On 1 May 2013 22:54, Pavan Yalamanchili <contact@xxxxxxxxxxx> wrote: > For example, the following commit fixes a bug > > https://projects.archlinux.org/svntogit/community.git/commit/trunk?h=packages/cuda&id=09ae51b3d87b7cc8cce2bfd1026e711e29368a8f > > But that change would require a 400+ MB to be downloaded when doing a > pacman -Syu. The next version of CUDA is probably around the corner. I feel > a fix then would have been easier. > > I am not complaining, I can just do --ignore for now. I am just curios to > know if this kind of stuff is thought about. As Gaetan has mentioned, these sort of things usually depend on the maintainer. Now, I do agree that having some sort of consensus is good, but we can't really dictate every minor detail. I personally do delay updates if they introduce a little bit of extra baggage that some users might not want, but there's nothing in that which makes it a better decision than if I were to do otherwise. -- GPG/PGP ID: C0711BF1
