[LEAPSECS] Schedule for success
seaman at noao.edu
Wed Dec 31 11:25:31 EST 2008
Poul-Henning Kamp wrote:
> Until now I have never seen any indication anywhere that longitude
> or lattitude would be affected in any way.
> Pray, show us the documentation for this hyperbole.
Ah! That explains it. You believe documentation is the master
definition of a system.
Rather, requirements rooted in real world issues are the master.
Whereas the requirements listed in the documentation are always
incomplete and never completely self-consistent.
If this were a religious argument (which so many technology issues
resemble), there would be a fundamentalist versus a liberal
interpretation of every point.
Greenwich Mean Time has a clear definition that has nothing to do with
technology. The prime meridian likely has an official definition
distinct from GMT, but in origin and practice the two are inextricably
It is through the actions of CCIR as inherited by the ITU that UTC was
linked to GMT.
Presumably nobody disputes that the ITU seeks to sunder this connection.
The Earth has a large moon. As a result, the Earth's rotational
angular momentum is being transferred to the Moon's orbit. The Earth
Civil time is related to diurnal rhythms. As a result, a timekeeper
built on non-diurnal rhythms, like an ensemble of atomic clocks, will
inevitably diverge from a timekeeper for civil time.
The ITU proposal is to flip the interpretation of the last paragraph.
There are world's of possibilities in how this might be done. The ITU
has only ever considered the most garish option.
So civil time is to tick, tick, tick at a steady rate and the Sun in
the sky is to fend for itself. On some schedule (some of us prefer
quick and some slow), intercalary adjustments will have to be made.
The implicit ITU position appears to be that these adjustments will be
made through the system of timezones that were conveniently set up in
the 19th century. Consider this a preadaptation in the evolutionary
Well, it just so happens that one of the timezones, actually the
keystone of the whole system, is tied to GMT. If the ITU's implicit
assumption as to the future mechanism for intercalary adjustments is
to work, then GMT must also be redefined to continue to serve as an
approximation to the new UTC timescale.
So the ITU demands both that GMT != UTC and that GMT == UTC.
If the definition of GMT changes to be ITU-centric, rather than
astronomical, the question of the definition and meaning of the prime
meridian is left in limbo. I certainly don't think the nice folks at
the ITU have missed this point, so I am requesting that they make it
evident in their proposal. It would also help if the proposal were
made publicly accessible.
So, I just typed that in without review. I'll read through it once to
look for typos, bad phraseology and so forth. I had no problem
navigating from top to bottom - and could continue to add levels of
details - because the logic of the message is rooted in the logic of
how the world works (and 9 years of practice :-)
The world, that is, is automatically self-consistent. Documentation
can be assumed to be always inconsistent. Just because nobody has
said anything about implications for latitude and longitude it doesn't
mean that time and space aren't related in various ways, prosaic and
Ok - my read through is finished. The only change I made was to add
the smiley and break the last paragraph at that point. By comparison,
a message layered on an analysis of ITU standards would be very
laborious to construct since one would have to keep referring back to
other documents and glosses on those documents.
More information about the LEAPSECS