[Openicc] Promoting colormanagement for LINUX

Chris Murphy chris at colorremedies.com
Wed Feb 13 16:16:41 PST 2008


On Feb 13, 2008, at 5:54 PM, Graeme Gill wrote:
> Hmm. I think the X11 folks have decided that CM is exactly NOT what  
> X11
> should be doing, at least going by the "State of X" google talk.
> See <http://www.youtube.com/watch?v=-oFxhqYn-g0> at the 17:00 mark.
>
> And to a fair degree I think they are right. There might well
> be architectural reasons for placing certain parts of the
> functionality in the X server, but the primary issues with
> doing this at the moment is the standards one. Color isn't a
> done technology that everyone agrees on. Build something like
> a current CMM into the X server, and it will only be a short
> while before people are fretting about how it then locks them
> into an "old fashioned" work flow for old hardware and old
> applications, and is an "extra layer of complexity" that has
> to be bypassed.


In less than 10 years anyone using a digital camera and a computer  
will expect any application that can display images to be able to  
work as easily with scene-referred HDR images as easily as  
applications across the board have been able to work with JPEG and  
TIFF. It's an issue for photography, for video, and for games. It's  
not like it's pie in the sky stuff. People will want this and the  
technology will be there - short of a meteor hitting the planet and  
setting us back 500 years...

I don't know how the x11 people actually think this is going to get  
done, except by total piecemeal by various developers. That is not a  
good leveraging of finite resources in my view.

I agree that consideration needs to be made to keep any  
implementation modular because it may have to be supplemented or flat  
out replaced, and perhaps in the not too distant future.

While this is ostensibly an openicc group in reality is a color and  
imaging group more than anything else. There's nothing that says the  
open source community can't integrate other color management  
technologies along side full ICC support as those technologies become  
available and can be implemented into the architecture, even if the  
ICC chooses not to or just goes too slowly for the open source  
community. I have no problem with that. Use the best we've got,  
whatever that is.


Chris Murphy


More information about the openicc mailing list