Sam Roberts

seems reasonable, or else it should be an option. note that it is close on "exec" not close on "fork"

@nodejs/http

I think it would be useful to refer to what is being limited with the names from RFC's BNF. What's the "start line"? Nothing in RFC 2616 is called that....

Could you track down the actual patch that is being applied to the http_parser and link to it? None of the links above lead to it, they just mention it...

Ouph. No tests. I'd say "libgit2 should move to llhttp", but that doesn't have a URL parser ;-). I don't personally object to the change, if it matches RFC, but...

This repo is currently "lightly maintained"... I'm not sure how quickly you can get a definitive answer. Sorry. But PRed code usually gets more comment than a statement of intent....

Am I mistaken, or did everyone who has done a release in the last year respond in the thread above? I wonder if a yearly check that people are still...

> Suggestion to add semverness to the commit message. Reasonably often we change our mind on the semverity of a commit after it has landed. In theory, that shouldn't happen,...

I understand its not the goal to make it immutable, but it would be a side-effect. I agree that having semverity associated with entire PRs instead of individual commits also...

Why not just check that a module is using the latest version of nan, since the latest will cover all node versions, and anything other than the latest will not.