[LEAPSECS] "China move could call time on GMT"

Ian Batten igb at batten.eu.org
Mon Jan 9 07:43:22 EST 2012

On 9 Jan 2012, at 09:30, Clive D.W. Feather wrote:


> If paying the price of an additional one-hour shift in 12 generations time

> is the price to pay for getting rid of leap seconds, I'd happily pay it in

> a heartbeat.

Quite. The invocation of "won't somebody think our of great^8 grand children's watches!" is disingenuous. It's raised not because the people raising it are moral paragons with a concern for ten generations' time (because if they were, they'd been worrying about more immediate issues like energy security); it's raised because the other argument they have lack sufficient force.

It is a hell of a stretch to claim that re-aligning time zones once every ten generation is unreasonable. How many countries are there whose time zone has been stable for two generations? How many countries are there which have even had national time for more than five generations? Almost all technological systems are configurable for time zone (*); changes in DST rules happen with almost no impact and the DST changes themselves with, for practical purposes, no impact. Why is this any different?


(*) For me, everything except my elderly bedside alarm clock, similar to http://goo.gl/rwPl8, a very early Junghans MSF receiver from the late 1980s, and a couple of cheap MSF analogue clocks I have around the house. But as MSF is defined to broadcast UK civil time, not UTC, they should continue to work quite happily in the event of the UK changing time zone, providing that my great^8 grand-children can get AA batteries and that the MSF transmitter is still functioning, both of which seem to be even less likely than the clocks still being in working order. Although, as a footnote to a footnote, I can imagine that if the UK _does_ transition to CET that the debate on what MSF should broadcast could become quite heated, because I suspect a lot of equipment relies on it being UTC --- my memory of looking at the NTP driver for an MSF receiver was that they didn't provide for any offsets other than subtracting an hour if the DST flag was set. But that's a UK-specific problem, caused by the slippery assumptions about civil time = GMT+DST = UT1+DST = UT+DST, for some value of =.

Most similar products sold today contain the means to set the timezone, because that way they can be sold as a single product in multiple territories. Such products sold in the US have to be switchable for timezone, because there isn't a distinct WWV transmitter for each of the timezones. In Europe, there will be products that rely correctly on MSF being UK civil time, or less correctly on DCF77 being civil time for most of the remainder of the EU, but they are becoming less common and are not, in any event, the sort of things that Danny Hillis and Brian Eno would regard as up to the mark for The Long Now.

More information about the LEAPSECS mailing list