[LEAPSECS] Civil timekeeping before 1 January 1972

[Brooks Harris]

On 2015-03-08 01:09 PM, Zefram wrote:
> Brooks Harris wrote:
>> It seems to me NTP and POSIX as well as other timescales concerned
>> with "civil time", are essentially disconnected from "reality",
>> expressing "idealized" measurement scales.
> That's very much what they're not.  TT is idealised, and TAI less so.

I meant "idealized" in the sense of convenient counting methods. They 
are not accurate in any respect before 1972. If it were up to me I'd 
just abolish legacy timescales and force everybody to use UTC. That's 
not gonna happen, we obviously must accommodate existing systems.

>
>>                                                                   I
>> think none of the "civil" timescales are counting in UT - they are
>> measured in SI Seconds, even when prolpetic to 1972.
> NTP doesn't deal with pre-1972 time at all.  (No, the nominal epoch
> doesn't count.)
Right. You can, if you like, generate a date and time from a given 
seconds offset from NTPs' prime epoch era 0. Its inaccuracy with respect 
to some other pre-1972 timescale you wanted to compare it to could be 
calculated. That might be useful for some purpose, but its not the 
objective.

>
> POSIX time_t notionally can represent pre-1972 times, but in practice no
> Unix system of that era was synchronised to UTC.
We're not caring about what systems did then. We're caring about what 
they do now and in the future.

> Any use of time_t for
> precise pre-1972 time is heavily retrospective, and the interpretation
> is more governed by the application than by the POSIX standard.
Exactly.
> Wild
> pre-1972 Unix time_t values heavily predate the POSIX standard, and their
> interpretation has little to do with UTC.  They are understood to be vague
> UT with usually very poor synchronisation (via the operator's wristwatch).
Right. So what?
>
> Actual civil time, as used in the real world, was rarely precise pre-1972,
> certainly not much to do with SI seconds.  Historically civil time has,
> over the long term and with large short-term excursions, been synchronised
> to the solar day, so effectively to UT.

And that's what we'd hope to make more clear, explicit, and reliable in 
the future. The past is past. What can we do now?

>
>> I understand the proper SI second sprung into existence as of 1972.
> You understand incorrectly.  As with the existence of TAI (which I
> discussed in my previous message), the development of the SI second
> happened in stages.  We can start in 1948, with the definition of
> Ephemeris Time, which is the first time we got a definition for the
> second that was not derived from the solar day.  The ephemeris second is
> instead defined in terms of the Earth's orbit around the Sun.  ET was
> adopted by the IAU in 1952.  In 1958 the caesium second was defined,
> the definition being chosen to match the value of the ephemeris second.
> SI was established in 1960, and at that time adopted the ephemeris second
> as the SI second.  In 1967 SI was revised to adopt the caesium second
> as the SI second, and in 1997 the definition was refined to account for
> blackbody radiation.
Yes yes I know.

>
> So the value of the SI second was established in 1948; the present
> definition originates in 1958 but its full present form only goes back to
> 1997; SI (and so the concept of "SI second") originated in 1960; and the
> SI second acquired its present basic definition in 1967.  Any of these
> dates could be taken as the inception of the SI second, but the only one
> that could really be described as the SI second "springing into existence"
> is the establishment of SI in 1960.  Nothing relevant happened in 1972.
Yes, I know. What happened in 1972 was the magic moment when TAI - 10s = 
UTC. All the complex, if interesting, history of the development of TAI 
and UTC is over the damn. Discussion of it is all down in the weeds 
until it becomes mostly locked-in in 1972.

I was responding to Steve's "UT" discussion and his comment "Before 1972 
civil time was not SI seconds".

-Brooks

>
> -zefram
> _______________________________________________
> LEAPSECS mailing list
> LEAPSECS at leapsecond.com
> https://pairlist6.pair.net/mailman/listinfo/leapsecs
>
>