Re: ghc now almost ready for migration
On 07/06/2011 18:14, Joachim Breitner wrote:
> Am Dienstag, den 07.06.2011, 14:17 +0200 schrieb Mehdi Dogguy:
>> pandoc is not migratable in its current state in sid. And, gitit depends
>> on pandoc. I tried to remove both gigit and pandoc from testing during my
>> tests to see if ghc could migrate, but it doesn't because we need
>> haskell-dummy, which needs both gitit and pandoc through
>> libghc6-{gitit,pandoc}-{dev,doc}. If you want to keep those old -dev
>> packages, then you'll have to fix pandoc.
>
> For fixing pandoc, removal of pandoc on these three arches is
> sufficient, right?
>
and its reverse dependencies, or they have to downgrade the relation to
Recommends/Suggests, afaik.
> BTW, is it a problem to have the dummy packages from haskell-dummy,
> which are arch all, depend on their libghc-*-dev counterpart even though
> it might not exist on all arches? Or is britney satisfied if the
> dependencies of an arch:all packages are satisfiable on at least one
> architecture?
>
iirc, they are only checked on i386, but I could be wrong here… Adam? :)
--
Mehdi Dogguy مهدي الدڤي
http://dogguy.org/
Reply to: