Hi, Am Montag, den 21.12.2015, 18:28 +0300 schrieb Dmitry Bogatov: > * Joachim Breitner <nomeata@debian.org> [2015-12-21 15:57:46+0100] > > Hi Dmitry, > > > > just checking: Why do you re-add maybet? It was just removed from > > the > > archive a few days ago: https://bugs.debian.org/807768 > > Just because it is in packages.txt. Not any more: $ grep MaybeT packages.txt and before, it was marked as "obsolete", which is the tag for something that is packaged, but scheduled to be removed. Are you on the latest version of the package plan? > > Also, I would like to discuss following points: > > * `stack` in out packages.txt is 0.1.8.*, but > new version is 0.1.10.*. Shouldn't we upgrade to lts-3.18? Yes, once the current migration is over. We can of course already start planning upgrading to LTS 3.18 in the package plan. > * Why should we follow stackage, instead of keeping only > packages, needed for executables, like Go team do? > > Inevitable we are lagging behind, and mere mortal, like me, who > lives in stable/backports, can't get benefits of packaging, that > happens in sid. > > In pre-stack times, it was clear, why maintainace by DHG is good > thing: cabal is unreliable, and debian packages are rock-solid. > But now stack is reliable also. Are you suggesting to stop maintaining any library that is not strictly necessary as a build dependency of a program written in Haskell? I have thought about this, and it is definitely worth considering. On the other hand, the better our processes are, the less expensive it is to just provide a few extra libraries as well. Note that a large number of our packages are indeed required as dependencies of executables such as gitit or git-annex. Greetings Joachim -- Joachim "nomeata" Breitner Debian Developer nomeata@debian.org | ICQ# 74513189 | GPG-Keyid: F0FBF51F JID: nomeata@joachim-breitner.de | http://people.debian.org/~nomeata
Attachment:
signature.asc
Description: This is a digitally signed message part