RE: [LEAPSECS] the legacy of ephemeris time

From: Seeds, Glen <Glen.Seeds_at_Cognos.COM>
Date: Mon, 8 Dec 2003 10:36:58 -0500

Everyone would be happier if time were an easy thing to measure and
communicate. What we're discissing is the most practical way to deal with
the fact that it's not.

I agree wholeheartely with Steve Allen. The confusion could be eliminated by
mandating that whatever we pick for uniform time, it will always be
represented as a simple number, never be directly converted to anything that
looks like civilian time representation, and never directly consumed by air
traffic controllers or any other human. Human time should always be based on
something that is (1) directly convertible to local time, and (2) has a
relationship with the aforementioned uniform time that is (a) historically
known, and (b) predictable at least a few months into the future. Based on
these criteria, TAI seems to be a good candiate for (non-human) uniform
time, while UTC with leap seconds predicted a year in advance, plus fixed
local offset (0 for the military and other global human time bases), is a
good candidate for human time.

Of course, we still have this massive legacy of systems that don't follow
these rules...
  /glen

-----Original Message-----
From: Leap Seconds Issues [mailto:LEAPSECS_at_ROM.USNO.NAVY.MIL]On Behalf
Of Poul-Henning Kamp
Sent: December 7, 2003 4:27 AM
To: LEAPSECS_at_ROM.USNO.NAVY.MIL
Subject: Re: [LEAPSECS] the legacy of ephemeris time


In message <20031207064453.GA6682_at_ucolick.org>, Steve Allen writes:

>For the sake of air traffic controllers, however, I wonder if the
>solution is not to demand the use of a uniform time scale and to
>enforce a lack of confusing by expressing that in a notation that is
>suitable for a uniform time scale. I.e., rather than use the
>inappropriate (and confusing) hh:mm:ss notation, why not require that
>the uniform time scale be reported as 5 significant digits of decimal
>fraction days or the 4 least significant decimal digits of a count of
>seconds?

...

So to go to decimal time (again, didn't the french try this during
the revolution of Paris ?) you would have not only traffic controllers
and their computers, but also a major slice of the airline industry
to convert.

As long as leap-seconds only result in "a few seconds of glitches",
they don't care much about it, the system is designed to cope with
glitches and the staff trained for it, so as long as it gets back
on track fast, they don't worry too much about it.

Still, they'd be happier if they didn't have to cope with leapseconds.


--
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.
This message may contain privileged and/or confidential information.  If you
have received this e-mail in error or are not the intended recipient, you
may not use, copy, disseminate or distribute it; do not open any
attachments, delete it immediately from your system and notify the sender
promptly by e-mail that you have done so.  Thank you.
Received on Mon Dec 08 2003 - 07:38:13 PST

This archive was generated by hypermail 2.3.0 : Sat Sep 04 2010 - 09:44:54 PDT