Fwd: [Re: Compliance of apache release policy in practice]
Forwarding from members@ (with permission) some comments about PMC's and
releases, apropos the announce_at_a.o moderation discussion last week. The
bracketed edit is by the original author.
> Roy T. Fielding wrote on Fri, 29 May 2020 03:52 +00:00:
> > [...] we expect TLPs to be self-governing.
> > Generally speaking, that means we expect a TLP to adhere to ASF
> > policies when making decisions about what to release, and to use their
> > own judgement regarding how that applies to a specific package. In
> > turn, the PMC relies on the individual judgement of each PMC member to
> > evaluate a package and decide whether the proposed release is better
> > than the last release by the same name (within the same version scope).
> > That decision is made by binding vote of the majority of PMC members.
> > It's very unlikely that anyone outside the specific project will have
> > sufficient understanding of the existing code base and releases to make
> > a "better" decision, unless there is something in the package that
> > would be illegal to release and the project is not yet aware of that
> > fact.
> > [...] and it simply is not true that a release needs to be held up
> > until it meets one person's ideal. A release should go out as soon as
> > it is legal and better than the last release.
> > Specifically, a project performing a beta release is in a better
> > position to make a release decision than a drive-by member with an
> > opinion based on their experience in other projects/podlings. [Three people
> > voting +1 and a majority positive is enough to make a decision.]
> > > What is [our] general opinion about that? Are there regulations in place to address such issues in TLPs?
> > We don't have a general opinion. We have people who are responsible for
> > making judgement calls for their own projects. As you can see, someone
> > like me (who wrote the license and the original Apache release
> > guidelines) might have very different minimum requirements than someone
> > else. That's fine. What matters to us is that the people on the
> > project, who have been granted the power to make this decision on
> > behalf of the ASF, have done so as a PMC (using a lazy-consensus
> > majority vote).
> > We trust each PMC to do the right thing. They are the ones doing the
> > work to improve each release. They are the ones who decide when to
> > release. They are the ones who will help out their users if a release
> > turns out to be a dud.
Received on 2020-06-06 07:21:12 CEST
This is an archived mail posted to the Subversion Dev