[LEAPSECS] Focus in the debate, alternative proposal

[Zefram]

Nero Imhard wrote:

>So here's a proposal for making UTC more workable in the long run: let's

>alternate positive and negative leap seconds!

...

>Somewhow I can't imagine mine is a novel idea.


It's not.  In the message at
<http://www.ucolick.org/~sla/navyls/0197.html>, Ed Davies proposed
leaping every month subject to maintaining |UT1-UTC| < 0.9 s, thus
hitting about 80% of months given current scheduling lead time.  About two
months ago, Tom van Baak proposed leaping *every* month, at the expense
of a moderate increase of the |UT1-UTC| bound.  It's probably come up
on other occasions too.

The basic concept, of having leap seconds at
every opportunity, achieves its goal of providing plenty of real-world
opportunities to test leap-second-sensitive systems.
The concept also fits nicely with Rob Seaman's proposal
at <http://iraf.noao.edu/~seaman/leap/> to explicitly abandon the
preference for leaping in particular months, and schedule strictly to
optimise |UT1-UTC|.

However, it doesn't fundamentally make leap seconds any easier to
handle.  It *only* makes it easier to test whether the handling has been
implemented correctly.

-zefram