Subject: Re: Avoiding microtime(9) global state
To: Frederick Bruckman <fredb@immanent.net>
From: Nathan J. Williams <nathanw@wasabisystems.com>
List: tech-smp
Date: 06/30/2004 21:47:57
Frederick Bruckman <fredb@immanent.net> writes:
> We did that, too, until very recently. monotonic time is weird, and
> hardly anyone should use it, so we could just as well make it strictly
> monotonic, which would cover the use you described.
Any time you declare that something is going to be monotonic
increasing (as opposed to merely nondecreasing), you need to stop
calling it "time", because it will always, eventually, become possible
to call it at a rate higher than it can represent while keeping pace
with the clock.
There's a place for functions that return new values on every call,
but I think they should be divorced from any sort of timekeeping.
- Nathan