back to article Hold on a sec - leap seconds granted a last-minute reprieve

A decision to kill leap seconds and permanently change how time is measured has been deferred until 2015 by the International Telecommunication Union. A meeting of ITU Radiocommunication Assembly reps on Thursday was unable to reach a decision on whether to stop adding leap seconds to Coordinated Universal Time (UTC) to keep …


This topic is closed for new posts.
  1. Peter Mount

    All because they are lazy

    At the heart of this is where they say telecom equipment can't handle the odd leap second (most are added but they can be taken away as well), citing things like navigation - presumably GPS.

    The thing is that GPS doesn't use UTC - it has it's own time system strangely called GPST which doesn't use leap seconds.

    So since GPST started in 1980 it's now 15 seconds ahead of UTC

    A good example of the various clocks & how far they are out:

    1. Kevin Johnston

      Quite agree

      I don't see where the problem is with retaining the current system of GMT (yes I know they re-badged it just to annoy the British and other pedants).

      So there is a difference between Earth time and Atomic time, big whoop. Since everyone has known this for some time and they have been playing with leap seconds for a huge number of years then anyone building kit which can't cope deserves to fail, it's like trying to sell 'self-setting' clocks globally that can't do timezones. Without any training in this stuff I can see immediately that you could make stuff either accept leap-seconds when required or else refer to a counter somewhere which tells you what the current discrepancy is between the two.

    2. Yet Another Anonymous coward Silver badge

      That's the exact problem. You get the time from GPS which has to and look in it's copy of The Farmer's Almanac to find out how many seconds off it should be.

      If instead we all just adopted the sensible solution of counting seconds since 1970 and let the astronomers and druids worry about what local time the sun rises.

      Now we have won this battle we can start campaigning to have the 0deg meridian moved back to Greenwich

  2. Silverburn

    Treating the symptoms, not the cause

    I dunno why we're bothering...shouldn't we just fix our orbit, so it's perfect, instead of mucking about with the clocks?

    I hard can it be...?

    1. jai

      this! why _should_ we adapt to the imperfections of nature? we're top of the evolutionary tree, that gives us the obligation to give something back to nature, and fix it were we find faults.

      i read a story a few years about (i think it was some steampunk thing) where some evil genius had worked out a way of adjusting the orbit of the planet by causes all the volcanoes on a single continent to erupt at the same time.

      that was just science-fiction, but i'm sure the world's top boffins wouldn't have too much trouble applying the theory in practise.

      1. Silverburn

        As long as we don't leave it in the hands of the boffins who crashed the mars probe because they got their metres and yards mixed up...tha'ts all I'm sayin'...

        If we're really clever (or just lazy, and blag the idea from Futurama), we could fix global warming at the same time by bumping ourselves into a higher orbit.

    2. Dances With Sheep

      How much would it cost us to change the laws of physics?

      1. Evan Essence

        Ye cannae.

    3. tony2heads

      NOT the orbit

      Its the earth's rotation about its axis (day) not the orbit about the sun (year) that's the problem

      1. TeeCee Gold badge

        Well that's simple then.

        Instead of leap-seconds, we just get everyone to run towards the west at once for a bit to speed it up...

  3. ratfox

    Do seconds really matter?

    Couldn't we wait until there is, say, a difference of one hour, then simply skip one daylight-saving change?

    1. Evan Essence


      No, because when daylight saving starts and ends, it's the local timezone which changes, not UTC. Regardless of daylight saving, UTC always increases at a steady rate (except for leap seconds).

    2. Simon Neill


      Because that would take a bloody long time. These occur max once a year, so its a long time to get a whole hour together.

      On the other hand, seriously...we can cope with an hour change twice a year but not a twitch of one second? WTF?

  4. jai

    3 years?

    so, another 3 years to debate this? seems more like a vote for 3 more years of milking it. lets face it, it's going to be 88 years at the earliest before they next get a chance to debate this issue, might as well try and get as much out of it as possible while they can

  5. FutureShock999

    Worry warts...

    Do those people really worrying about the divergence pay any attention the to probability of the human race actually LASTING until 2700?!?!? Have they looked at the probability of a meteor strike big enough to kill us all? I mean, it's nice to have a human-centric view on all this, but let's face it, the odds are we won't be here long enough to notice the divergence...but maybe our machines will....

  6. Fuzz


    As anybody who has watched superman will know, time is governed by the rotation of the earth. Therefore it's the atomic clocks that are wrong so it is imperative that we keep the leap second.

