[time-nuts] GPS/UTC time

[Bill Hawkins]

-----Original Message-----
From: Brek Martin
Sent: Friday, July 03, 2015 11:46 PM

>If they skip a second UTC, surely the GPS time isn't run incorrectly
forever.

Correct for who or what?

There are two kinds of time (using some simplifying assumptions):
machine and social. Machine time is monotonically increasing forever,
because machines don't care about things like a noon break for lunch.
Social time is for the vast majority of people who do want noon and
midnight to correspond to their idea of time relative to the planet.

Two things change the relationship between machine and social time: leap
seconds and daylight savings time. Both can use a relatively simple
translation application to convert machine time to human social time
when machines display time to people. We certainly noticed it when
calendars had the wrong number of days in a year.

So GPS time is correct for machines and UTC time is correct for humans.

>ps. am I replying the way I'm supposed to with a new email and subject?

Yes. Thank you. The archives will be better for it.

Kind regards,
Bill Hawkins

P.S. It's true that the time difference changes very little (ignoring
DST, which cancels out) per year. Considering that there's a 90% chance
that humans will be extinct in 100 years, maybe it doesn't matter.