I have been pretty brutal in chopping a bunch of enhancement requests today. What's left in the 1.3 milestone is a few bugs with patches outstanding and about 20 feature requests, most of whioch are claimed by someone or have patches outstanding.

Of course, if a bug report has a patch against it, someone needs to update it for CVS, and commit it, before the bug can be closed. That takes time, so it would be nice if people who had a working build of 1.3.17 could come up with an updated patch, and attach it to the bugzilla bug report.

This is a summary of outstanding features/bugs for the feature freeze. There are a number which require some action but not much. Most of these are what city people call low-hanging fruit.

ASSIGNED/Currently being worked on:

100356 Marc Lehmann

84151 Sven Neumann 85579 Sven Neumann
107009 Sven Neumann
102822 Sven Neumann

52374 Maurits Rijk

Patch attached, or linked to code, requires update and commit:

28889 66367
78180
85285
98326
113033
115626

Solution in bug comments, no patch:

62988

Features which are pretty serious, and should be done if possible:

55364 91941
113564

Bugs which should be fixed before pre-releases:

78064 102058

Requiring comment. Possibly finished?:

51108 56443
71514
106991
115427

Apologies, I had intended to have a sumarry of the bug with each bug number, but time has run out for the evening, and I prefer to send it like this than wait until tomorrow.

It goes without saying that anything outside the 6 bugs that have names attached to them, everything is up for grabs. In particular, if the bugs 55364, 91941 and 113564 could be looked at, and comments added, and in an ideal world patches attached, that would be great.

If not, updated patches for the bugs with patches against 1.2 attached or old 1.3 releases are very welcome, and for the other "easy" bugs, patches are also welcome.

Thanks a lot,

Cheers, Dave.

Alan Horkan wrote:

On Sat, 26 Jul 2003, David Neary wrote:

I have been pretty brutal in chopping a bunch of enhancement requests today. What's left in the 1.3 milestone is a few bugs with patches outstanding and about 20 feature requests, most of whioch are claimed by someone or have patches outstanding.

It would be a really big help if those in the know could please add the easy-fix keyword whenever they know a bug is easy to fix? A pointer the relevant file in the source code is always a huge help to those of us less familiar with the codebase (and that goes for any project).

Agreed. We do use the PATCH keyword for bugs with outstanding patches attached, which goes a lot of the way to what you are asking for. When there isn't a fix, easy-fix should be used for the situation where the fix is described in detail in the comments.

I dont know if "Count Colours" is anywhere in the list of bugs Dave mentioned but the "ColorCube Analisys" plugin effectively gives this functionality but unfortanately was only available for GIMP 1.2 on windows. This seems like it might be a relatively easy feature to get working but then again I dont really know, and it is probably one of the features that got bumped.

Sorry Alan, that's one of the ones I bumped. It could be brought back into the 1.3.x milestone, on the condition that (1) the Colorcube analysis plug-in is GPL, and (2) someone agrees to maintain it.

That brings up a major point which should be discussed. One of the big TODO points to come out of the last gimpcon was that plug-in distribution needed a complete reworking.

There are several unmaintained plug-ins in the main gimp distribution, and there are several well-maintained plug-ins distributed separately. It would be nice to provide just the core, and have a kind of on-demand plug-in installation process for other plug-ins (kind of like CPAM does for Perl). Of course, someone said this all 2 years ago and no-one took it under their wing.

Including plug-ins in the main gimp distribution requires a maintainer. If someone says they will maintain it, then it can go in. The other requirement is that it is of general interest - in the case of the colorcube analysis plug-in I think that can be assumed. So please, feel free to re-badge the "Count colors" bug as 1.3.x, but in that case, please send a mail to the devel list requesting a maintainer (also, link to the sources from the bug report, if the link is not already there).

Thanks, Dave.