«
»

He Who Has Two Clocks is Never Sure

January 8, 2013, 1:20 pm

I had no idea there were so many time standards

We already have way too many time standards, including:

  • TAI, time based on an atomic clock, which ignores all motion of the Earth
  • UT0 and UT1, time based on precise measurement of the Earth’s rotation
  • GPS, the time standard used by GPS satellites
  • UTC, the standard used in computing, which is like TAI but with leap seconds to keep it in sync with Earth
  • TDT, TBT, TCB, and TCG, which are all even worse

This leads to all kinds of little headaches, particularly for programmers. For example, the clock in your smartphone’s GPS is 16 seconds out of sync with the phone’s system clock. This is because the system clock uses Coordinated Universal Time (which has leap seconds), but GPS time doesn’t. They were in sync in January of 1980 and probably never will be again.

Tags: , , , , , , ,
Category: Science

4 Comments

  1. HenryBowman419:

    Some of these time standards are more-or-less the same: TAI and GPS time are basically the same, as GPS time is based on atomic time; in fact there are very expensive rubidium clocks on the satellites. The other times are, in effect, application-specific, and, of course most people could care less.

    I'm not sure that phones actually use GPS time - I think they use GPS time adjusted to UTC. Most Windows computers (don't know about Linux or Apple, but they are likely the same) adjust their time periodically using various UTC time servers. My own observation is that my phone's time is basically the same as my computer's time (within 1-2 sec). Try it on your to see if that is your experience.

    January 8, 2013, 1:56 pm

  2. Earl Wertheimer:

    I work in EDI (Electronic Data Interchange). We have an expression: "The great thing about standards is that there are so many to pick from".

    January 8, 2013, 2:42 pm

  3. john mcginnis:

    HenryBowman419,
    Unix/Linux. Depends on deployment. If using internal clock there are two epochs to consider. Differnet from those used by windows. If device is using the NTP service then it is the average of the time server sources provided by either the ISP or internal network clocks, generally GPS.
    Such precision! /snark

    January 8, 2013, 6:08 pm

  4. mesocyclone:

    I suspect that most or all consumer devices using time from GPS adjust it to UTC. It would be pretty dumb not to. I just compared three sources and they were all within 2 seconds (why not the same, beats me):

    For the following, the WWV wrist watch is the standard:
    +0 sec: "Atomic" wrist watch - sets time daily to UTC via WWV
    +0 Sec: NTP (Mac mini toolbar clock)
    -1 Sec: Android Phone on Verizon
    -1 Sec: Android Phone GPS time display

    I can't explain the Android variance - especially the GPS derived time. But it isn't the difference in standards.

    BTW, my previous phone (also on Verizon) varied around 20 seconds slow. Verizon could never explain that to me. I also notice that the phone apps I downloaded, and the native app, don't show seconds. Not sure why. I'm an Android developer, and the framework adds no significant delay.

    January 8, 2013, 6:25 pm