On Wed, Jan 14, 2009 at 8:56 PM, Bryan Ischo <bji-keyword-pacman.3644cb@www.ischo.com> wrote:
OK, I can modify my patch according to this feedback.
What is the best way to do this? Should I make a new patch assuming that my patch 1 was accepted and applied? Or should I make a replacement patch 1 and send that out on the assumption that my other patch will not be applied?
When we give feedbacks on patches on the ML, we always expect you to resubmit fixed patches. And not fix on patches. For example, patch 5/5 is a fix of patch 2/5 so that's not perfect. Otherwise about your patchset, I am still not convinced that this is the smallest and best implementation possible. I would say the patches themselves are readable and well written, with good comments, etc. But I would still like to give more thinking to it and to explore other alternative ways. Last thing, Nagy seems to have given these patches a closer look, and he did a lot of good work on that deps.c file in my opinion, so I would like to know his overall feeling about this patchset now 8)