On 01/26/2011 08:15 AM, Stephen Greenwalt wrote:
I have been reviewing Gimp / GEGL source code . . . to get familiar with
everything . . . so that I have some context to understand where I might
help the project.
But I am operating in a void because I don't understand:
* How the development effort is organized.
o
Is there a team leader?
Strategic decisions are taken in concert, although due to human nature,
the opinions of some individuals have more weight depending on their
reputation. We have two official maintainers, cat MAINTAINERS.
o What tools, etc. are being used to manage things such as:
+
Feature requests, bug reports, etc.
We prefer feature requests on this mailing list, so that they can be
discussed and evaluated, in particular in the context of our product
vision [1]
Bug reports are managed in GNOME Bugzilla [2]
+ Component development.
In general, all development happens on the git master branch directly,
although personally I think we should review our development methodology
for GIMP 3.0 because we have problems in our 2.8 development cycles due
to this.
+ Testing.
I have a long term goal of improving the GIMP development environment,
and have lately put efforts into writing regression tests for both old
and new functionality, and I have been setting up a nightly builder that
runs all our tests [3]. Curiously enough, you'll notice make distcheck
failed tonight due to changes yesterday, but I already fixed that...
* What the development priorities are, and where programmer
resources are needed.
Right now, the priority is getting GIMP 2.8 out. The two major work
areas left is finishing the implementations of layer groups and
single-window mode, see [4] for the main missing bits.
After that, we will start working on GIMP 3.0, which will be GIMP 2.8
running on GTK+ 3.0 and using GEGL-buffers for all data. That is, we
will throw out our 8-bit only image/tile data structures. For GIMP 3.2
we will start looking into serious support for non-destructive editing.
The main Gimp website says that the new priority is expansion of the
GEGL engine.
But, does that mean that major feature development for main Gimp app is
being stopped until the GEGL engine is ready?
See above
I, for one, love Gimp but would really like to see one or new features
such as:
* Support for sub-layers (i.e. nesting of layers in tree-style control).
Fundamental support for that is already in place in git master
* Ability to delete multiple layers at the same time, rather than
one-by-one.
Yes that would be nice, but support for higher bit depths and
non-destructiveness has higher priority
* Layer-by-layer history . . . so that you can undo changes within
the current layer only.
Sounds like a work-around for GIMP not having non-destructive editing yet
Should I offer help in those areas?
Any help is greatly appreciated.
Best regards,
Martin
[1] http://gui.gimp.org/index.php/GIMP_UI_Redesign#product_vision
[2] http://bugzilla.gnome.org/browse.cgi?product=GIMP
[3] http://gimptest.flamingtext.com:8012/waterfall
[4]
https://bugzilla.gnome.org/buglist.cgi?product=GIMPbug_status=UNCONFIRMEDbug_status=NEWbug_status=ASSIGNEDbug_status=REOPENEDtarget_milestone=2.8
--
My GIMP Blog:
http://www.chromecode.com/
Nightly GIMP, GEGL, babl tarball builds
___
Gimp-developer mailing list
Gimp-developer@lists.XCF.Berkeley.EDU
https://lists.XCF.Berkeley.EDU/mailman/listinfo/gimp-developer
