[Openicc] meta data in test chart

[Richard Hughes]

On 26 January 2011 09:38, Alastair M. Robinson
<blackfive at fakenhamweb.co.uk> wrote:
> Yes, that's what I'm trying to avoid - the need for a handoff space. The
> sRGB / AdobeRGB / Device space switch I suggested is a default input space,
> not a handoff space.

I also think a handoff space is a really bad idea. That said, deciding
on a fallback space (e.g. adobeRGB,sRGB,etc) is also fraught with
problems. Would the fall back space be:

* session preference?
* per-queue configurable?
* system policy?

Surely we just have to enforce that the applications set and send the
document colorspace when the page is printed. GIMP already knows my
document colorspace and my display colorspace for each monitor, and
corrects the image automatically.

It makes complete sense for GIMP to ask the session automatically what
the default colorspace for new documents should be. This is what the
session DBus interface in GCM allows. This colorspace just has to be
sent to the CUPS device so we can do just the single device
conversion.

Richard.