4 ms·
"At this level, maintaining absolute time scale on earth is in fact turning into nightmare," Ye says. This clock they've built doesn't just look chaotic. It is
by jessaustin 12y ago
"At this level, maintaining absolute time scale on earth is in fact turning into nightmare," Ye says. This clock they've built doesn't just look chaotic. It is turning our sense of time into chaos.
This seems a bit sensationalist. The new clocks will be more accurate than any we have now, for existing applications. Those who develop new applications that actually use the increased precision will just need to be more careful about which frames of reference they use. Those who built GPS already solved some version of these problems.
- sp332 12y agoIf you can't synchronize the clock with any other clock in the world, timekeeping gets a bit meaningless.
- droopybuns 12y agoIf we designed houses around planck lengths as the unit of measurement, we'd get nothing done. It's not that timekeeping gets meaningless- it just means that there is a threshold where the variability of unsynchronized clocks disrupts certain highly precise problems.
- treve 12y agoI was looking for someone to make a comment like this. I also felt that it simply meant that the last few digits were less meaningful for applications such as sync, but it definitely doesn't mean that that's the case for all the digits.
- GauntletWizard 12y agoBut you can predict and account for relativity; We already have good protocols for dealing with expected drift. NTP already takes into account that your PC is not keeping time as accurately as the upstream servers, and keeps a driftfile that guesses at the time differential needed. The NTP daemon silently adjusts the time every second to deal with this. If two clocks are at different heights, they can compute the differential in their rates and agree on a time unit. Sure, it might be 1 quadrillion ticks in on one clock and 1 quadrillion and 1 ticks on the other, but it's still a 'second' so long as both sides see it as simultaneous.
- sp332 12y agoThat doesn't help if one clock is drifting because it's being moved by tectonic activity, or being vibrated slightly by gravitational waves.
- jessaustin 12y agoIf you have a protocol and enough fancy clocks, you don't have to know what's causing the perturbations or how big they are. At some level, there isn't actually a "real" time at all (partially, for the reasons you state). Rather, there is a time that we can agree upon. If some precise operation requires better time coordination than we can obtain via a particular configuration of these fancy new clocks, then we just need a different configuration, probably with more of them.
- sp332 12y agoIt's not a perturbation in the timekeeping of a single clock. This clock keeps incredibly reliable time. The fact is that time passes differently for each location in space. If you put another clock at a different altitude, they would keep different time, and they would both be correct.
- jessaustin 12y agoLet me put it this way: if we had really fast trains (or just imagine some other schedule that must be coordinated over some distance), then it would matter that time passes differently for each location. However, we could still deal with that, with an appropriate time communication protocol and enough fancy clocks (probably spaced out along the train track).
- SiVal 12y agoEven if you could measure altitude to centimeter precision, the clock speed would depend on gravity, not altitude, and the Earth is too lumpy (in terms of density) for that centimeter to tell you much when comparing two different locations. I also think (speculate) that with this degree of sensitivity, we will be able to measure some gravitational fluctuation due to the currents of molten iron swirling inside the Earth. With all this uncertainty, all you could do, after getting two clocks to approximately the same altitude, would be to say that clock A is currently running a factor of X faster than clock B. What you wouldn't know is how much of the delta was due to altitude, how much due to unequal density underneath, and other such factors, compared to how much was due to the internal differences in the clocks themselves that you were trying to factor out by using multiple clocks.
- ithkuil 12y agoIf you can't synchronize the clock with any other clock in the world, the increased precision of this kind of clock gets a bit meaningless (or at least has only a very local usefulness).
- aqme28 12y agoThat's only for measuring the time since 0 BC. If you're trying to measure the duration of a very precise experiment to a very precise degree, this kind of clock can be very useful.
- Strilanc 12y agoYou can still synchronize them, it's just that "synchronize" takes on a more complicated meaning. Both sides can use their clock, and the known positions of the two clocks over time, to predict the time that the other clock should be reading. The time deltas are no longer the same on both sides, but you can still transform between the two in order to compare them and detect errors occurring. More pragmatically, we'll probably agree on a standard centralized time (i.e. International Atomic Time) and everyone will translate between their own locally perturbed time and the standardized time when synchronizing.
- deciplex 12y agoIt's not like the issue they're dealing with doesn't exist with literally every clock in the world - it's just that we didn't have a clock that was sensitive enough to notice before. In a sense, this is not a property of the clocks, but rather just the way nature is. We can't keep these clocks in sync at various places around the world, because the thing they measure is not in sync at various places around the world. But, we've known that for about one hundred years - what's the big deal? nullc is right, this is a bad article.