[LEAPSECS] Caveat emptor!

[Ian Batten]

On 12 Apr 2011, at 00:48, Tony Finch wrote:


> On 11 Apr 2011, at 20:19, Gerard Ashton <ashtongj at comcast.net> wrote:

> 

>> If leap seconds are abolished, and Britain decides to maintain an approximation

>> to mean time 0 degrees longitude, +/- a fraction of a second (modulo time zones, as currently defined),

>> over time the difference in timekeeping systems will become clearly perceptible and clearly

>> distinguishable to people with no special equipment beyond a carefully set wrist watch.

> 

> But note that Britain does not currently have the capability to maintain GMT since the relevant department was closed decades ago and the institution that housed it was abolished. If the international consensus is to change UTC then UK civil time will have to follow.


I think it's messier than that.  De facto UK time is UTC, as we've discussed, but the only source of de jure legal time now is MSF: it broadcasts UTC(NPL), but DUT1 is available in the payload (and that's (UT1-UTC), not (GMT-UTC), because as you say GMT is unmaintained).  So of itself, the non-maintenance of GMT hasn't affected anyone, as the (UTC, DUT1) pair  analogue    If UTC were unhitched from UT1, the MSF format couldn't carry |DUT1|>0.9, but let's imagine that was resolved somehow for the small community of users to whom it matters: a DUT1-like quantity would be available.   So if there were applications for which something very close to GMT were needed, it would be no less available in the future than it is today.

However, a lot of UTC clocks (either NTP or MSF) have been sold as "UK Legal Time" clocks, for trading and banking systems, and so long as DUT1 is small that doesn't really matter: it's hard to imagine a real legal action (as opposed to some geek's fantasy) revolving around quantities less than a second.  And for those scenarios where it might matter, the solution is easy.  The Interpretation Act 1978 S.9 has nothing to say about contracts, so case law, "reasonable man" tests and so on would resolve the first action, and thereafter contracts would add "times in this contract shall be assumed to be UTC" to the definitions section.  Wise lawyers might add that clause anyway.  Unless someone can produce something other than the Interpretation Act, "uk civil time" is only GMT insofar as it affects the interpretation of primary legislation which does not specify another timescale, and even then (the thing engineers do badly when assessing UK law) case law is almost certain to override it.   

The NPL/BIS document someone posted to this list a few months ago included a comment about the political difficulties of formally changing GMT.  They're right:  an unstable coalition government with a profoundly Euro-sceptic wing (this would smack of "Brussels" and "gold-plating", even though it's obviously neither) and a desire to remove regulatory red-tape is hardly going to pass legislation which on the face of it affects everyone with a clock, even though in fact it makes everyone's lives easier (UTC is ubiquitous, processing DUT1 isn't).   So in the grand British tradition of compromise and fudge, UK Legal Time would converge on whatever was most easily available, without any primary legislation being needed.

ian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://six.pairlist.net/pipermail/leapsecs/attachments/20110412/fc76a860/attachment.htm>