Re: [LEAPSECS] Precise time over time

From: Rob Seaman <>
Date: Fri, 12 Aug 2005 12:06:39 -0700

On Aug 12, 2005, at 10:36 AM, Poul-Henning Kamp wrote:

> I have tried to reach out to you, tried to offer various models
> which might work for both you and me, but all I get back is "they'll
> just ignore us", "we can't do it", "we don't know if this will work"
> etc etc.

I wouldn't have invested my own time writing a point-by-point reply
if I weren't attempting to be responsive to the need for action.
That said:

     - Before changing policy, a coherent and comprehensive proposal
should be composed.

     - Before drafting a proposal, a community consensus should be

     - Before reaching consensus, the full range of options should be

You have offered a single "model" that amounts to faith that we can
predict UT1 50 years in the future, and posits that leap jumps of
uneven height will be acceptable to various stake holders of civil
time, and exhibits a childlike belief that builders of civil time
dependent systems will simply do the right thing.

I have refined a suggestion I made 4.3 years ago and allowed myself
to explore an option (leap minutes) that I truly believe to be
unacceptable based strictly on the merits.

There are clearly many more options that could - and should - be
explored before even attempting to build a new consensus. I'm sorry
if you don't like my emails - but there is nothing about them that is
anything but open to considering options. "Considering" includes
pointing out negative ramifications, not just barging ahead to some
imposed solution that is slightly less evil than one that would be
laughable if it didn't appear to be being taken seriously.

> Here is a concrete proposal:
> A.
> Until 2009-12-15 we use the current rules, after
> that date we use the rules in section B (below)
> B.
> 1. leapseconds can be inserted at the end of any month.
> 2. All leap-seconds for the period [2009-12-15 ... 2060-12-15]
> must be announced before 2009-05-15.
> 3. All leap-seconds for after 2060-12-15 must be announced
> 50 years before it is scheduled to happen.
> 4. There is no upper bound on DUT1. DUT1 depends only on
> the quality of the model(s) used to schedule leapseconds.
> 5. Timeservices and GNSS systems should broadcast DUT1.
> Is there anything in this which is unclear to you ?

There aren't things here that are unclear to yourself? This isn't a
proposal, it is a notion. A proposal would be fleshed out with
discussions of implementations, legal implications, a detailed plan
for deployment, risk analysis, economic impact statement and so forth
and so on.

> Will you work actively to get such a proposal in front of WP7A
> and work to get it adopted instead of the current proposal ?

No. I will work to get the current proposal rejected outright and a
process adopted that is responsive to real world concerns and
communities, not just a narrow and obscure set of special interests.
In the mean time, I will value discussions of the boundaries of civil
time possibilities.

> Yes, you have the time from now until 2009-12-14 to do that.

Are you really of the belief that four years will make any difference
whatsoever to the worldwide response to either leap seconds or a
growing DUT1? Will the ITU identify funds to notify all potentially
affected communities?

> Nobody is in a position to mandate the format or contents of
> time signal transmissions [1].

This is precisely what the current standard exists to do. It is only
that the ITU is asserting authority over UTC and that the IAU has
abandoned same that we are in this situation. Any civil time
proposal that does not include an analysis of instrumentalities is
void of meaning.

Rob Seaman
National Optical Astronomy Observatory
Received on Fri Aug 12 2005 - 12:07:00 PDT

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