[ntpwg] draft minutes from Vancouver NTP WG meeting

[David L. Mills]

Danny,

At the risk of being briefly off-topic, a few years ago I and students 
were working the problem that, given the ephemeris of a spacecraft with 
synchronized clock and another spacecraft with given ephemeris but 
unknown clock, how could that clock be synchronized? Our solution was an 
iterated algorithm that first deterimed the lightime relative to an 
assumed unknown clock, then determined the position at that time, 
compared to the calculated lighttime and iterated. Generally, only a few 
iterations were necessary to converge. This did not consider gravity and 
time dilation effects.

Dave

Danny Mayer wrote:

> Rob Seaman wrote:
>
>> On Dec 24, 2007, at 8:39 PM, Danny Mayer wrote:
>>
>>> I forgot to comment on this one. The packet follows the geodesic
>>> which is a 4-dimensional path so you cannot just locate the 3-d
>>> coordinates of each endpoint and calculate the distance between them.
>>
>>
>> Even if the Newtonian approximation doesn't apply, there is a physical
>> model that can be run through the Runge-Kutta mill to deliver an
>> estimate of the light travel time to whatever precision needed.
>>
>
> I admit to not having looked at RK for quite a few years!
>
>> More fundamental is that even Einstein had to punt on the central
>> conceit of NTP. It may seem logical that the timestamp "midway"
>> between event A and event B is (A+B)/2, but this is the "standard
>> synchrony" of relativity - that is, an inductive assumption, not a
>> deductive conclusion.
>>
>
> Agreed. I'm not really sure what the midpoint really means in such a
> case never mind what the timestamp would be for it. That boggles the mind.
>
> Danny
>
>> Rob Seaman
>> National Optical Astronomy Observatory
>> _______________________________________________
>> ntpwg mailing list
>> ntpwg at lists.ntp.org
>> https://lists.ntp.org/mailman/listinfo/ntpwg
>>
>
> _______________________________________________
> ntpwg mailing list
> ntpwg at lists.ntp.org
> https://lists.ntp.org/mailman/listinfo/ntpwg