On Thu, 2011-03-17 at 00:23 +0100, Jacek Poplawski wrote:
> I know nothing about politics and fights between Gimp developers, I
> just want to understand the development process (I use Gimp almost
> everyday, so it's very important to me).
It's not a question of fights and politics.
It's a very small number of people with a lot of other things to do in
life, trying hard to work together, and making a really good product.
> I am also willing to donate
> my free time to this project (as a developer), but reading this forum
> (complexity of development process, not fights) holds me back.
GIMP is a fairly large piece of software; it's also true that there seem
to be some bottlenecks in development right now. As I see it these are:
(1) all UI proposals (and actual changes) need to be reviewed by an
experienced UI designer; this does happen, and all the developers seem
to agree that it improves the user experience considerably, once the
negotiations are complete, but it can take a lot of iterations;
(2) the move to gegl means lots of patches and contributions get turned
down, because they will help gimp in the short term, but will then soon
need to be rewritten;
(3) there are between 0.5 and 3 people-equivalent working on GIMP coding
in any given week. That is, GIMP generally gets only a few hours of
progress each week right now.
(4) it's true there's not 100% agreement on priorities, although there's
for sure broad agreement. Some high priorities are clearly fixing the
partly-implemented features (e.g. single window mode, on-canvas curves,
layer groups), and the move to gtk+ 3.0. This work is mostly in the
"very hard" category, so it needs core developers.
Changing this means getting past a difficult hill in the road, so that
then smaller changes (even if equally important to some or many users)
become possible again.
Maybe one day a month of "gimp love" easy bug fixes would be worth the
time cost of having the core developers spending the day mentoring.
If it were possible to hire Sven, Mitch, the Michaels, Alexia, Simon,
Wilber and the others for a solid month, we could probably drive over
that first hill, and maybe even get within sight of the Mountains Of
Gegl. For the Mountains we'll probably need a lot of
not-necessarily-core developers to move plugins, tools and scripts.
Saying, "but this bug is absolutely critical" is really only OK if (1)
you've already got a patch that's well-formatted :-), and (2) if it's
really true that absolutely no-one can use GIMP without the fix, e.g. it
crashes on startup. That may be true for GTK+ 3, for example.
This is my own perspective, and I am not by any means a core gimp
developer, even if I might like being in a cage sometimes :D
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Gimp-developer mailing list