[LEAPSECS] Definition of Standard time - Brooks Harris

Brooks Harris brooks at edlmax.com
Sun Feb 16 14:14:20 EST 2014

On 2014-02-16 10:39 AM, Warner Losh wrote:

> On Feb 16, 2014, at 11:20 AM, Brooks Harris wrote:

>> Only a comprehensive plan which aims to fix the obvious and well known problems is going to head off the "kill Leap Seconds" movement.

> I think the momentum and general conservatism of the powers that be will do more to kill the plan than any other comprehensive plan. the status quo is powerful enough and works well enough that people are unwilling to risk a change.

Maybe. But we're still left with vague and difficult implementations
that have known problems.


>> Or will we just roll over and watch 4500 years of timekeeping tradition evaporate?

> The kill the leap second stuff doesn't kill 4500 years of timekeeping tradition. Leap seconds broke with tradition my making minutes longer than 60s. It accepted there's an error between the time in London and the time we coordinate on and that's OK. The leap second moved one step away from the sun by averaging out the noise into discrete steps.

Right. Well I see it slightly differently. For centuries timekeeping
sought to align the observed positions to some "absolute time
reference", long recognizing the days length didn't divide into a year.
Pope Gregory (partly) fixed that. Then when atomic clocks could keep
very accurate time and astronomical observations were extremely precise
we were finally in a position to quantify the difference between
absolute time and observed rotational position. Integral seconds (Leap
Seconds) was deemed accurate enough to solve the problem for general
purposes. So I see UTC as the latest and greatest version of 4500 years
of timekeeping tradition.


> Moving to atomic time doesn't undo 4500 years of timekeeping tradition. In fact, it restores the tradition of all minutes being the same length.

I think it abandons the long sought goal of aligning civil time to the
position of the Sun. The "tradition of all minutes being the same
length" was the best anybody could do. Everyone knew it wasn't exactly
right. UTC solved that.


More information about the LEAPSECS mailing list