Hi, I’ll reply to other points on this thread later, when more opinions are in (especially from Clint, who is a major contributor in the DHG), but I’d like to quickly adress Am Sonntag, den 06.04.2014, 23:37 +0100 schrieb Colin Watson: > On Sun, Apr 06, 2014 at 01:13:16PM -0400, Joey Hess wrote: > I agree with Joey's position > expressed on his blog a while back that there's basically no convincing > reason not to use the upstream git repository, where it exists, as the > basis for the one used for packaging. There is in this context: Alone the act of checking which libraries on hackage have git repositories, where they are, and what tags (if any) correspond to what hackage release is too much hassle if you are attempting to handle dozends of packages efficiently. Our starting point is hackage and its tarballs; that’s our interface to upstream – an abstraction layer, if you want, and one reason that a handful people can mange the large number of packages we do. I don’t think it would be helpful if some our packages are handled differently from others (depending on whether upstream has a git repo or not). > > For what it's worth, I've had excellent experiences in a few months of > using git-dpm... I like git-dpm as well, much more than git-buildpackage. But it also demonstrates that there is no genuine standard here. And the simplest choice is to _not_ add a layer over git itself... but I digress from my promise made above :-) Greetings, Joachim -- Joachim "nomeata" Breitner Debian Developer nomeata@debian.org | ICQ# 74513189 | GPG-Keyid: 4743206C JID: nomeata@joachim-breitner.de | http://people.debian.org/~nomeata
Attachment:
signature.asc
Description: This is a digitally signed message part