A lurker surfaces

From: Jim Palfreyman <jim2_at_tasmail.com>
Date: Sun, 31 Dec 2006 12:41:46 +1100 (EST)

Hi Folks,

I've been lurking for a while and I'd like to add my thoughts. Whilst I'm a professional software engineer (and was very active during the y2k "bug") I also am an amateur astronomer, and one could call me a time nut - my wife certainly does. I have built a GPS disciplined oscillator and clock and I even happen to own one of the old speaking clocks that were used in Australia for 15 years. So what I'm saying is that I have a foot in all camps.

With my software engineer hat on, having a time scale with discontinuities in it is a pain. For example you never schedule jobs to start on or between 2am or 3am for daylight savings reasons. You never schedule jobs on Dec 31 or Jun 30 for leap second reasons. Even if we make UTC non-discontinuous, the daylight saving problem is still there and always will be. Even with leap seconds removed I could not schedule a job to start exactly 5000 hours from now, because the government may change the daylight savings regulations in two months time. Keeping tables of past daylight savings rules around the world is a pain, but we can work work with it. So we can also keep tables of past leap seconds, we can work with that too. The basic premise of being able to know these changes in advance would be lovely - but (especially with daylight savings) it won't happen.

Still with this software hat on, when dealing with these discontinuities, one thing I can say is that it is important to deal with them regularly. The y2k "bug" was expensive because we hadn't done it before. We deal with daylight savings twice a year, and leap seconds on a regular basis. Once a month they literally cut the power to the main telecommunications building in my city to make sure the backup generator works. So the idea of a leap hour would be particularly "expensive" and one to avoid because it won't happen for a long time.

With my time hat on, having time that is discontinuous pains me. It doesn't make sense "in my heart". But at least these discontinuities are in whole seconds. The earlier concept of rubber seconds gives me the creeps and I'm glad I wasn't old enough to know about it then! It is heart warming to know we have TAI or GPS time and that these are nice and steady. At least they are there if I need them.

Now my astronomy hat comes on. The fact that the difference between UTC and UT1 is very close to zero at the moment and when I watched a recent lunar occultation and it actually was "on time". That was nice. The fact that I can watch these and even see how UT1 is going - I just love it. As an amateur astronomer I could handle these diverging no worries. But only to a point. 28 hours? No way. 5 seconds? No worries. Somewhere in between? Maybe.

So in summary, what I think is important is:

Any discontinuities must be regularly done. So they are part of all computer systems and are tested and used all the time. Don't let them build for a decade - that is bad bad news.

Keep the discontinuities in second multiples. Once we start playing with that basic SI unit (e.g. leap milliseconds) tracking things could become a real pain.

And with all my hats on at the same time what would I suggest? Keep it the way it is. Even my 20 year old speaking clock handles leap seconds with no worries!

And finally for a bit of sarcastic humour in this argument. The "Americans" are suggesting the abandonment of leap seconds to make things easier. You'd at least think they'd get their own house in order and start using the metric system in their country first. That would have far bigger impact and save much more money. :-)


Jim Palfreyman
Tasmania, Australia

Received on Sun Dec 31 2006 - 00:22:41 PST

This archive was generated by hypermail 2.3.0 : Sat Sep 04 2010 - 09:44:55 PDT