[LEAPSECS] the big artillery

[Dennis Ferguson]

On 6 Nov, 2014, at 20:16 , Sanjeev Gupta <ghane0 at gmail.com> wrote:
> Anyone who wishes to believe 60 secs always to a minute can continue to do so.  Anyone who needs the extra accuracy (time-nuts, astronomers, pedants, old men like me) will learn the fact that a minute is not always 60 secs.  Anyone who is unwilling to learn this should not be pointing the scope on Mt Palomar anyway.

So you'd be okay with a hectosecond that wasn't 100 seconds?

The problem with the UTC minute is that it is a polysemous use
of the word "minute".  In most contexts the word still means
precisely 60 seconds, as it always did.  An SI-compatible minute
is always 60 seconds and 1/60 of something else.  The constant
86400 that keeps appearing in equations is inexplicable without
that definition.  A Julian date with many digits to the right of
the decimal point can have those digits unambiguously expressed
as an hh:mm:ss form for pretty much any timescale except UTC, where
the meaning of the digits becomes a bit fuzzy.  The UTC definition
of "minute" isn't more accurate, it is just a different definition
of the same word.

I realize, however, that an argument about polysemy isn't a
particularly strong one.  Avoiding polysemy is a benefit which
needs to be balanced against its cost, and making up new words
to describe the new definition of the units of UTC (and civil
time) would have had a high cost compared to just redefining the
traditional units to be different in that context.  I would just
point out that the argument that a new definition of our civil
timescale to retain SI seconds but eliminate leap seconds should
have a new name to avoid making "Universal Time" polysemous is
one that seems very similar since there would be a price to pay
for not continuing to call the redefined universal timescale UTC.
I'm not sure what I think about that trade off, but I do find it
a little inconsistent for a person to strenuously object to
that polysemy while simultaneously arguing that the older one
is perfectly acceptable.

You could learn the fact that UTC is no longer as close to mean
solar time as it once was if you needed the extra accuracy.  Most
people probably won't need to bother.

Dennis Ferguson