Unittests and colorspaces

[Cyrille Berger]

On Tuesday 10 July 2007, Boudewijn Rempt wrote:
> Then your test will run for all available colorspaces. This is pretty
> useful since it turns out that not all colorspaces are that stable. Gray8
> and Gray16 fail on the bltMask test, for instance.

What is surprising me is that they fail only on that one :/ In fact composite 
op assumes the presence of an alpha channel. Unless there is an emergency, I 
will look at fixing that one only once I have started unittesting pigment 
(which is arround #2 or #3 pigment priority)

-- 
Cyrille Berger