Is this the end of the leap second?



The leap second is a periodic adjustment to the world’s clocks that’s been done every few years since 1972…but an upcoming negative leap second could upend the world’s time-keeping. To decide whether to abolish the leap second altogether, the world’s time-keeping representatives are gathering at the General Conference on Weights and Measures (CGPM) on October 15th. Time will tell…

– Written & Presented by Lizzie Gibney
– Produced by Maren Hunsberger
– Stock Footage: Getty Images / ReflectFilms / Jan Homola / Cristina Bustiuc / sebastian coll / Peresmeh / keybal / Discovery Access / primeimages / sarawuth702 / Vitali Kasporski
– Stock Music: Triple Scoop / Richard Ames
– Supervising Producers: Shamini Bundell and Anne Marie Conlon

source

41 Comments

  1. Wrong definition for the "day" used for millennia. It's not one turn around on its axis; it's always been enough more than that to have the sun come around again. And that was only if they knew the earth turned.

  2. It seems like the problem is that we use time for two broadly different purposes: scheduling (wall clock time) and precision (scientific measurement, high frequency trading, etc.). For scheduling time, it's more important to stay in sync with the rotation of the Earth (people want lunchtime to be roughly halfway between sunrise and sunset, etc.), whereas for precision time it's critical that it be consistent. The ideal solution would be to decouple these fully. Then we could standardize a solution similar the one I believe Google uses: on the day of a leap second, the length of scheduling seconds are adjusted slightly to accommodate the addition or removal of a second. Because it's spread over 86,400 seconds, the error when compared to precision seconds is minimal, but you still need precision seconds to be available, otherwise a time-based experiment would give a different reading depending on what day you performed it.

    It would take some implementing, but perhaps not as much as it seems: I believe computers already differentiate between the two purposes to some extent. The functions used for high-precision timing (e.g. to measure software performance) are distinct from those used for coarse timing, so it seems like you could make an adjustment to one and not the other, perhaps at the OS level.

  3. But even atomic time fluctuates based on the gravity well it experiences. So since the gravity across Earth is not equal, hence even atomic clocks are not stable.

  4. So i keep wondering, why not just let it drift back toward 0?

    Could we not just skip taking away the second this time and skip adding a second next time the clock has drifted 1 second backward?

    The clock is at +1 second and we let it drift back to 0.

    But i guess the first question is "do we expect the clock to get ahead more times?"

    If not or if it will fall behind more often than get ahead, we might be able to keep adjusting in only 1 direction.

  5. Can someone explain why we can't just smear an extra second over a span of a minute when we need to adjust negatively? Isn't that what we already do with video that is slightly different in FPS than the output Hz of a TV channel signal?

  6. As someone who absolutely hates daylight savings time, I think the leap hour is a great idea. Even if it were necessary far more often than it will actually happen, The occasional correction is easily tolerable. As far as daylight savings time is concerned, pick a ** time and stick with it.

  7. It's true, leaving the problem to future generations has never backfired on the decision makers of the time.

    I have no bitterness about other similar decisions made in the past

  8. I've never understood why this is presented as a problem? The vast majority of clocks are not accurate enough to notice a leap second because they drift by seconds each month. Even GPS and online time services regularly drift and need to regularly resync to UTC. Just add or subtract a leap second as needed and the rest of the world will catch up the next time it syncs up with UTC. I don't understand why this would cause an issue.

  9. Why not simply not change anything but not do the negative leap second? Sounds like it's just giving us more time before we need to do the next positive leap second, so isnt it the opposite of a problem?

  10. While we are at it we should also remove daylight savings, time zones and make year 13 months. Every monh would be exactly 4 weeks. Also i live far enough in the north that its already dark in the middle of the day during winter and bright in the moddle of the night during summer

  11. Atleast if we do this silently and add no compensation any AI overlord and their army in the future will have a stroke and fall over due to them not remembering such a minor detail that could take them down and is so specific you'd need to directly mention it to notice it

  12. can't we just wait until the next leap second is positive again? and when it doesn't happen, just subtract the leap second when testing has been done
    also make it skip 00:00:00 and not 23:59:59, it's weirder to see the computer skip 59 rather than 00

Leave a Reply

Your email address will not be published. Required fields are marked *

You might like

© 2026 Cantinho do Vídeo - WordPress Video Theme by WPEnjoy