On Wed, Sep 19, 2007 at 04:10:35PM +0200, Nagy Gabor wrote:
Well, I don't want to hurt you, but my list would be:
- review/apply/write bugfixes
- add new features/code clean-up/improvements...
I'm sure that you are busy, but you may add git access to other "trusted" devels (1 vote for Xavier) to help you. I'm not familiar with pacman developing philosophy, but as I see, only you and Aaron are the active devels who have git access.
There is no need for that, and I wouldn't want it anyway. I push my own changes to my git tree, and when they are good/simple enough, they are merged very fast. Maybe even too fast sometimes, I have messed up several things recently : 1) http://projects.archlinux.org/git/?p=pacman.git;a=commit;h=d34b2c4ed84bc40f4... one of the newline fix there is actually a newline breakage 2) http://projects.archlinux.org/git/?p=pacman.git;a=commit;h=26441cf65ca10d4bf... I noticed there was an issue there, but didn't notice it was done twice.. It's a bit stupid to require 2 commits for this, but well :) 3) http://projects.archlinux.org/git/?p=pacman.git&a=search&h=HEAD&... it took me at least 4 commits to get this right (hopefully it is now, I'm not even sure), the last one making all the previous ones obsolete.
So you shouldn't trust me, I can't even get my own stuff right, even for simple things :) Your patch are much more complicated, especially the last two big ones : fix for sync 1003 and conflicts resolving cleanup. I've spent hours trying to understand the existing pacman code edited by your patches, and it seems I'm still far from understand it perfectly. I have a basic understanding of it, and how you changed them. It does look good to me, so I pushed them to my tree, with the associated comments : http://chantry.homelinux.org/~xav/gitweb/gitweb.cgi?p=pacman.git;a=shortlog;...
About the resolveconflict stuff, I've been willing to rework it a bit, with my suggestions there : http://www.archlinux.org/pipermail/pacman-dev/2007-August/009087.html But I only fixed the two little gcc warnings. I think some refactoring would be nice, but it isn't critical. I still have another question about it, but I'll reply to your previous mail instead.