>The point is, Jonathan, that if you publically announce that NetBSD
>performs less well than FreeBSD in a certain situation, you should be
>prepared to back your statement up with verifiable information that
>allows us to determine (a) whether we care about that case and (b)
>whether we agree with your conclusions.
>
>Saying ``esteemed researcher X gets the following results with test A,
>and I get the following less good results with test B'' doesn't really
>add a lot of information - it just gets people riled up.

   [holds breath  and counts to 20]

Ted, perhaps this is too subtle, but has it ***really**** escaped
everone's attention that if we DID have TIODCDTIMESTAMP support, then
we COULD run such head-to-head tests like this very easily?

That one of the reasons I wanted to do TIODCDTIMESTAMP is to make it
easier to get good timings on NetBSD? And to compete with FreeBSD in
the  arena of timekeeping support?

That if we could support that, intsead of having Charles crap all over
something he knows nothing about, we could actually get good
measurements?