Is there a good way of getting the system time as the atomic time (TAI)
rather than UTC? Or even a bad way? Atomic time is essentially UTC without
leap seconds.
Are leap seconds really just compiled into NT? Can they be inserted without
a service pack?
Thanks,
Jan Mikkelsen
xxxxx@transactionsite.com
At 11:48 PM 6/26/00 +1000, you wrote:
Is there a good way of getting the system time as the atomic time (TAI)
rather than UTC? Or even a bad way? Atomic time is essentially UTC without
leap seconds.Are leap seconds really just compiled into NT? Can they be inserted without
a service pack?
My understanding is NT actually uses something closer to TAI time than UTC
time, except I believe the time service sets the current TAI time to
current UTC time. It doesn’t handle leap seconds at all. So the reported
UTC time glosses over leap second events, and calculation on time don’t
account for leap seconds.
I had a big discussion here about the lack of correctness of NT’s time
handing a while back. Most of the reactions here were I had way too much
free time if I could worry about such things. I though there were a number
for real applications where you need the system to really track UTC time,
and just resetting the clock daily was not such a great option. A lot of
programmers didn’t seem to want to face the reality that some UTC minutes
will have 61 seconds, and you can’t forecast when these minutes will
happen. You can look in a table of when they happen in the past, and have
some warning about upcoming leap seconds. The whole concept of time
calculation on UTC time values is just amazingly ugly.
Perhaps we should teach children (or even adults) about the reality of time
measurement. Personally, I learned about the magic of base 2 counting at an
early age (like 5th grade). My teachers and schoolmates were kind of
annoyed that my belief was fingers actually were great for counting up to
1023, instead of 10 (guess my fingers were two orders of magnitude better
than many fingers, just because of how I perceived them).