Hi, Am Montag, den 15.02.2010, 09:44 +0100 schrieb Joachim Breitner: > This means we need to do binNMUs, and we need to do them “blindly” > because we can not tell from the outside what ABI versions ghc provides. I just did sourceful uploads of mtl, X11, and xmonad, because users were complaining. I’m a bit reluctant to schedule binNMUs, because it’s hard to get figure out the right ordering of binNMUs: Just scheduling will give lots of FTBFS because packages are tried before their dependencies have been fixed an uploaded. > We could consider using dh_haskell_provides also for ghc6 and ghc6-prof, > so that they list the ABI is they provide. Any other ideas? I really think we need to run dh_haskell_provides in ghc6: Haskell packages just don’t work with non-matching ABI ids, even for base, so the packaging relations should reflect that. Note that haskell-devscripts is probably not a good build-depends for ghc6 (it depends on ghc6, haddock and hscolour). But I guess dh_haskell_provides could be copied to ghc6 without a problem. I know that ghc6 provides a bunch of packages (33 to be exact), and it would look slightly ugly to have such a long Provides line, but I think it’s worth it. Also note the nice benefit that, for example, a patch that changes the Template Haskell ABI will leave the base ABI hash inplace, so only packages actually using Template Haskell need to be rebuild. Kaol, do you agree so far? 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: Dies ist ein digital signierter Nachrichtenteil