Ben Harris <bjh21@netbsd.org> writes:
> > I'd strongly suggest we fix that to use getrusage.
> 
> Agreed.  This doesn't help with existing binaries, though
> 
> > It is one of only two apps that actually use times in our sources,
> > actually.
> 
> What's the other?  Presumably it should use getrusage too.

timed, which uses it to get time (it actually uses the return value
and ignores the usage information.)

> > I don't think that is necessary -- old sources assume CLK_TCK, as was
> > defined in the old POSIX, and not _SC_CLK_TCK. Anything properly using
> > _SC_CLK_TCK is new.
> 
> Erm, _SC_CLK_TCK has been in POSIX since 1996 at least.

Yes, but before, CLK_TCK was the defined interface, and in NetBSD the
man page said the units were CLK_TCK.

> > I'm very unsure it is. We are talking about something that really,
> > really impacts almost nothing. I suspect that no one would actually
> > notice except for our paranoia about it.
> 
> Perhaps.  I feel that this is an instance of "if a thing's worth doing,
> it's worth doing right", so if we're going to provide times(), we should
> do so with full backward-compatibility.

Fine. Please version it, then. I want to bump CLOCKS_PER_SEC to 1E6 at
the same time -- would you be willing to version clock() at the same
time?


--
Perry E. Metzger		perry@wasabisystems.com
--
NetBSD: The right OS for your embedded design. http://www.wasabisystems.com/