[LEAPSECS] op eds on leaps

Steffen Nurpmeso steffen at sdaoden.eu
Sat Sep 10 19:21:00 EDT 2016


Steve Allen <sla at ucolick.org> wrote:
 |An op-ed in favor of the leap second from last Sunday
 |https://beta.welt.de/print/wams/kultur/article157946393/Wie-die-Zeit-vergeh\
 |t.html
 |
 |There have been op-eds against the leap second, but someone please
 |remind me if any of those have been poetic.

Journalism is hard
Latte Macchiato is expensive luxury
so here's the why
the press is free
to live a life of fee

The article starts with "Lesedauer: 6 Minuten" (time to read: six
minutes).  I wonder if that indicates a general problem.
Amazing that mankind gets weak and sentimental when they have
babies, especially daughters: Each second is precious;  Humans
should care, humanly;  You can die in a second, or fall in love.
As if you could get a system going with plain soppiness, and keep
it ticking the right way.  Especially now that alternative home
planets come into sight one should really look ahead for the great
future that is to be expected.  New watches can also adhere to
more sophisticated rules, like mars time, for example.

Just like the head of the Tibetians, the Dalai Lama, said in the
interview for his 50th anniversary in Indian exile: it is all
there, you just have to see it.  Maybe leap seconds should simply
be celebrated, for example with a vegetarian barbecue at
Stonehenge.

It is a bit unfortunate that the article doesn't find -- or
haven't had -- the time to describe that it is a problem of
definitions rather than a problem of time by itself.  It mentions
Qantas and the Linux kernel bug, but only mentions a complete
switch over to TAI as an alternative.  My opinion on this is that
systems should gain a CLOCK_TAI, and that protocols should start
providing this, if they don't do already.  Now, i know the
proposal [1] of Markus Kuhn, and that is almost two decades old.

  [1] https://www.cl.cam.ac.uk/~mgk25/posix-clocks.html

Eventually we get there, that so much damage has been done because
of the slowness of democratic plus systems is a real problem.
E.g., just think about the immense amount of oil that has been
wasted because sequential cylinder deactivation, overrun cut-off,
combined turbo and compressor, downsizing etc.  hasn't been
enforced politically over 25 years ago!  In, e.g., Japan, 600ccm
engines with about 70 hp are fostered since about
35 years so (or others are disadvantaged), and see how the
Japanese car industry broke down due to this decision.

And then fuel cells with wheel hub engines and a sandwich vehicle
floor that hosts the tank, that also was possible back then.  (The
tanks were not capable of storing the hydrogen for a long time,
iirc, but for one that is just a problem that could have been
worked around, and it has been solved in the meanwhile.) So that
would mean hundreds and hundreds of millions of these middle age
combustion engines are in the wild, and will stay for many more
decades to come.  Instead of just improving a few dozen
refineries.  A different story, but the same s..t, or never made,
decisions.

So i would still hope for some nice NTP which distributes all that
easily, a standardized CLOCK_TAI that users which need it can
easily access, and a normal UTC clock for normal human behaviour,
which eventually is adjusted to adhere to our home planet.  You
need a table of past adjustments, but that is all wrong anyway.
E.g., who knows whether a file time of a tar archive member from
the early ninetees is really correct.  You wouldn't know even if
it were TAI, most likely.  Having _TAI is only useful with a set
of routines which can be used to create datetimes etc.: isn't it
a pity that even ISO C 2011 didn't follow a bit of the other
proposal [2] from almost twenty years ago?

  [2] http://www.cl.cam.ac.uk/~mgk25/time/c/

It is plain that i was joking once i saw, just one or two years
ago, an ultra-right Japanese politician holding a glass of diesel
(murky) in one, and a glass of petrol (clear) in the other hand.

--steffen


More information about the LEAPSECS mailing list