[LEAPSECS] Leap seconds decision deferred until 2015

Warner Losh imp at bsdimp.com
Thu Jan 19 20:49:46 EST 2012

On Jan 19, 2012, at 4:36 PM, Rob Seaman wrote:

> Warner Losh wrote:


>> On Jan 19, 2012, at 10:57 AM, Stephen Colebourne wrote:

>>> Consider it an opportunity to find consensus.


>> That is unlikely if people believe it is axiomatic that time is fundamentally "time of day" and not "elapsed time."


> But I don't. Time is both. I believe it axiomatic that systems built assuming otherwise will be frail and fail. Additionally, as I understand the requirements for civil timekeeping, time-of-day-ness is one of those requirements. But just as a requirement is not a specification, it also isn't an axiom. An axiom describes the thing. A requirement describes the people who will use it.


> Which is to say that even in an Asimovian subterranean "caves of steel" scenario, our lives would be tied to diurnal rhythms.


>> These are two fundamentally opposing views of time. And sadly they both agree to about a minute over the next 100 years, so the difference is small. This difference will matter less and less if we become a space faring race,


> This was one of the first talking points put forth by Matsakis. It is unsupported. As you said in an earlier message:


>> Universal time is an abstraction, as you say. It models the synodic day.

This is the axiomatic part I'm talking about. The synodic day does govern our lives, but at the level of +/- 2hr, not at the level of a second. Many schemes could be devised that alter UTC in a less radical way so that it is still steered, but steered ala the calendar, not ala measurements on the middle term (say decade), but with varying rules every 10-25 years to keep the long term average in line.

> The synodic day (or something very, very close) is used on Earth and has naturally been used on Mars for rover operations. In fact, the amplitude of the equation of time on Mars is much greater than it is on Earth due to the greater eccentricity of the Martian orbit. This emphasizes that it is indeed *mean* solar time on Mars that is used for rational scheduling.

This works only when mars is an isolated world. If there were commerce between mars and earth, or other need to regularly communicate, then there would be demand for a uniform time scale for both, just like there was demand for a uniform time between cities that shared proximity 150 years ago.

> There are fully 25 terrestrial worlds - rocky/icey bodies, gravitationally rounded - in our solar system within the orbit of Pluto. (And apparently billions in the Milky Way Galaxy.) Four planets, Mercury through Mars. Two dwarf planets, Ceres and Pluto. Nineteen moons: our Moon, Charon and 17 around the gas giants. The Sun on Pluto is dimmer, but still 250 times the light of a full Moon on Earth.


> On each of these worlds the only common yardstick for time is the mean synodic day: the sidereal rotation period of each, adjusted by one day per year for lapping the Sun (or exosun). This is independent of the tilt of the axis, the eccentricity of the orbit, orbital resonances, retrograde rotation or even of being in orbit around another world. Some variation applies to worlds orbiting binary stars. Io and Europa and a handful of the other moons orbit with a faster velocity around their planets than the planets do around the sun; as a result the moons make loop-de-loops where they literally move backwards. And it is mean solar time that brings clarity to the entire orrery.

And yet elapsed time would be universal for all these systems, to a very high order (ignoring the small relativistic effects).

>> I also don't have much hope for things getting better unless people are willing to budge on other issues. For example, we could not change a thing, but announce things 1 year in advance rather than 6 months in advance. And if that goes well, we could push that to 2 or more years.


> Details remain to be sorted and some folks may find it only part of a solution, but we already have good consensus on this.


>> If we relax the DUT1 to 2s or 3s, then things could be announced even further in advance, which would ameliorate one of the operational difficulties of the current system. History shows that DUT1 of 1s is an arbitrary limit. Some folks wanted .1s, others .8s, some 2s. There's nothing magical about DUT1 <= 1s fundamentally, so that axis of the problem should be explored (yes, I know it is a change, but not one so fundamental that it couldn't be phased in with proper studies before).


> I believe there is a willingness to explore the engineering phase space here.

That would be good.

>> So I'm not too optimistic since all the focus has been on 'let's just junk them entirely' with little middle ground explored.


> Exactly! We've spent a dozen years able to do nothing else but fend off the most ridiculously over-simplified proposal from SG-7. If we flesh out different options we could make significant progress in a trade-off study in time to inform RA-15.

I think the ITU should decide today that eliminating leaps on a scale less than decades is unlikely to succeed and start looking for small steps to move in that direction today. Increasing the tolerances would allow for greater freedom in devising solutions to the real operational problems that we have today. Maybe they could be solved without eliminating leaps, but by making them more regular so they are more likely to be supported, tested and working.


More information about the LEAPSECS mailing list