  7. Robert E A Harvey
    Thumb Up

    And also

    " Other member states wanted further probing."

    You really can't resist, can you?

    Keep it up.

  8. Anonymous Coward
    Anonymous Coward

    The solution is simply to recast UTC from being a human readable time into just a number (for instance, 2012-01-20T10:39:49.1234 into 1327055989.1234). This will operate perfectly well for the people who actually need to use UTC who have the current technical complaints, since they in the main don't care about absolute times, only relative times. UTC can merrily proceed without leap seconds, intervals between two UTC times will always be exact, and we won't run into problems with the sun rising at midnight any time ever as the association with human time will have been removed. Of course, this would also mean restoring GMT as the basis for time zones, which will surely please the French... so a win all round!

    1. Evan Essence
      Thumb Down


      There's no need to make such a change to UTC, because there's already a time standard for the job. This is TAI (atomic time). As I say below, let systems which need such accuracy use TAI directly.

      And there's no need to revert to GMT because it really is very convenient that there's always an integral number of seconds' difference between UTC and TAI; the difference between TAI and solar time (on which GMT is based) is, as we know, subject to the uncertainty of the earth's rotation.

    2. Darryl

      @AC Just a number

      Even better:

      Stardate -310946.87345046555

  9. Evan Essence

    Crazy, stupid proposal

    Has the ITU considered what would happen at some unspecified future time when a leap hour occurs? Effectively, UTC would stop for one hour, so financial transactions needing sub-second timestamps couldn't occur, world wide, for that hour.

    Let systems needing a regular time standard use TAI or GPST, and let humans keep their existing time, based approximately enough on the position of the sun (GMT is *not* exactly UTC, but for human purposes they're more than close enough). TAI-based systems would be able to keep absolutely unambiguously-timed records, and computers can convert where necessary to our ambiguous, messy, human timescales. It's what computers are *for*.

  10. Jason Bloomberg Silver badge

    Divisive approach

    Why not leave UTC alone and simply add another time measurement system so people can use whichever they want and is most appropriate for their uses?

    One group or the other isn't going to be happy with whatever is decided so simply let them co-exists. Seems more wanting to control the world and intolerance of others' than living in harmony if you ask me. What's so wrong with agreeing to disagree?

  11. Velv

    Time is an illusion

    Lunchtime doubly so.

    1. bazza Silver badge

      Neutrino ride?

      Did that joke just step onto a neutrino and ride back in time to another forum a couple of days back?

      1. TeeCee Gold badge

        No, you're just in the wrong frame of reference.

      2. Blofeld's Cat

        @Neutrino ride

        No, it's still on Puckoon Time (Currently 4:32)

  12. Steven Jones

    How about two time standards...

    Could they not just establish some form of time standard which is decoupled from "Earth Time" and then maintain an appropriate offset. For those applications where an "absolute" time standard matters, then use that in conjunction, where required, with "Earth Time". For everyday apps where "Earth Time" matters, it's relatively straightforward to "stretch" time a little to provide for the odd "leap second" without creating big jumps.

    It's not as if computer systems don't have to deal with multiple time standards anyway. Many systems have to cope with different time zones and so-called "daylight saving".

    1. Anonymous Coward
      Anonymous Coward

      I vote this

      Can we call it stardate?

  13. ridley


    Using Sunrises to set the clock would be a bit tricky (as the time the sun rises changes every day) much easier to use noon instead.

    Which of course is why they did use noon to set the clocks.......

    I will get my coat.

    1. StephenD
      Thumb Down

      But the time mid-way between sunrise and sunset or when the sun is at its highest in the sky isn't fixed, either, but varies considerably (more than 30 minutes) during the course of the year.

      1. ridley

        Can you please explain how the time between noons, when the Sun is highest in the sky, varies during the year.

        Are you saying that the rotation of the Earth speeds up and slows down by up to 1/2 hour during the year?

        1. Evan Essence


          No, the rotation of the earth doesn't change (at least, not by that much!)

          The time between noons changes due to two effects: the tilt of earth's axis relative to its orbit around the Sun, and the fact that the orbit is elliptical. The effects combine to produce "The equation of time", which gives the deviation from the Mean time (as in Greenwich Mean Time).

          More details on Wikipedia:

          1. ridley

            Thanks for that, reading it now. Having a bit of difficulty visualising it though.

