On Thu, Aug 09, 2012 at 12:02:38AM +0200, Tollef Fog Heen wrote: > Am I understanding you correctly in that your answer is «A Recommends > can never be upgraded to a Depends for a metapackage»?
I'd like to respond, somewhat belatedly, to this characterisation of the proposed resolution, since I don't think my position on it has been expressed clearly elsewhere. I do not subscribe to the position that a Recommends may never be upgraded to a Depends for a metapackage, and I think such a position would be unreasonable. There are many cases where requiring the installation of additional software is a clear improvement for pretty much everyone with that metapackage installed, or at least where the costs of undoing a new upstream dependency outweigh the costs of having the additional package installed (which in many cases are negligible, for example the installation of an additional application). However, when an added dependency turns out to be controversial, maintainers are expected to listen to those users who have a problem with it, try to understand why that is, and consider whether the change should be reverted. If the responsible maintainers then decide to keep the change anyway, they should be prepared to justify it to their colleagues, and (in the case where a dispute is escalated within Debian's governance structure) persuade the TC that their position is better. In other words, I don't see this as any different from the common case of collaborative software development. In general, maintainers have the freedom to make changes that they consider sensible, within the general framework of things like Debian Policy. If those changes are disputed, they should provide clear reasoning, or consider reverting them if they find upon reflection that the change was not as beneficial as they thought. If there is a dispute that can't be resolved in the normal way, then they should be prepared to justify themselves as part of the project's dispute resolution processes. That is certainly what I was led to expect when I joined the project. We should not prejudge future disputes by assuming that they will all be resolved in the same way. I don't think there is any reason to make such an assumption. -- Colin Watson [[email protected]] -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/[email protected]
