Talk:Time synchronization in North America

NIST and USNL 'implementations'

UTC is not an algorithm; one does not 'implement' it. It is a standard timescale.

NIST provides a UTC time-signal, which they have (unhelpfully, IMO) branded as UTC(NIST). USNL has done the same thing. It seems to be claimed that UTC(NIST) is intended for USA civil time, and UTC(USNL) is intended for navigation.

But they are the same timescale! These are not two different 'implementations' of UTC - they are two measurements of the passage of time by two different sets of instruments. To the extent that they are not exactly the same, they are simply inaccurate (and all such measurements are intrinsically inaccurate, to some degree).

I want to eliminate the language in the lead that says that UTC(NIST) and UTC(USNL) are distinct standards. They are not; they are actually distinct time-signals. The standard is called 'UTC'. I want to eliminate the language in the lead that says they are standards.

This seems to me to be part and parcel of the persistent efforts of some to insist that GMT is a standard. I suspect it is motivated by jingoistic nationalism.

Actually, I think this article should be deleted; time-synchronisation in the North America is no different from time-synchronisation anywhere else in the world. You can use a telephone, you can use a time-signal, you can use the internet. It's the (hard) problem of keeping clocks in sync across geographical divides. There are way too many articles about time standards in wikipedia, and this is one of the worst ones. MrDemeanour (talk) 17:40, 27 September 2018 (UTC)

I'm inclined to agree that the article, in its present form, doesn't provide much of value that is specific to North America. This article attempts to give an overview of time synchronization methods and characterize the relative accuracy and expense. Perhaps the article should be improved so it does a better job of that, and broaden it to cover other parts of the world.

"Standard" can mean many things. In the world of measurements, it can mean an artifact which one compares to whatever is being measured to carry out the measurement, as in the case of placing an object to be measured on one pan of a balance and laboratory weights on the other pan. Usually, there is a hierarchy of standards, so the weights that are actually used on a daily basis are compared several times a year to a higher quality set stored in a locked cabinet, and once a year those are taken to the state office of weights and measures to be certified. The weights used at the state office of weights and measures were supplied by NIST, who (through some intermediate standards) compared the weights to the United States National Prototype Kilogram). Each and every artifact in this comparison chain is a standard.

It isn't much of a stretch to regard a time signal as a standard. The USNO Master Clock 2 is the primary standard of the USNO, which is used to derive other time standards, such as the atomic clocks at GPS ground stations, from which the atomic clocks on GPS satellites are set, from which the display of my recreation-grade GPS is set, from which my wrist watch is set. I then use my wrist watch as a working standard to measure the occurrence of events (being an EMT, that might be the time of a birth of a child, which will be recorded for posterity on the baby's birth certificate).

As explained at https://www.usno.navy.mil/USNO/time/master-clock/international-time-scales-and-the-b.i.p.m UTC, with no parenthetical designation after it, is a correction to the time kept by the various national time laboratories and isn't available until about 2 weeks after the time in question has passed. The only available official time signals are always derived from some real-time version of UTC, such as UTC(USNO) or UTC(NIST).

The way the article is currently worded, it implies that UTC(USNO) is somehow "unoffical" for civilians, and UTC(NIST) is somehow "unofficial" for the Department of Defense. I believe it would be difficult or impossible to find any authoritative source to back up this idea. My guess would be that if this civilian vs. defense dichotomy has any real meaning at all, it would be in how the two organizations would respond to requests to develop new timing services, or to refrain from discontinuing existing ones. For example, the latest NIST budget proposal would eliminate WWV and WWVB.[1] If the Navy, for example, were to ask to keep those running, NIST might insist the Navy should pay them, since serving the Navy isn't part of NIST's budget. Jc3s5h (talk) 18:30, 27 September 2018 (UTC)

@Jc3s5h "The only available official time signals are always derived from some real-time version of UTC, such as UTC(USNO) or UTC(NIST)"

Yes, these terms designate time-signals, not standards. The standard is UTC. I don't think any international body promotes as a standard UTC(USNO). If you choose to use it as your master clock for setting your watch, that's fine; but the standard you are relying on is UTC. UTC(USNO) is the proxy for UTC that you are using, because UTC itself is known exactly only retrospectively, a few months in arrears. But hey, USNO and NIST both have caesium clocks - drift over the coordination interval really isn't an issue. USNO and NIST keep UTC to within the margin of error of those clocks. They are not promulgating two different standards that are competing with UTC. They are each promulgating their own best estimate of UTC. MrDemeanour (talk) 21:00, 27 September 2018 (UTC)

MrDemeanour wrote "Yes, these terms designate time-signals, not standards. The standard is UTC." No. UTC(USNO) and UTC(NIST) designate time scales that only exist within the respective time laboratories; it is dubious to call them signals. Signals are typically understood to be things available to those outside the laboratory, such as WWV, WWVB, the signals emanating from GPS satellites, NTP transmissions, etc. Also, I reject your narrow definition of "standard". Time signals are standards. Jc3s5h (talk) 21:08, 27 September 2018 (UTC)

OK, @Jc3s5h, I note your explanation of 'standard', with example. I'm not convinced, but perhaps you simply know better than I do. MrDemeanour (talk) 02:14, 28 September 2018 (UTC)