[ntp:questions] Re: GMT: UT1 vs UTC

David L. Mills mills at udel.edu
Sat Nov 1 02:21:40 UTC 2003


Take a look at the DUT1 format in the WWV/H/B timecode. Only three bits
plus signbits, so at least those stations tank out over +-0.7 s. A leap
has to be programmed before the DUT1 gets to 0.7 s.

Terje Mathisen wrote:
> Michael Shields wrote:
> > In article <bnt9v3$vd5$1 at osl016lin.hda.hydro.com>,
> > Terje Mathisen <terje.mathisen at hda.hydro.com> wrote:
> >
> >>With hindsight, which almost by definition is hard to get, it would
> >>have been better to delay the last leap second by 6-18 months, that
> >>way the current gap wouldn't have exposed that Oncore firmware bug.
> >
> >
> > Although the Oncore is important, I don't think it would be a good
> > precedent to start adjusting international timescales to work around
> > bugs in one vendor's product.
> Sorry, I didn't want to imply that!
> As noted on one of those pages, Leap seconds used to be added
> preemptively, i.e. quite a while before the 0.5 second offset point,
> instead of closer to the final 0.9 second limit.
> The last leap second we had was added 6-12 months too early, if the goal
> had been to minimize the UTC-UT1 at all times.
> Terje
> --
> - <Terje.Mathisen at hda.hydro.com>
> "almost all programming can be viewed as an exercise in caching"

More information about the questions mailing list