Re: [LEAPSECS] building consensus

From: Rob Seaman <seaman_at_noao.edu>
Date: Mon, 5 Jun 2006 17:13:56 -0700

On Jun 5, 2006, at 4:00 PM, Zefram wrote:

> The algorithm is well-known and immutable. It is independent of any
> authority. One might equally refer to it as the "ISO 8601 calendar"
> rather than the "Gregorian calendar".

ISO 8601 is certainly not immutable. Gregory himself is pretty
immutable,
now that he's dead and departed.

I'm also not sure a proprietary standard can be regarded as "well-
known".

Doubt I can lay my hands on the copy of ISO 8601 from my Y2K remediation
days. Anybody want to comment on whether it actually attempts to convey
the "Gregorian" algorithm within its pages? If so, this would indeed be
similar to ITU-R TF.460-6 in an attempt to assert ownership over an
external standard that should be regarded as "well-known and immutable".

I don't dispute the ITU's right to control time signals layered on
UTC. This
appears to be a treaty obligation that has been historically well
mediated
by the IERS. On the other hand, the underlying definition of UTC seems
to me to be an algorithm "independent of any authority" since its
original
commissioning.

> The international standard, in this sense, is actual planetary
> rotations.

Eh? What's that you say? Can't hear you! Could you say that a bit
louder?

THE INTERNATIONAL STANDARD IS ACTUAL PLANETARY ROTATIONS.

> The difference between days of the Gregorian calendar and seconds of
> UTC is merely that the seconds are an artificial phenomenon.

Well no - seconds are *two* phenomena - one "natural" and one
"artificial".
Would be happy to argue either side of which is which :-)

A second is either the fraction 1/86400 of one solar day, OR
a second is the SI unit of time. The first second varies with time,
while
the second second requires effort to be denumerable with the length
of the day at any arbitrary epoch. Tough nuggies.

We wouldn't be having this discussion if the SI unit of time had simply
been called the "essen" as originally proposed. Then the distinction
between these rather different meanings would be obvious and the need
for an actual time-of-day solution indisputable. Eventually one second
will equal 1.5 essens. What would the suggestion then be for a
zero fuss, lotta muss "solution"? (Ein gedankenexperiment,
don't try this at home, kids.)

Nothing about the ALHP mitigates the originally described quadratic
dilemma.
A leap second is a leap second whether it is delivered one or 3600 at
a time.
In fact, the 600 year estimate for the first leap hour needed by the
ALHP assumes
that a leap hour will be triggered when TAI diverges from UT1 by a
half hour.

Which is to say that the Absurd Leap Hour Proposal demands 3600 leap
seconds
be issued over a period that would otherwise require only 1800 :-)

Rob
NOAO
Received on Mon Jun 05 2006 - 17:14:22 PDT