[ntpwg] Further to the timestamping issue

[David L. Mills]

Rob,

You imply a burning desire to predict leap events possibly years in 
advance. From the IERS bulletins I gather announcing leap seconds is 
somewhat of a black art and the announcements are only a few months in 
advance, depending on the DUT1 crossover. Also, the IERS can announce 
multiple bulletins even now should they choose; however, the time 
dissemination community has no means at present to announce more than 
one at a time. In fact, it is possible to announce more than one 
occasion in the NIST leapseconds file. Should that possibility arise, I 
would be delighted to have NTP properly respond to them.

Dave

Rob Seaman wrote:

> TSG wrote:
>
>> Do you have any idea of the cost of implementing a log management
>> practice that will address multiple Leap Second's being added???
>
>
> No, and neither do you.
>
> You are also jousting with a windmill. Nothing would stop Gambis from
> announcing a leap second for December 2011 at the same time as a leap
> second for June 2009. (Dates chosen randomly.) Each would be handled
> in sequence whether or not NTP is aware of both at the same time.
> This is precisely a feature of UTC that has been requested. The
> question is how best would NTP implement it.
>
>> Considering also that Leap Second's are never added such that there
>> would be more than one per month added,
>
>
> Nice argument against leap hours :-)
>
>> the accidental adding of a leap second (assuming NTP automatically
>> tracks them) could cause billions or trillions of dollars of damage
>
>
> Somebody who places auditing above all other requirements might want
> to actually characterize the cost/schedule/performance benefits and
> risks of the various alternatives :-)
>
> Rob
>
> _______________________________________________
> ntpwg mailing list
> ntpwg at lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/ntpwg