@rbowen I'm a little concerned about the change that Fokko pointed out here. This isn't just a "refactor" as it really is changing the from a consensus vote to a majority rule. That's a significant change that should be called out.

I'm also of the opinion that consensus vote is a good thing. @potiuk pointed out some examples of where bad actors can abuse the consensus model, but we shouldn't assume bad intent and allow the board to step in when there is abuse (which is true for any scenario where the vote mechanics are being abused).

I believe that consensus is a better model than majority rule because the latter sets up an antagonistic model where you just need enough votes to "win" where as the consensus requires that you work together to arrive at an agreed upon path forward. Majority just results in vote stacking and attempts to "control" the project.

Either way, I think this needs to be addressed more publicly as opposed quietly changing the documentation.

It was discussed on dev@community.apqche.org and it is my belief and that of the list that the document was in error. I would welcome your thoughts there if you disagree in any way.

Is this the discussion thread or was there more discussion somewhere else?

I feel like the changes aren't clear from the title or description of the PR and that there is a substantive change being made.

In practice, with consensus voting, people discuss the person first, and an agreement is reached. It is not just a matter of a particular group of PMC members having the most numbers. You rarely see a -1 vote when voting on a PMC member, but I can recall it has happened on a couple of occasions.

That's a valid point. I see why, but that also (and I will keep on repeating it) - it means very likely that I think we should have a separate entry on the "voting" page about it here https://www.apache.org/foundation/voting.html
This page can be interpreted and suggests something completely different. If you take that page literally as "all the types of voting you can have in the ASF" then the only matching type of voting that fits PMC voting is "procedural" - which is majority voting, not consensus.

If it has been discussed in a number of places, and it has not been captured and agreed upon then it's really a bit of "tribal" knowledge and also your interpretation might be different, depending in which "tribe" you are and which discussions you manage to find about it.

Maybe that's a good opportunity to clarify that and capture this type of voting there - explaining WHYs as well (of course as long as we are going to reach some consensus on it).

BTW. Does it also apply to "new committer" voting and if no - why it would be different ? (this is yet another angle of "this is so prone to interpretation").

I think possibly one of the points of confusion here is the use of the word "consensus", which has been a point of confusion for years, since that word has several substantially different uses. Re-reading the discussion here, it appears that I have completely misunderstood the point being made. The notion "consensus" means "universal agreement" is weird to me, and inconsistent with what I have understood for a long time.

https://www.apache.org/foundation/voting.html pretty clearly says that for procedural matters, you need more +1's than -1's.

Consensus, as I understand it, means "an opinion reached by the group as a whole", rather than "can be derailed by one disgruntled person who disagrees with everyone else." The notion that on a PMC of 30 or 40 people you need complete agreement by every single person seems like a recipe for a tyranny of one person, which is completely antithetical to what community governance means to me. I can point to three example ASF projects today, where a single voice is overriding that of everyone else, and that is completely broken.

So, yeah, perhaps this needs to go to a larger audience for discussion.

Two questions in response to Justin's comments here:

1. Yes, I'd like to see those references please.
2. Who are we to say that a particular individual is "without merit" if a majority of the PMC says that they have earned that merit? That's exactly the kind of tinkering that we say that we don't do.

FWIW, I have reverted the portion of the PR that suggests only a majority vote is required. (so that we can move forward, and possibly discuss the other issue at leisure.)

Note that I think that vetoes, rather than majority, is what consistently leads to abuse in our actual projects, while the notion that a majority vote could lead to abuse is purely hypothetical and implies that "we" know more than the actual members of the PMC. But I sense that I am, ironically, outvoted here. :)

LGTM.

Thank you for all of the reviews, and expecially to @westonpace for the grammar/phrasing tweaks and attention to detail!