Date: Sun, 01 Jan 2006 14:23:37 -0700 (MST) From: "M. Warner Losh" <imp@bsdimp.com> To: dfr@nlsystems.com Cc: phk@phk.freebsd.dk, freebsd-current@freebsd.org, current@freebsd.org Subject: Re: FreeBSD handles leapsecond correctly Message-ID: <20060101.142337.33509370.imp@bsdimp.com> In-Reply-To: <200601011643.35578.dfr@nlsystems.com> References: <73774.1136109554@critter.freebsd.dk> <200601011643.35578.dfr@nlsystems.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In message: <200601011643.35578.dfr@nlsystems.com>
Doug Rabson <dfr@nlsystems.com> writes:
: On Sunday 01 January 2006 09:59, Poul-Henning Kamp wrote:
: > http://phk.freebsd.dk/misc/leapsecond.txt
: >
: > Notice how CLOCK_REALTIME recycles the 1136073599 second.
: >
: > Happy new-year!
:
: Mmmmm... Leap seconds... I can hear Warner grinding his teeth from
: here :-)
Having experienced on has not changed my deep, and abiding feelings
for leapseconds in the least. I've spent about 120 hours on leap
seconds in the past couple of years. For something so damn simple,
these are a huge pita. Most of the time has been determining what
real devices will do over a leap second, which information is reliable
and which information lags and how. There's too damn many variables
to know what information you can rely on and what information you have
to 'filter' and how. You can make the data streams more reliable at
the cost of code complexity and inflexibility.
I guess that's a long way of saying that I hate Leap Seconds. They
are too damn complicated for such a simple concept :-(
Wanrer
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060101.142337.33509370.imp>