US DoD Poll On Leap Seconds

@10u8 writes "For time scales to leap, or not to leap, has been the question here before. The ITU-R will be considering leap seconds again in a few weeks. This week the USNO posted a survey about leap seconds by the US DoD. The issue has civil implications as well as technical ones, and there is a demonstrated way to respect the history, remove leaps from navigation and POSIX time, yet keep the sun overhead at noon."

Comments

Are leap seconds really all that important?

[NoobixCube]

I thought we had leap years to take care of the discrepancy between our calendar and the actual orbit around the sun. Would a leap second even be made longer by any noticeable amount? What about sporting events? Someone who misses out on a world record by a tiny bit would complain that the record h older had more leap seconds in his race! (Okay, that one was a joke, but the rest I'm serious about)

Re:Are leap seconds really all that important?

[maxume]

The leap seconds do the same thing as the leap years (each leap day moves the calendar closer to the orbit, but not exactly to the orbit).

Re:Are leap seconds really all that important?

[Hurricane78]

I want to solve: There is no such thing as absolute time.

Kill DST instead!!!!

[wealthychef]

I'd be more interested in killing Daylight Savings Time than dealing with Leap Year.

Re:Kill DST instead!!!!

[illegalcortex]

DST can be "fixed" by recording time in UCT. No such "fix" exists for leap seconds. With leap seconds, you're getting down to the fundamentals of how time is recorded, not how it is translated to local time.

Re:Kill DST instead!!!!

[Kjella]

Killing it? I want to change them completely, and wintertime too. Now, I live a bit further north than most people (60 degrees latitude) and what happens in the winter is that I, like most people, head to work in the dark and come home in the dark. Maybe you get to see some sun on your lunch break, but unless you got an office with a view you won't see much of it otherwise. If we have like 6 hours of sun, they should be 4PM-10PM so you can do some outdoor activity after work. What happens now is I sit indoors during the day because of work, and I sit indoors in the evenings because it's dark and cold outside. I haven't got any stats to back it up but I'd think most people work indoors these days, the reason to have light == noon so you could run around outside just isn't there. I'd be happy with mornings that suck (some more) and evenings that were bright and nice all year round.

Or played with GPS etc

[EmbeddedJanitor]

Keeping leap seconds synced is pretty important across comms networks.

Re:Or played with GPS etc

[snowraver1]

You missed the point. In a system like GPS accuracy is paramount. The birds need to know where they are, and are updated as they deviate. Likewise the need to be synched and need to be kept in synch. They do not run off the same clocks as humans though, so do not need to be updated with a leap second update. To the satellites, time goes on...

Re:Automated and consistent leap seconds

[klapaucjusz]

There should be a planned algorithm that kicks in,

This assumes that we know when, in the future, we'll need to insert leap seconds. And we don't.

Leap seconds are introduced in order to compensate for medium-term variations in the earth's rotation speed. We don't have a good understanding of the way the earth rotates -- knowing what UTC time it will be in ten years' time is about as difficult as predicting the weather for next week-end.

Yep...

[msauve]

and if anyone doesn't like leap seconds, all they have to do is use one of the time scales which don't use them, like TAI.

It's exceedingly silly and stupid for people to keep trying to change UTC so it doesn't track solar time. That what it was intended to do. If you made the wrong choice, live with it, or change time scales. If it's being forced on you, quityerbitchin', and convince whoever decided on UTC to change. Stop trying to turn UTC into something it isn't, there are other people out there who made an intelligent decision, and depend on it's characteristics.

Re:Yep...

[Phil Karn]

You are absolutely right. UNIX and its derived systems simply made a big mistake in choosing UTC for internal use. They should use an atomic time scale, either TAI or GPS time internally, I don't really care which.

If they really have to, UNIX could define their own epoch with a zero offset to UTC as of right now. Then timestamps made in the past few years won't have to jump in the changeover. This would give exactly the same benefit as no longer applying leap seconds to UTC without removing UTC's ability to track earth rotation time.

Whatever timescale UNIX chooses, it MUST have a known offset to TAI that remains fixed for all time. Period.

It's just absurd that every time there's a leap second it ripples through the whole NTP network for hours. GPS receivers ride smoothly through leap seconds because they don't see them. Why should glitches happen in NTP/UNIX?

It should be up to the library routines to properly handle conversations between internal time and human-friendly UTC representations, driven by updated tables of leap seconds in the same way they're already driven by updated tables of daylight savings time. Both are unpredictable and subject to administrative whims. You can't base internal timescales on them. I'm tired of having to write these routines myself for my satellite tracking programs.

It's important to remember that timescales based on the rotation of the earth simply didn't exist before certain specified dates. Before 1961, UTC simply didn't exist. There's simply no proper way to date an astronomical event back in 1900 in UTC.

Even worse, between 1961 and 1972, frequent ad-hoc frequency offsets were introduced into UTC to keep it close to earth time. The UTC second and the TAI second differed slightly, and this difference was constantly changing! Only in 1972 did the present leap second system start, with the lengths of the UTC and TAI seconds exactly equal. It was an improvement over the previous system, but it's still no substitute for an atomic time scale for basic use.

Re:So long as we don't have leap nanoseconds

[Martin Blank]

Changing the length of a second will end up changing almost everything in our lives. It would be an enormous undertaking, redefining, among many other things, electromagnetic wavelengths and the speed of light. Speed limits would change, computers would have to handle travel time calculations differently, and the length of the workday would change slightly.

It was hard enough to get the world to change to the metric system (with notable holdouts still remaining). Changing the very definition of one of the six core SI units would be nearly impossible.

Re:Why not just change time pieces to include the

[John Hasler]

> We currently cannot predict those variations, and as such, the leap seconds are
> determined based on astronomical observation and applied as needed.

I know that, but zoneinfo has to be updated frequently anyway to accomodate the whims of princes.

Re:Who cares about all this timezone crap?

[mr_matticus]

I think the US should have one time zone, and it should be based on NYC time. -5/-4 UTC for everybody! Who cares if California would have the sunset at 2230 today?

Fine by me, and I'm a native Californian.

There's no rule that says business hours need to be 9 to 5. Since you already have to convert what "time" it is in a faraway place you're calling, it's not a big deal. In San Francisco, I can't make any calls after 1pm to East Coast offices and expect to get anything done. What the clocks say over there doesn't really matter. For all intents and purposes, New York business hours are 6am to 2pm from my perspective.

I'd be content for the entire planet to move past the idea of time zones entirely. It's an outdated idea from a time when you needed physical references to the passing of time, and when it didn't matter that the times didn't line up in faraway places. Just think of all the things it would simplify: flight arrivals/departures, conference calls, news stories--and it would make am/pm an unnecessary distinction, too. 0514 would really be 0514. Everywhere. I'm okay with "business hours" for me being, say 0100 to 0900, and 2200 to 0600 in some other place. They're just numbers.

Tradition and conditioning, however, are unbeatable--and the idea of "noon" being the middle of the day has undeniable intuitive appeal (even if it's rarely accurate).

Re:Leap seconds fix a diferent problem

[n3tcat]

so our clocks are more precise at measuring how fast the earth does a twirl than the earth is at twirling?

wait, what?