On Jan 8, 2006, at 6:41 PM, Tom Van Baak wrote:
am not sure I like the idea that eventually my car, traffic lights,
airlines, television, and my thermostat will have to be reliably
tied to the IERS in order to function properly.
This is a general issue with the increasingly tight coupling between
any number of networked systems. Certainly not unique to
timekeeping. On the other hand, rarely should any system (certainly
not a safety critical application) depend on a clock that must remain
tightly slaved to *any* external signal. Consider what evolutionary
path technology would have to take such that significant numbers of
traffic lights and thermostats would ever require direct contact with
the IERS. It's easy to speculate about pathological engineering
practices - A requires access to B, immediately and always, perfect
and inviable. But real designs result from real requirements.
What if some killer app 40 years hence requires 100 ms or 1 ms time
accuracy.
Don't think "accuracy" is the word you want. The short answer to
your question is that all the time wonks would celebrate their
newfound employability.
Do we still want UTC leap seconds when it will infect ten billion
devices?
Perhaps you meant "affect"? :-)
What we want and what we need are two different things. And as is
currently true, those devices aren't required to use UTC unless they
need UTC.
the current scheme does not scale well into the future;
No, it does not - but there are two caveats:
1) the current scheme (also known as an "international standard") is
good for several hundred years
2) no other scheme scales better
either a technological future (way too many devices affected by
unscheduled time steps) or an astronomical future (way too many
leap seconds a year).
As I pointed out close to five years ago, the ultimate long term
remediation will likely involve redefining the length of the second:
http://iraf.noao.edu/~seaman/leap
Nothing over the years of intervening discussions has given me cause
to change my opinion.
Rob
