On Wed, Jul 11, 2007 at 09:55:01PM +0200, ngaba@petra.hos.u-szeged.hu wrote:
Your own patch has 11 empty lines. I removed a few of them to check, and it resulted in a malformed patch, eheh. Oh well, I shouldn't have asked this question, now I messed up your thread :d Well, the .diff file has no empty line (it may contain at least one space), and the .py file contains 6 empty lines :-P
I try to save the thread now :-D: As you noticed, here we must assume again, that alpm_list_add adds entry to the end of the list. You suggested earlier (http://www.archlinux.org/pipermail/pacman-dev/2007-June/008543.html <- and then you can remove the ready variable from that patch), that we should use alpm_list_add_last or similar to be perfect. I like your idea, and an O(1) alpm_list_add_first also would be useful (I mentioned here: http://www.archlinux.org/pipermail/pacman-dev/2007-April/008202.html)
I agree, I don't see why the current list implementation couldn't be improved a bit (without major changes everywhere around the code), even if the goal is to eventualy move to kernel list, which would probably require much larger change. The first discussion about this stuff I found is 4 months old already : http://www.archlinux.org/pipermail/pacman-dev/2007-March/007374.html But maybe that move to kernel list could be done quicker than I think, I actually don't really know.