Subject: Re: Please Revert newlock2
To: Bucky Katz <firstname.lastname@example.org>
From: Nathan J. Williams <email@example.com>
Date: 02/17/2007 18:48:47
Bucky Katz <firstname.lastname@example.org> writes:
> "Nathan J. Williams" <email@example.com> writes:
> What's wrong here is that it shouldn't go into head until it at least
> passes a minimum of tests off head.
You asked for several things:
> 1) Compiles and passes simple tests on all archs
This is not reasonable to ask a single developer to do.
> 2) Retains m:n support for uniprocessor as an option
This is an unrelated feature request. Given the reasons around
switching from M:N to 1:1, I think the burden of making this work
should be on the people who want it, not the people who want 1:1. I
wrote the M:N support, and I'm not even bothered enough to keep it
> 3) a reasonable amount of pthread testing
This is more reasonable, but again I think that it's just out of scope
for one developer. That doesn't mean that that one developer shouldn't
introduce the change; it means that the developer should help move
things along as problems are reported. From what I've seen, that's
> Fundamentally, you've missed a state. There's "far too broken",
> "broken", and "less broken." If you introduce "far too broken" code
> into current, your desire for user testing backfires as people move
> away from it because it's too unstable.
> It should be a ground rule of any project that you don't introduce
> code if you know it breaks architectures and you supposedly have those
> fixes coming.
This assumes that -current needs to work all the time. I'll repeat
this: I don't agree. This is *exactly* what -current is for. Again, if
this bothers you, wait for a release.