Fwd: My KImageEffect code

Zack Rusin zack at kde.org
Fri Nov 26 22:39:38 GMT 2004

On Friday 26 November 2004 12:20, Maksim Orlovich wrote:
> > The main issue is maintainership. From what I gathered, Maksim is
> > the only one works on that stuff every now and then (I apologize in
> > advance in case I overlooked anybody :-), so the last call is
> > probably his. I for one am not
> Actually, Zack is your man for most of this stuff.

I do feel guitly about not being more proactive with this... 

And yes as Mosfet was saying current state of affairs in KImageEffect 
can be desribed with such popular phrases as "huh?" or "oh, hell no". 
And I think I talked to some of you about this a few times. On my todo 
I had a number of things which just have to be done there, but I'm not 
known for too much organization and with spare time that can be counted 
in seconds on a weekly basis I immerse myself in chaos.

So for KDE 4:

- we need to get rid of most of the algorithms there. I'm guilty of the 
worst abomination of them all since I added the bump-mapping algorithm 
(I wanted tot play with bump-mapping some style elements, like labels 
or menus, but of course didn't have time for that). Personally I'd like 
to leave in kdefx only blending and gradients.

- I'd like to create a more sophisticated image library where all the 
other ops are moved. So that graphics apps can share it.

- add tests for all ops in both libraries.

The problem that we're facing right now stems from the fact that most 
ops in KImageEffect have simply never been used. As someone who worked 
with them all, trust me when I say that a suprising number of them 
simply doesn't work. So the lack of reports is directly caused by the 
fact that all KDE applications that I've seen essentially use only 
blending, gradient and maybe blur ops.

So, going back to the original query: Mosfet if you have some time and 
would like to work on this, I'd be _very_ happy. We should think about 
which algos to leave in KImageEffect and which ones to move as well. 


Those who live by the sword get shot by those who don't. 

More information about the kde-core-devel mailing list