William A. Rowe, Jr. wrote:
Henri Yandell wrote:
Interested in what other views are there.
My view:
* Not mirroring tar.gz releases is dumb. Infra aren't worried about
the bandwidth and it's pointless to not put things on mirrors.
I'll concur.
* Pulling releases because something didn't graduate is bad.
I laugh - of course it isn't 'bad', but there's no reason not to pull
down a disbanded /dist/incubator/deadproject/ site, since they live
for next-to-eternity in archive.apache.org.
My only concern is that incubator releases are more prone to IP and
other issues of a legal nature than a graduated project's release, so
I still think they should be cloistered. But I'm done advocating for
it, so unless others have legit objections, fix it already. Roy's
already done a good job advocating -for- www.apache.org/dist/incubator/
as the distribution point of incubating projects.
I have to completely disagree with this last statement. Incubator
releases are vetted 1000% more than your average TLP release I would
venture. To get a release out of the incubator, you have to run it by
the entire IPMC, and have it reviewed. That, certainly in projects I've
been involved in, has led to a lot of legal introspection. That level of
introspection isn't demanded of a TLP (by virtue of not having to get an
external agency to approve a release).
Thus I'd suggest that incubator projects stand some chance of being in
better shape than those of TLPs.
Personally, I think that is quite natural, and to be expected, in terms
of how meritocracies tend to be better at managing/policing their
borders than what happens inside.
Regards, Upayavira
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