  14. micheal

    all this

    Another great British invention lost to the Eurotwats

    It's called GMT cos it starts at Grenwich.

    Like us moving to European time, all measures by them Germans to take over Not So Great Britain again, like in 1939 and 1918

  15. Old n Cynical


    I don't really think the Americans are best placed to lecture on good practice here [links follow]:

    There are other such examples :)

    1. David Harper 1

      And let's not forget Idaho

      The U.S. state of Idaho is in two time zones. The southern part of the state -- which includes most of the population and the state capital Boise -- keeps Mountain Time (7 hours behind GMT), but the northern part keeps Pacific Time (8 hours behind GMT), so you have to change your watch as you travel north/south, not east/west.

      The reason is that north Idaho has closer economic ties to eastern Washington state (which keeps Pacific Time) than it does to the rest of Idaho, so the clocks in north Idaho match those in Washington.

    2. Anonymous Coward
      Anonymous Coward

      There is a reason for the discrepencies/anomalies in Indiana, and a few other places, mostly because they lie on the boundaries of time zones, and certain parts of the state, with small parts having stronger economic ties to communities outside the state's boundary. Indiana is MOSTLY rural farmland, with just a few areas built up. A few counties in NW Indiana have much stronger economic ties with Chicago, IL than further into IN, so, it makes more economical sense to have their clocks set the same as Chicago.

  16. David Harper 1

    The real reason why we're in this mess

    The fundamental problem in the UTC versus TAI debate is that the length of the second in the SI system was defined in 1967 so that it matched the then-standard astronomical second.

    That in turn had been defined in 1952 in such a way that it matched 1/86,400 of the length of the average day sometime in the middle of the 19th century.

    So the SI second doesn't even begin to approximate 1/86,400 of the current day length, and this is why leap seconds are needed every few years.

    It gets worse. The Earth's rotation rate is diminishing at a roughly linear rate, so the UTC-TAI clock error is growing quadratically. By the end of this century, we could be adding leap seconds every two or three months.

  17. Kubla Cant
    Black Helicopters

    Cui bono?

    If this proposal is adopted, then the prime meridian will drift (very) slowly towards Paris, where the French wanted it to be in the first place.

    Revenge, as they say, is a dish best eaten cold.

  18. Chazmon

    detective novel?

    I can see a pottential Dorethy L Sayer style detective story with a murder's alibi dependant on the discrepancy between the two times.

    'How could I have murdered proffesor Andre? You can clearly see from this security camera footage that I was in the resturant eating pudding at this point....'

    At this point something silly with flying cars and robots will happen as the story would have to be set in 2700.

  19. Mister Cheese

    Tectonic shift

    Not sure how relevant leap-seconds are - Greenwich isn't even on its own meridian any more...

  20. jake Silver badge

    I run three clocks.

    The first is TheWife's monthly cycle. If you are married, you'll grok.

    The second is the seasonal clock handily provided by the Solar Year & the Earth's axial tilt with respect to its orbit. It is totally out of my control, but I plant my fields & breed my critters by it, as humans have since time immemorial. Trying to change this is a fool's errand.

    The third is the clock provided by the Master clock on my network, which syncs up to an atomic clock once per day ( works for most purposes ... I use something else), which all of my machines adhere to. This is for computer record keeping more than anything else.

    Context is the key. There is no "SingleTimeStandard[tm]", and never will be. With the exception of TheWife's, of course ;-)

    As a side-note, I don't wear a wristwatch ... and haven't in over a third of a century (since my HP-01, back in 1977, in fact). In my mind, they are completely pointless. Everywhere you look these days you can see something giving you a pretty good approximation of "local time". Humans living life to the second or minute (or even ten minutes!) is counter productive. Even when baking bread ... Relax, be patient, learn to make homebrewed beer :-)

    1. Anonymous Coward
      Anonymous Coward

      I've worn a watch for two thirds of a century and find it immensely useful.

      1. TeeCee Gold badge

        It must be getting a bit smelly under the strap by now.....

    2. Heironymous Coward

      Second the three clock system

      and add one (between WifeClock and SolarClock in importance) - a stomach clock, which tells me it's lunchtime / dinnertime / pubtime.

      1. jake Silver badge

        @Heironymous Coward

        Agreed. Four clock system it is :-)

        This round's on me, and I almost never use icons.

  21. Tom 7

    So clever they can make a GPS system.

    So clever they can make an atomic clock.

