[time-nuts] Re: UTC - A Cautionary Tale

Chris O'Byrne obyrne at iol.ie
Tue Jul 19 08:48:34 EDT 2005


Since you have asked me to extrapolate the leap second question to leap
days, allow me to do the same. Would you settle for a system whereby you
had to wait on some scientists' verdict before finding out if your next
birthday was going to be on a Sunday or on a Monday?

> It wasn't clear you had based your argument on a bunch of unnatural,
> unstated qualifications. It's always possible to argue a manufactured
> exception, but it's the generalized case which applies most broadly.

OK - let's try this one out for size. This is an actual, real-life
situation that could quite concievably happen next March 29 in the
middle of the Lybian desert.

There is an eclipse-chaser who has their cameras all set up to capture
the eclipse on film. Their cameras are on timers, so that they will
automatically capture the event while they themselves just sit back and
watch. They have a mobile phone with them, and they have synchronized
its clock against a GPS system that is displaying UTC. They certainly do
not have access to the Internet, nor to any other reference systems. On
that mobile phone, they have a wonderful piece of software that they
downloaded off the Internet months before, which is giving them the
information they need to synchronize their automatic camera control

Unfortunately, the version they downloaded is the penultimate version
that was released. They don't know this - they have better things to be
doing than constantly checking that every scrap of software they have is
up-to-date. With the result that some of the most interesting of their
pictures were ruined, simply because their software was just one version
out of date! They didn't know that the time they keep on their watch -
the time displayed on their GPS system - isn't a regular time! (By
regular, I mean each minute is 60 seconds, each hour is 60 minutes - you
get the idea).

(In case you don't remember, the significance of the penultimate version
is that I released a version of the software hours before the most
recent leap second was announced).

Now, can you come up with a scenario extolling the virtues to the
average person of leap seconds? Or a scenario in which an
ever-so-slightly variable second being used by a member of the public
proves disasterous?


More information about the time-nuts mailing list