald Braumann wrote:
> On Wed, 05 Mar 2008 12:55:56 +1100
> Graeme Gill <graeme2 at argyllcms.com> wrote:
> > But there is no need for late binding at the low system level. Doing
> > the binding (== device profile linking) at the high system/graphics
> > API level is fine.
> I'm not intimately familiar with the technical details, but from a user
> perspective I would expect that I create a colour profile for a monitor
> and then everything is colour corrected on that monitor. I guess I
> wouldn't get that result if I had to depend on every single GUI toolkit
> to implement colour matching. If a window is displayed on multiple
> outputs, either in full or partially, I would also expect it to be
> colour corrected on each output. Now I'm not sure if that is at all
> possible with the current architecture, as the window would have to be
> rendered independently for each output, but it sure would be nice.
> Anyway, you couldn't achieve this at the application level.
You could achieve it at the application level but it would be difficult and
would add significant complexity to how applications deal with rendering
images. This would entail using different transforms for different parts of
the image which would require adding addition logic to track what parts of
the image appear on what actual display device. At this point applications
like GIMP, CinePaint and the like make an assumption that the user will only
display images on a single physical display.
More information about the xorg