----- Original Message ----- From: "Markus Kuhn" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Thursday, June 05, 2003 2:33 AM Subject: Re: [LEAPSECS] pedagogically barren?
> "Seeds, Glen" wrote on 2003-06-04 15:00 UTC: > > It's also true that changing to SI units for weight and volume is a lot more > > technically tractable than for length. Public opposition would still be a > > big barrier, though. > > That's what the UK have done. The imperial units of weight and volume > are not legally recognized any more in Britain (only pints are still > permitted for drinks volume), whereas inch/yard/mile continue to be > legally recognized for length and speed. > > To bring the topic closer back to the scope of this mailing list: > > One international standard related to time keeping that I would like to > advertise for is the international standard numeric date and time > notation (ISO 8601), i.e. 2002-08-15 and 14:14:57. > > Whereas both the modern 23:59:59 and the old fashioned 11:59:59 p.m. are > equally widely used in Britain, the modern notation seems to be mostly unknown > in the US outside the military and scientific communities (and the US > military seem to drop the colon as in "1800" and say strange things like > "eighteen hundred hours" instead of "eighteen o'clock"). The uniform > modern 00:00 ... 23:59 notation is now commonly used in Britain for > almost any publically displayed timetable (bus, trains, cinemas, > airports, etc.), and on the Continent they haven't used anything else to > write times for many decades. > A propos of both the topic and the discussion of notation, I've observed that in the U.S., hospitals (where 24-hour notation, or "military time" as civilians inevitably call it) are one of the few businesses where wall clocks are nearly always set to the correct time (within+/- one minute, and often within +/- 10 seconds, as checked against my WWVB watch). The correct time on birth and death certificates is important, but I was not aware of how important until I saw a posting from Prof. David Mills on comp.protocols.time.ntp in which he said that UT1 (not UTC) is the legal standard for death certificates. My reaction was that this is fascinating if true, but even if it is (I couldn't find any documentation of this), I would have to wonder how DUT1 becomes an issue if the tolerance is (as currently) less than one second? Does anyone have any firshand knowledge of forensic medical issues related to DUT1? The implications of removing the 0.9s limit are clear if Prof. Mills is correct, but my impression was that time-of-day need only be precise to within one minute for birth and death certificates. Brian Garrett