[Lazare]

On procedural grounds.

It hadn't been accepted, as far as Ben knew, by someone with the authority to make that decision, it didn't follow project guidelines, and the contributor hadn't signed a CLA. The revert had nothing to do with grammar and language.

[chc]

AFAIK this is not the case. The contributor had not signed a CLA when he first submitted the pull request, but evidently did sign it well before Isaac merged in the change (he claimed to have done so very early in the original PR thread). As far as I can tell, procedure was followed.

It seems pretty clear that Ben acted somewhat inappropriately — maybe not with his initial rejection, but when he reverted Isaac's merge and "chided" him. The important thing to remember, though, is that one bit of prickly behavior shouldn't be enough to brand someone "an asshole." Heaven knows I've gotten annoyed and said things I've regretted on more than one occasion.

[Lazare]

Well, I admit to not being 100% clear on the CLA. I was going of comments in one of the other threads from someone who seemed to know what they're talking about, but didn't show proof.

Still though, I would make the reverse point: I think the initial rejection was a bit dodgy, but the reversion was clearly justified.

Really, the reversion had nothing to do with the CLA; rather it was made by Isaac, who did NOT have authority to make that commit, and as far as Ben knew, it was unauthorized. Reverting it was justified. (In fact it was authorized by Bert, but Ben didn't know that. Yeah, Ben could have checked with Bert first, but I doubt I'd have done that in his shoes. If I was making a commit to a project I didn't have authority to, I'd probably make a public note who had told me it was okay. Isaac didn't.)

...still, either way, I agree with your conclusion.

[HaloZero]

I thought Isaac accepted the Pull Request and merged it in? Doesn't he have the authority?

[Lazare]

No. Not on the libuv project.

At the time of the controversy, the core developers of libuv were Ben Noordhuis (bnoordhuis), and Bert Belder (piscisaureus), both working for StrongLoop. Nothing is meant to be committed to the libuv tree without one of them signing off on it. To quote Ben's commit message: "@isaacs may have his commit bit but that does not mean he is at liberty to land patches at will. All patches have to be signed off by either me or Bert. Isaac, consider yourself chided."

But the very first comment on Ben's revert is from Bart: "I signed off on it. Just leave it as-is, no need to revert."

In short: Isaac did accept the pull request. He did not have the authority. Someone with authority (Bert) did accept it, however it was reverted (by Ben) under the mistaken belief that no one with authority had accepted it. A simple communication breakdown; why Bert didn't commit it himself (or why Isaac didn't note that he was committing it with Bert's authorisation) I'll never know.