On Thu, 29 Jul 1999, Joseph Carter wrote: > On Thu, Jul 29, 1999 at 01:49:13PM +0200, Santiago Vila wrote: > > > > I dislike the idea of packages pre-depending on a specific version of > > > > base-files, this imposes a certain order in the upgrade. > > > > > > > > I think there is no hurry and we should stick to /var/spool/mail for > > > > potato. > > > > > > Even if we do this we must still declare the dependency I would imagine. > > > > Yes, and it would still have to be a Pre-Dependency, but a package in > > potato+1 which Pre-Depends on the base-files in potato is not a great > > deal, considering that almost everybody will upgrade from slink to potato > > first. > > I don't think we should try to enforce that people use /var/mail now, it's > just good to get it into the policy now.
So: Could you modify your proposal so that the change from /var/spool/mail to /var/mail is postponed after potato? I will not object to create a symlink /var/mail to /var/spool/mail, but the right time would be in the frozen stage of potato. Or maybe we could modify policy so that it reads: "Migration from /var/spool/mail to /var/mail is postponed until further notice" (i.e. Waiting for FHS 2.1 to be finallly released :-) > FHS issues should not be bugs in potato---next version maybe, but not > potato, I agree. For this reason we have to be careful in the wording. -- "63777cea2305edfd2b4a474033ac6aab" (a truly random sig)
