Re: FreeBSD handles leapsecond correctly

In message <20060102211956.GA10928_at_merlin.emma.line.org>, Matthias Andree writes:
>On Mon, 02 Jan 2006, Poul-Henning Kamp wrote:

>No offense, but I simply don't get your point.

Right, I'm getting that feeling here too.

I think the only way for you to catch up is to study the subject
matter much more carefully.

In particular you need to understand the "ownership" of the various
time/frequency related units, standards and timescales much better
than you do now.

A good place to start is the classic 1968 Essen article in Metrologica:

   http://www.leapsecond.com/history/1968-Metrologia-v4-n4-Essen.pdf

And next read Steve Allens very comprehensive but rather politically
slanted web-page about the leap-second problem:

   http://www.ucolick.org/~sla/leapsecs/dutc.html

Next you want to read RFC1305 to understand how NTP works, pay
particular attention to the timescale/representation parts.

Finally, you may want to buy and explore "Calendrical Calculations"

   http://emr.cs.uiuc.edu/home/reingold/calendar-book/index.shtml

And then, feel free come back and tell me and the rest of the world
how easy this problem can be solved correctly.

>Another question: Did you
>mean to write "FreeBSD handles leapsecond in POSIX compliance" for the
>subject?

POSIX compliance is assumed for FreeBSD and no other relevant
computer standardization exists for this area.

I assumed this to be self-evident, but I guess I need to lower
the bar in future emails.

Poul-Henning

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk_at_FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.