Santiago Vila <[email protected]> wrote: Hi,
> This package contains just architecture independent data, so it would > make sense to make it Architecture: all. The set of backends installed can and does differ between architectures, so libsane-common does not qualify for arch: any. I haven't decided yet how I'm going to handle this wrt full multiarch. The current solution is OK for architectures of the same family (i386/amd64, mips/mipsel, etc) and allows us to get rid of ia32-libs. Unfortunately there isn't even a (meaningful & useful) common core that is guaranteed to be available on all architectures (hurd-i386 screws that up pretty badly). There are a number of potential adjustments to be made once it'll be possible to try out the packages in a real multiarch setting; there are quite a number of dependencies and I don't think they have all transitioned to multiarch yet. Wait & see. JB. -- Julien BLACHE - Debian & GNU/Linux Developer - <[email protected]> Public key available on <http://www.jblache.org> - KeyID: F5D6 5169 GPG Fingerprint : 935A 79F1 C8B3 3521 FD62 7CC7 CD61 4FD7 F5D6 5169 -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
