What to do about KColor?

[Ingo Klöcker]

On Monday 28 May 2007 21:34, Thomas Zander wrote:
> On Monday 28 May 2007 20:40:58 Alex Merry wrote:
> > I agree with this - the function name is very confusing, because it
> > _doesn't_ blend two colours.  It overlays the second on the first.
>
> That entirely depends on your definition of what a color is.
> Beginning in Qt4, and in other toolkits quite a bit longer, the
> definition of a color is *including* the alpha channel.  And with
> that in mind, the blend method indeed does exactly what it says.

Great. Then I guess this method fulfils all the qualities of a statement 
made by a mathematician: It's 100% correct [from a color theoretical 
point of view], but completely useless in practice.

> As I (and Aaron) stated elsewhere;  lets see where actual apps end up 
> using this and avoid this 'design by committee' re-painting of the 
> bikeshed.

By committee? Yeah, right. Only that the committee is apparently you, 
Aaron and Zack. I find it extremely irritating how this committee is 
stubbornly ignoring the comments of people like Andreas and Alex who 
actually want make use of a color blending method in their code. 
Pigheadedness is not the way to create a good API. A good API is 
designed by listening to the people who want to use the API. That's API 
Design 101. And if they say the current API is unintuitive and not easy 
to use then it's not because they are too stupid to use it. Bah. This 
makes me sick.

I'm not able to tell you just how disappointed I am about how things are 
apparently decided nowadays in kdelibs.

Regards,
Ingo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/kde-core-devel/attachments/20070529/659c27e2/attachment.sig>