    Yet they try and change something that isnt broke..

    with a comitteee???????

  22. billm2257

    Did you hear that? "two to three minutes in 2100 and about 30 minutes in 2700" Run for your lives!!!! It's 2012 for god's sake. 700 years from now, our clocks will be "off" by a half hour. This is such a non-issue it's not even funny. In the year 3000, people will be waking up and going to work at 7am instead of 8am. Oh no, where will it end? Stop the insanity.

  23. Anonymous Coward

    Good result

    You all miss the importance of stepping back from an intellectual precipice yesterday from an idea which assigns the wrong number of rotations to the equivalent amount of days.No point is discussing the accuracy of atomic clocks if scientists are 4 rotations out in 4 years as they assume the Earth turns 1465 times in 1461 days when everyone here will experience 1461 days/rotations.

    The fundamental unit of time is not a second or any fraction of it,it is not the minute,hour or 24 hour day and it is not even the orbital period of the Earth in 365 days 5 hours 49 minutes.The fundamental unit of time has been known for thousands of years and it is the proportion of days to years,in this case 1461 to 4 years where an extra day is added to a constant stream of 365 days to keep us fixed within the orbital points of equinoxes and solstice.

    Before everyone gets in a jam again,I suggest they start with the natural cycle and then work in how very intelligent humans overlaid a calendar structure on this natural phenomenon -

    " on account of the precession of the rising of the Divine Sothis by one day in the course of 4 years.. therefore it shall be, that the year of 360 days and the 5 days added to their end, so one day as feast of Benevolent Gods [the pharaoh and family] be from this day after every 4 years added to the 5 epagomenae before the New Year, whereby all men shall learn, that what was a little defective in the order as regards the seasons and the year, as also the opinions which are contained in the rules of the learned on the heavenly orbits, are now corrected and improved.." Canopus Decree 250 BC

    So,when Feb 29th comes around in a number of weeks as the 1461st rotation and within that 24 hour period there will be another sunrise and sunset due to the rotation of the Earth,somehow you all may learn how we managed to screw it up with the most impeccable credentials -

    "The Earth spins on its axis about 366 and 1/4 times each year, but there are only 365 and 1/4 days per year." NASA

    The 'leap second' merely exposed an issue that is many,many magnitudes worse than the Piltdown Man episode with far more at stake.The problem is that the same people are acting in the same way -

    "Anthropologists refer to the Pilltdown man hoax as 'another instance

    of desire for fame leading a scholar into dishonesty' and boast that

    the unmasking of the deception is 'a tribute to the persistence and

    skill of modern research'. Persistence and skill indeed! When they

    have taken over forty years to discover the difference between an

    ancient fossil and a modern chimpanzee! A chimpanzee could have done

    it quicker." Daily Sketch

  24. hrjacobson

    Compromise proposal

    Why not compromise? Publish a list of leap seconds that will be added during the next 10 years, based on the best estimates of Earth's rotation. Every few years, extend the list farther into the future making adjustments as needed. This gets rid of the uncertainty of getting only six months' notice. The solar time may occasionally be off by a second or two, but in the long term it will continue to agree with UTC.

    1. Ken Hagan Gold badge

      Re: Compromise proposal

      Why limit yourself to 10 years? Those arguing that the sky is about to fall in seem to be pretty confident about how many minutes we'll be out by 2100, so why not write up a schedule for the next 90 years and revise it in 2060?

      This "six months notice" business arises from a desire to keep UTC and solar time within 1 second of each other. Nobody outside the clocks community gives a monkeys about that level of accuracy. Loosen that to about 10 seconds or a minute and surely it all becomes more predictable.

      1. hrjacobson

        The length of the schedule in the future is determined by the lifetime of computer software and hardware. If someone is foolish enough to hard-code the leap-second table into software or burn it into hardware, with no easy means of update, they would like to know that their system will continue to function for X years. I am personally aware of software that has been in use for more than 20 years so you're right, 10 years is probably too short. Aside from astronomers and operators of Earth-observing satellites, I doubt that anyone really cares if Earth's rotation is off schedule by a few seconds.

      2. This post has been deleted by its author

      3. DavCrav

        Minutes, not seconds

        "Why limit yourself to 10 years? Those arguing that the sky is about to fall in seem to be pretty confident about how many minutes we'll be out by 2100, so why not write up a schedule for the next 90 years and revise it in 2060?"

