I've noticed that a lot of the patches that are submitted to the patch
tracker are quite sloppy. This makes them much less likely to be
incorporated into the core. Here's a couple guidelines which should
help:
- use diff -uw (-u provides context, which is good for humans and the
patch program. -w ignores whitespace changes)
- the patch should only included changes that are relevant. There
should be no other changes, include whitespace changes in the patch.
- the patch should be for one feature. It is much easier to figure out
what's what, and then it will less likely be skipped because one
feature in a multi-feature patch was nixed.
Way back at pd~conf, we all agreed to give this a shot, I think it can
really work, but everyone has to do their bit. Do you think Linux
Torvalds reads sloppy, untested patches? I think Linux has a well
established model which we can emulate on a smaller scale.
.hc
________________________________________________________________________
____
If you are not part of the solution, you are part of the problem.
- Eldridge Cleaver