Mathias Lindner ( wrote:
> Can I add something?

Sure.  :)

> All fine to now, two controversial opinions that may be solved by some 
> trade-off... But I still have to give some critique: I have experienced 
> similar incidents quite a few times. Again, you may all be right and I 
> understand your perspective. But if you want to get/keep contributors 
> you want to consider *how* to stress your workflow. Please try some 
> empathy: somebody invests a lot of time (and even more important does 
> the step to actually start doing something) and is rejected with not too 
> much thankfulness. Or at least it sounds like that. I know about some 
> people that wanted to contribute but also gave up due to bad help and 
> harsh conversation.

I understand your concerns and I do share them - at least partially. So
with all clearness:

I do appreciate the effort from しげっち (I hope that comes out
correctly...  :)  and I love the fact that he actually dived into the
code and made this work. We get a lot of "why don't you? It is easy!"
requests, and having some existing code is nice for a change.

Unfortunately we have a tendency to be a bit blunt in our wording and
rejecting stuff - even for hopefully good reasons - is hard. It does
have the advantage of being very clear, but yeah - it can put people
off, which is something I hate.

Especially with regards to the UI we currently are in a tricky
situation: On one hand the collaboration with Peter is very fruitful
and we got some enthusiastic responses to the stuff that has been
developed with his help. Since it is clear that a consistent UI does not
happen with people developing on random bits in the UI we - in my
opinion - need a person that coordinates the efforts in the UI, and that
person is Peter. On the other hand, yes, this unfortunately creates a
bottleneck (which scares me), but I don't see a good way around this.

So I really recommend to discuss stuff (on this list, in irc etc.)
before investing lots of time into code, that might get rejected for the
reasons mentioned above. This not only applies to GUI stuff, but to
other functionality as well.

In my opinion the current Gimp source code is in a very good state and I
tend to attribute this to strong maintainership and a reluctance to
accept random patches. Yes, this is sometimes hard on new contributors
(which we urgently need), but I don't think just opening the floodgates
would be a good strategy here.

What currently is needed is some more work on the single-window mode.
AFAIK there are some bugs in there and Martin (Enselic) can no longer
dedicate as much of his time to it as he used to. It would be awesome if
this work could be picked up and continued, I think a lot of the GUI
issues have been worked out already and there is a spec for it at .


Gimp-developer mailing list

Reply via email to