[fitsbits] Re: leap second alert

[Steve Allen]

On Thu 1999-12-16T20:05:37 +0000, Rob Seaman hath writ:

> First - what does astronomy need?  (And therefore what must FITS support?)

FITS would surely like to be able to name a timescale which will be
widely available far into the future.  Two years ago UTC looked like
the right thing.  So long as something called UTC remains widely
available I doubt that the FITS Y2K agreement need be amended because
its appendix provides for specification of timescale when necessary.

> Second - what are the resulting implications of this for civil time?

I think that "astronomy" has set a pretty good precedent following the
1925 fiasco which destroyed the utility of GMT as a technical term.
Look at the activity of the IAU over the past few decades and the
proliferation of new, subtly-better-defined timescales.  I'm not privy
to the actual discussions of the committees, but it looks like the
astronomy community learned a lesson in 1925:

	Never redefine an existing timescale.

Don't risk confusion.  If the understanding, theory, and or
specifications on which an old timescale was based turn out to be
inadequate, define a new timescale that addresses the problem.

The result of this is that anyone throughout history can make sense of
a record using these timescales.  If someone now records an event time
using TDT (say because the data-taking software hasn't yet been
upgraded) we may say "tsk, tsk, the IAU says you should be using TT
now", but we can also go through a well-defined bit of algorithmic
grunt work and convert that time to whatever timescale we want.

One might counter that UTC has already changed its rules, and/or that
the rules for UTC have always been subject to interpretation by the
international committee, but I'm sure this would violate the spirit
and intent if not the letter of the original UTC definition.

--
Steve Allen          UCO/Lick Observatory       Santa Cruz, CA 95064
sla at ucolick.org      Voice: +1 831 459 3046     http://www.ucolick.org/~sla
PGP: 1024/E46978C5   F6 78 D1 10 62 94 8F 2E    49 89 0E FE 26 B4 14 93