[LEAPSECS] Inside GNSS published an update of my CGSIC talk

[Poul-Henning Kamp]

--------
Warner Losh writes:

> So my first choice is
> always 'none, cope with shifting civil time on the scale of centuries' but
> my second choice is 'schedule for the long-term average and don't worry
> about going > 1s' .

The long-term average used to be "a leapsecond every 18 months -
give and take" and based on that I have previously also advocated
for the "announce leap-seconds at least 10 years in advance" idea.

We have now seen "patently geophysically impossible" negative
leapseconds become even money in slightly less than two years.

The nearest we get to a scientific explanation for that, is hand-waving
in the general direction of "climate change", even though not even
the most dire climate models move enough water around.

So do we in fact even have a "long-term average" to steer towards right now ?

I expect we would end up repeating the "Oops too little - Oops too much"
steering we saw back in the rubber-second years.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.