[ntp:questions] Re: Make NTP timestamps leap-second-neutral (like GPS time)

Terje Mathisen terje.mathisen at hda.hydro.com
Wed Jan 7 18:24:40 UTC 2004


Danny Mayer wrote:

> hack at watson.ibm.com (Michel Hack) wrote in message news:<61706c80.0401062158.720939a7 at posting.google.com>...
> 
>>Proposal:
>> ---------
>>
>> Currently, NTP timestamps are defined to represent UTC based on a sliding
>> epoch such that UTC can be derived from seconds-since-epoch using simple
>> Gregorian conversion (where each day has exactly 86400 seconds).
>>
>> I propose to redefine NTP to be tied to TAI (International Atomic Time)
>> but referenced to 2000, so that:  NTP(2000) = UTC(2000) = TAI-32 = GPS-13
>> and from now on:  NTP = TAI-32 = GPS-13
>>
>> This is a good time to propose such a change.  There have been no leap
>> seconds since July 1999.  I wish I had done so sooner (this here is based
>> on an internal memo I wrote a year ago).
>>
> You need to turn this into an RFC proposal and then argue it out with
> Dave Mills who is the real expert in this area. Interoperability between
> old and new protocols is essential for NTP to continue to work. In addition

This is crucial part, I do like your suggestion that the TAI(NTP)-UTC 
offset should default to zero, unfortunately I'm afraid that we might 
not have any free room in the packets. :-(

My suggestion would be to use some previously unused combination of the 
leap bits, and either simply forget about broadcast (at least for now), 
or send out two different versions of the broadcast packets, with the 
new format setup in such a way that they will look to be invalid to 
current clients.

> if a proposal is put out I'm sure there are other features that people
> might want to add to such a change in the protocol. 

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