Pixel operations and colorspaces

[Boudewijn Rempt]

On Tue, 16 Aug 2005, Casper Boemann wrote:

> I'm just not convinced that we need this in the first place, but you obviously
> do. I'm not going to argue over that, but I feel that our problems are more
> growing pains than an actual problem.

I wish I did think that it wouldn't be necessary either, but as soon as I start
thinking of ways to warn the user that the filter relies on operations that are
not implemented by the current colorspace my thoughts veer in this direction.

It _should_ be pretty lightweight... Even though it's rather a lot of work that
I wish I could postpone, but I'm getting hopelessly bogged down by the current,
big, colorspace classes.

Boudewijn