On 23/06/15 15:32, Christian Brenninkmeijer wrote:
By itself a few hardcoded version numbers are NOT reason to delay a release.
But making sure the license ect are Apache consistent are.
I leave it to others who know the Apache rules better than me to judge if these
are incorrect.
Which is why my vote is none binding
Christian
The real decision for a release comes down to the release manager and
community norms.
The requirements for a release vote are majority of PMC, be legally
clean, and at least 3 +1 PMC but that's just the foundation minimum.
http://www.apache.org/dev/release.html#approving-a-release
For example, many (most?) projects require the code to be buildable,
code to be tagged, tests pass - whatever the community here decides.
http://incubator.apache.org/guides/releasemanagement.html#best-practice
It is the release managers decision.
Once over and above the minimum, they decide whether to respond to
comments, whether to redo a release, what to fix, what to leave until
next time.
Binding/non-binding:
"Binding" would mean PMC.
In the incubator, the release eventually needs a Incubator PMC release
vote. That is best if it is more of a "please check our process".
The PPMC should act like a PMC even if it does not yet count as such.
It's part of showing a project is ready to graduate.
There is the -0 vote: 'I won't get in the way, but I'd rather we didn't
do this.' Counting wise, it's a 0.
Fractions when determining majority are "unclear" and mainly for
indicative points in discussions.
http://www.apache.org/foundation/voting.html
Discuss.
Andy
