I'd also like to know some of the stuff Dov is asking about...
I will try.
Let's say I submit a patch against 1.4.3 --- what is the process that the patch has to go through until it is committed?
svn co svn://svn.lyx.org/lyx/lyx-devel/branches/BRANCH_1_4_X write code, test, write, test,... :-) svn diff > mypatch.diff email mypatch.diff to the devel list, with explanations. There are some coding guidelines that you should follow, and will probably miss for the first several submissions. Some people can be picky about this, get use to it and do not take them personally.
Who actually commits it?
At first, we will exam and test your patch. If it is OKed, some of us will apply (with credit to you of course). If you have enough patches, someone will say: give this guy write permission, I do not want to commit for him repeatedly. Then, Lars may give you write permission to the repository.
to which branch?
You usually work on the trunk. Only proven patches will be backported to the stable branch.
How does it get merged into the main trunk (I mean, how are the patches going to be applied to future versions of LyX not in the 1.4.X series)?
See above. Patches go to trunk first, then branch.
Should I be submitting patches against 1.4.X at all, or should I work against the trunk (or some other branch)?
Usually trunk, until you see a need to backport something to 1.4.X. You then produce a patch against the branch and ask JMarc whether or not it can go in.
Now that the trunk is in a (at least semi) feature freeze, how does that affect patches / bug fixes against 1.4.X? (Note that I'm not talking about any major features, or anything that touches core-functionality; rather, it's just minor features or bug fixes that I see myself working on --- if at all --- in the near future.)
Bugs for 1.4.x will be fixed (first in trunk, then in branch.) New features will rarely go to 1.4.x, even if they are accepted for trunk. There are of course exceptions, like the really obvious ones.
For example, the patch I submitted (for chars-transpose), which is now pointed to from bugzilla (http://bugzilla.lyx.org/show_bug.cgi?id=2939) --- should I be picking it up, or I guess it has to be someone with svn commit-access at this point?
Attach you patch to bugzilla, yell at this list and ask people to test it. If not heard, yell again. :-) Cheers, Bo
