Re: Seeking clarifaction of tai64nlocal and leap seconds

On Wed, Sep 19, 2012 at 12:34:29PM -0700, Peter Wolfenden wrote:
> Brian,
> 
> Leap-seconds should really only matter when converting from a "pure" TAI64,
> TAI64N, or TAI64NA timestamp (which simply counts seconds and fractions
> thereof and does not care about hours, days, years, etc.) into a
> human-readable format.

When you put 'pure' in quotes, it leads me to believe there are
non-'pure' TAI64x timestamps.

Is that correct, and is that what multilog presents?  (Non-pure,
that is...)

Arguably, this all moot; what I'm doing it trying to resolve
timestamps as produced by various applications, and those presented
by multilog/tai64nlocal.

It looks like things are actually well-behaved, but I'm getting
caught out by java having it's own timezone files, and not providing
a distinction between the two sets of timezones as written up on the
tai64nlocal webpage:

  http://cr.yp.to/daemontools/tai64nlocal.html

I do appreciate the prompt feedback...

> 
> Cheers,
> 
>     Peter
> -- 
> Peter Wolfenden
> 650-801-7714

-- 
Brian Reichert				<[email protected]>
BSD admin/developer at large