ietf
[Top] [All Lists]

Re: Last Call: Date and Time on the Internet: Timestamps to Proposed Standard

2001-11-11 23:10:03
The use of a UTC timestamp is promoted on the grounds
that a "local" time alternative may have a relationship to
UTC that is "dependent on the unknown or unknowable
actions of politicians or administrators."

However, the relationship between UTC and TAI
(monotonically evenly increasing Atomic Time) is
dependent on measurements collected by the
International Earth Rotation Service (IERS), which
result in the application of leap seconds to TAI
to obtain UTC.  In general, without access to an
up to date table of leap seconds, it is not possible
to determine the difference between two UTC times,
since one or more leap seconds may lie between
them.  And, while I grant you they may not be
politicians or administrators, the actions of the
astronomers at the IERS are equally unknowable
more than a few weeks in advance.

I would argue that UTC, fixated as it is on the bobbles
of a spinning ball of rock, is yet another display format,
and that timestamps should be a form of Atomic Time.
At the very least, it should be recognized that UTC is
as unpredictable in its own way as Daylight Savings Time.

cheers,
Don Craig
Omneon Video Networks



<Prev in Thread] Current Thread [Next in Thread>
  • Re: Last Call: Date and Time on the Internet: Timestamps to Proposed Standard, Donald Craig <=