        You can guess how many minutes out, 2-3 to quote the article, it will be in 2100. But the whole point of this is incredible accuracy, and saying "well, the number of leap seconds will be about 150" isn't going to cut it.

    2. J.G.Harston Silver badge

      If you *know* how the irregularities in the earth's rotation is going to change in the next ten years so that you can tell when leap seconds are going to be needed, can I have next week's lottery numbers please.

  25. MrHorizontal

    Wait a sec...

    ... both the status quo and the 'new' have it wrong.

    First and foremost, the Gregorian calendar is wrong with the concept of a leap year. The earth takes 365.25 days to revolve around the sun. We can make an existing 'day' 1.00068493150685 days without too much impact. That fixes the February 29th every four years issue, and just never have it.

    Now leap seconds. As has been said, the earth's spin on its axis around the sun causes the issue with leap seconds. The issue at hand is what 'time' sunrise and sunset happens. As we know full well in Britain, getting up in darkness and leaving work in darkness in winter is a spirit-sapping experience. The point being is while it will take a good 100 years before there is a noticeable difference, given that at some point in time in the future (granted, very, very far in the future) under the proposed no-leapsecond time you will go to sleep at 11 pm in the UK and the sun rises, it's not representative of 'time'. By this virtue, technology 'because it's easier' doesn't help solve the problem.

    Personally, just state that at 12 midnight on New Years, the second that strikes is X+Y seconds, and if Y is a +2 second or we miss 1 second for a given year, it happens. The entire world is drunk out of their minds, so who the hell cares for +/- 1 second? Certainly not technology, if it's internationally proclaimed that all time is 'fixed' at a given point in the calendar, then technology can do what it's designed to do: serve humankind. And in the meantime, you and I can continue moaning about British winters, and in the meantime have to drink a little bit faster or slower by a matter of a second or two every New Years.

    Does it matter? It's all relative. I say, cheers!

  26. JeffyPooh


    Leap years where *an entire day* is inserted every four years (excluding every hundredth year, but including every four-hundredth year), ...this is okay.

    Daylight Savings Time (sic, and inherently stupid), where entire time zones are chopped up into random segments based on latitude and attitude twice a year, ...this is okay.

    (My company's SAP ERP systems apparently have to be shut down for the entire day twice a year due to DST [??!!??].)

    Odd ball times zones with 30-minute or other essentially random offsets, ...this is okay.

    All this and so much more, and some noobies can't handle a 60th second ("...58, 59, *60*, 00, 01...) once in a while. A process that has already been going on for a very long time.

    All they need is a Best Practices Handbook to help the noobies. Or, worst case, add another 'GPS Time' like version of time and make sure people understand which one to use for different applications.

    No need to change UTC.

  27. Neoc

    I must be missing something basic

    My computer (Linux in this case, but I'm pretty certain it's the same for most modern PCs) keeps time by keeping track of the number of "ticks" (let's call 'em seconds for simplicity's sake) since midnight Jan 1, 1970 UTC. It then converts this count into a date or time when asked to do so.

    So as long as the piece of software which does the conversion knows how to insert leap-seconds (it certainly knows how to insert leap-days every 4-or-so years), then where is the problem?

    Force all time-keeping devices to work the same way - keep track of the ticks (second) and convert to "human" time when asked. So the internal count would align with the TAI, while the displayed time would align with the "time-of-day" and seasons. All time-based communications would be synchronised via the tick-count.

    Where's the problem?

  28. Mips

    Two things to consider

    In the golbal scheme of things the day, hour and second dos not matter to the universe only to Earth, so save for scientific purposes there is no point in having a "standard" second. There is also no point in trying to reconcile the variable rotational values of the Earth to a standard second. The leap second, minute, of hour is an inevitable part of living on Earth. So it is time to stop the argument.

    Has anyone noted the admission that the Earths rotation is slowing in a measurable fashion. It is something I have written about in a number of places. The rotation of the Earth and the tilt of the axis is responsible for the fact that the surface is habitable almost pole to pole; without these the there would be large areas baked or frozen to death with only small habitable zones. Loss of rotation is vitally important to life on Earth. One day we will find out if how life will manage with a 25hour day. Not well, I imagine.

    The tidal regime of the Earth is causing this loss of energy from the rotating mass yet we still have proposals to use tidal power which removes even more energy from the system. And lets be clear about this tidal power takes additional energy not the energy that would be lost anyway. Talk about mad! Dr Strangelove would be proud.

This topic is closed for new posts.