When will physics make 32-bit signed leap-second counts obsolete?
A while ago I vaguely remember a prediction on this list of when the TZif file format will become obsolete because we have accumulated 2**31 total leap seconds if we continue to just add leap seconds as needed to keep UTC roughly in sync with TAI. To some extent this is just a playful prediction, as the format will undoubtedly become obsolete for other reasons long before then, but it's still fun to predict. Unfortunately I don't remember the details, and a quick Google search didn't find it. Does anyone else remember this? If not, I'll have to recalculate. The topic has just come up again on the tzdist-bis mailing list, which is why I'm thinking about it again. See <https://mailarchive.ietf.org/arch/msg/tzdist-bis/Q7OVmd60wydqcJahS3QRNHwtXHM>.
On Thu 2018-09-27T08:45:23-0700 Paul Eggert hath writ:
A while ago I vaguely remember a prediction on this list of when the TZif file format will become obsolete because we have accumulated 2**31 total leap seconds if we continue to just add leap seconds as needed to keep UTC roughly in sync with TAI. To some extent this is just a playful prediction, as the format will undoubtedly become obsolete for other reasons long before then, but it's still fun to predict. Unfortunately I don't remember the details, and a quick Google search didn't find it. Does anyone else remember this? If not, I'll have to recalculate.
The topic has just come up again on the tzdist-bis mailing list, which is why I'm thinking about it again. See <https://mailarchive.ietf.org/arch/msg/tzdist-bis/Q7OVmd60wydqcJahS3QRNHwtXHM>.
Extrapolating predictions of Delta T is something that should only be undertaken by the bravest of souls. For visual examples of various different efforts I recommend viewing https://www.ucolick.org/~sla/leapsecs/deltat.html and even moreso https://www.ucolick.org/~sla/leapsecs/year2100.html Most of the predictions suppose a roughly linear increase of LOD, thus a roughly parabolic increase of Delta T. The few seconds of offset between Ephemeris Time, Atomic Time, and Universal Time right now are pretty much irrelevant, but strictly speaking the parabolic formula is CalendarYearOfParticularDeltaT = 1820 + 100 * sqrt((DeltaT + 20) / decel) where decel is expressed in second/century/century A reasonable value for decel is 32 s/cy/cy, and extreme values for decel are 25 and 42, so decel=25 year=930000 decel=32 year=820000 decel=42 year=720000 It is hubris beyond measure to suppose that any calendar or timekeeping system will survive more than 1000 years. -- Steve Allen <sla@ucolick.org> WGS-84 (GPS) UCO/Lick Observatory--ISB 260 Natural Sciences II, Room 165 Lat +36.99855 1156 High Street Voice: +1 831 459 3046 Lng -122.06015 Santa Cruz, CA 95064 http://www.ucolick.org/~sla/ Hgt +250 m
Steve Allen said:
It is hubris beyond measure to suppose that any calendar or timekeeping system will survive more than 1000 years.
Tell Julius Caesar that. -- Clive D.W. Feather | If you lie to the compiler, Email: clive@davros.org | it will get its revenge. Web: http://www.davros.org | - Henry Spencer Mobile: +44 7973 377646
participants (3)
-
Clive D.W. Feather -
Paul Eggert -
Steve Allen