> No, what allowed sarge to go out the door with DFSG violations was an unambigous GR by a majority of the debian developers who decided to include those non-free firmware (and GFDL docs, and some random fonts, and ...), into sarge even though they didn't quite meet the DFSG. > > That vote is not valid for etch though, as we decided to waive that only for sarge, so only a new GR will allow debian to release the current kernel with non-free firmware as part of etch, independently of the migration scripts you are so worried about above.
I disagree. If there's non-free material in main, that's a bug. Nothing in the Constitution or policy says that this class of bugs should be treated differently than all others, therefore the normal rule applies: for each unresolved bug, the release managers are ultimately in power to decide whether it's a release blocker or not. A General Resolution is only needed to _overrule_ the release managers' decision. The reason a GR was needed for Sarge was precisely that the then release manager, Anthony Towns, had stated that the non-free material in Sarge was a release blocker, so that had to be overruled. But if the current release team decides that Etch can be released, then it can be released. In fact, you'd need a GR to force them _not_ to release. Gerardo
