Subject: Re: Please Revert newlock2
To: Bucky Katz <>
From: Greg A. Woods <>
List: tech-kern
Date: 02/21/2007 14:46:39
At Tue, 20 Feb 2007 19:32:17 -0800,
Bucky Katz wrote:
> But I "or someone else" did not remove functionality that the NetBSD
> user community was relying on without even asking if they were.
> This is the crux of this whole discussion.

But that just shows that you really do not yet "get it" with respect to
the use of -current by a third party developer such as yourself.

You should not have been relying on -current in the first place, thus in
reality nobody has so far removed anything you should be depending on.

If your argument were originally stated as something like this:

	"Can someone please make sure that SA will work with newlock2 so
	as to provide efficient M:N threading via libpthread on
	uniprocessor systems before the newlock2 code reaches a release

then you you would have been on the right track and that would really
have been all you needed to say for the time being (except perhaps to
also offer your help in achieving that goal).

> Actually, since no one has even bothered to benchmark newlock2 against
> pthreads, as far as I can tell, you don't have any evidence that you
> haven't traded a decrease in threaded performance for a simpler
> implementation.

Since you're talking about the bleeding edge of -current I think the
following sound quote and its derivations over time, especially the last
one, apply:

Premature optimization is the root of all evil in programming.
	-- C.A.R. Hoare

We should forget about small efficiencies, say about 97% of the time:
premature optimization is the root of all evil.
	 -- Donald Knuth (re-quoting Tony Hoare)

The key to performance is elegance, not battalions of special cases.
The terrible temptation to tweak should be resisted unless the payoff is
really noticeable.
	--Jon Bentley and Doug McIlroy (paraprhasing Tony Hoare)

						Greg A. Woods

H:+1 416 218-0098 W:+1 416 489-5852 x122 VE3TCP RoboHack <>
Planix, Inc. <>       Secrets of the Weird <>