Review Request: include KolorManager in kdegraphics

[Boudewijn Rempt]

On Wednesday 14 March 2012 Mar, Daniel Nicoletti wrote:
> > No. There should be color management by default in KDE, that's really 
> 
> > important; and there should be only one solution by default. We shouldn't 
> > let distributions, or even worse, users decide which solution they use. That way 
> > madness lies. KDE's Color management solution shouldn't be in extragear.
> > 
> > As to which one is selected, there are a couple of ways to decide.
> > 
> > The first is, first come, first go. Kolormanager has been in development for 
> > quite some time now, and colord is an upstart, gnome-derived technology. 
> > Integration of colord in kde was only started very recently. Everyone is free to 
> > start a competing project, even inside KDE, but to make that project block a 
> > pre-existing project isn't the way to go.

> I'm not talking about blocking pre-existing projects, I'm really looking forward a
> solution where the two could live.

Well, I am really not looking forward to that situation. Sometimes having a choice is a bad thing. Sometimes starting a new project instead of helping out an existing project is not the right thing to do.

> Sure we don't want users/distributions to
> decide but they do.
> 
> > The second way to decide would be on technical merits. I'm not going to go 
> > into that discussion; I've seen too many tiresome discussions already, and I 
> > don't feel really competent anyway.
> > 
> > For me as an application developer, life sucks anyway, since I have to support 
> > Linux, Windows and OSX, so for the time being, the application will offer its 
> > own way to select profiles, in addition to using the X11 display atom that both 
> > colord and kolormanager support. (And I don't want to think about printing 
> > anyway.)

> Well if we go into the merits discussion I really think we will get nowhere,
> as we didn't sort this first we won't sort this out now, KDE and GNOME primary
> goals is Linux, so I really don't think supporting platforms where they already
> have good solutions for this is a way to go.

No, Gnome's primary goal is Gnome, while KDE offers a framework for building applications on many platforms next as well as desktop environment. My own desktop environment is KDE's plasma, but my goal for Krita is to have it run everywhere, not just on the plasma desktop. 

In fact, until Gnome 3 and Unity drove them away, most of my users were on Gnome... And that was fine, because colord and oyranos both set the relevant X11 atom, so Krita is fine, as long as I explicitly don't care about scanners and printing.

> So how do we go into the merit discussion without creating yet another flame war?

I don't know. I don't know of a extensive comparison between colord and oyranos. And I'm not sure it's possible anymore to find anyone who can do that competently, honestly and impartially. 

Realistically speaking, we'll have colord on Linux as the "standard" whether we want it or not, because it's in Fedora, and whatever Redhat puts in Fedora will be the "standard" for Linux. Of course other distributions will package it, because they will want to package gnome. Even if we had been faster to the finish line and had had kolor-manager ready for KDE 4.6 or 4.7, no way colord would not have replaced our own work.

Apart from all technical merits.

-- 
Boudewijn Rempt
http://www.valdyas.org, http://www.krita.org, http://www.boudewijnrempt.nl