On Fri, May 29, 2009 at 8:21 PM, Dario Freddi<drf54321@gmail.com> wrote:
Hello people,
Here goes another chapter of the fetch callback saga. I finally managed to get a patch working on my copy and completed the integration with Shaman. I have attached a patch that is almost completely based upon Sebastian's one, however it does not abstract the logic of xfercommand out to pacman. Now alpm works this way:
1. If the frontend registered a fetch callback, use that 2. If there is a xfercommand, use that 3. If there's nothing, usual behavior
Do you realize your patch is the first one Xilon submitted : http://www.archlinux.org/pipermail/pacman-dev/2009-February/008198.html Then Aaron made the suggestion to move xfercommand handling to pacman frontend, and to use fetch callback to register it. Why did you revert this change, why is it problematic for you? And why didn't you say anything back then?