I like sketching. There may be better artists, but I still like doing it. I
started working on Krita instead of Gimp because Gimp didn't satisfy my
painting urges (in 2003) and because the code looked decidedly unhackable.
Krita was nearly ready back then, I thought, only missing an actual painting
tool. I didn't work on MyPaint, because that application started in 2004. My
comparison application are art rage (nice, but unusable because the paint
never dries) and Corel Painter (nice, but not for tablet usage because of the
I don't do much with the pictures I take with my camera, mostly cropping and
scaling for my blog. I seldom even run unsharp mask on them, let alone color
The kind of interface I need actually varies with computer: on my tablet pc I
want some art-rage-like, but usable, on a computer with a bigger screen,
something Corel Painter like will do, though even there, I can do without the
many of the palettes and would like an easer way to select and manage colors
and brushes (like what Vera is now working on for Krita :-).
The advantage I see for our technology is the flexibility: our colorspaces and
brush engines are pretty unique, and they make it worthwhile continuing with
Krita. But I have often wondered whether making a second painting-oriented app
on top of krita/image and pigment wouldn't be easier than trying to develop an
all-in-one package, even with modes or profiles.)
I see the following problems with krita, apart from anything that might have
to with vision, focus or usability:
* performance: krita just is too slow to work with. 1.6 was slow, but not too
slow, 2.x is way too slow. This prevents me from actually having fun painting,
more than a lack of features of our brush engines.
* stability: we have serious stability issues. Going out on a limb, I'd say
that most are related to integration issues with KOffice libraries, instead
of problems in our own core code.
* being part of a suite: not a problem for me, but I keep running into people
who tell me that they won't use Krita because it is part of KOffice. Silly of
them, but it means we are losing out on users and potential constributors who
are way more part of the graphics/art scene than most people in the KOffice
* fit and finish: we have lots of really cool features, but they all feel a
bit unfinished (masks), or very unfinished (brush presets, color mixer). Which
means not many people can really make use of them. When we move to git, I
would like to have a main tree and a staging tree where all the new features
land first only to be moved to main when they are really good enough.