On Saturday 17 February 2007 17:24, Christian Perrier wrote: > > It seems to me that the changes in this case don't warrant a release > > exception at all. Why not just ask the RMs to not accept this > > version? > > Well, of course, also...but, this, I think is the RM decision which I > don't really want to interfere with. I think you should if the string changes do not fix an issue that really affects the usability of the package in testing. Or at least you could make RMs aware of the issue by replying to the unblock request. Something like "please note that this version introduced string changes that make x translations outdated that are currently up-to-date in testing". If one of the valid reasons for freeze exceptions is translation updates, then regressions in the state of translations is also a valid reason to block a migration. Again, this has to be weighed against the "bug" that was fixed by the change. IMO there is a valid alternative to always making translators do the extra work: ask the maintainer to upload a new version _without_ the string changes. > My point is that, whether or not the package gets accepted in etch, > the changes made to debconf templates warrant a further update, which > I intend to do (or at least prepare for the maintainers if they > guarantee that they will do the upload). Sure, the update will have to be done eventually. But it does not have to be done _now_. Instead translators will basically have the full cycle for Lenny to do their updates...
Description: PGP signature