[LEAPSECS] drop in what bucket?

Rob Seaman seaman at noao.edu
Tue Feb 11 01:28:22 EST 2014

On Feb 10, 2014, at 4:49 PM, Warner Losh <imp at bsdimp.com> wrote:

> The leap forward or back an hour due to increasing out-of-syncness with the sun would be a drop in the bucket.

A standard bucket is around 10 liters:


And a drop is 0.05 mL:


So if a leap-hour is 5 micro-buckets, that makes a leap-second 1.4 nano-buckets. Or alternately a leap-second is a drop in four of these size buckets:


That's even less than a drop in a bucket, isn't it?

>> Predictions of future events (say, solar eclipses) would be unable to engage with a local time that might differ +/- one hour rather than a few seconds.


> The same problem exists with leap seconds...

Well, no, both because the leap seconds apply worldwide not at the whim of local jurisdictions, and because they're only 1/3600th the size. If your quantizer has a step size 3600 times bigger, the quantization noise increases, too.

> The timezone decisions would be made on a national level, just like they are today. Salt Lake City and Denver are on the same time, even though the sun sets half an hour later in SLC.

Salt Lake City and Phoenix share the same meridian, yet for most of the year clocks in SLC are one hour later than clocks in Phoenix. Without leaving Arizona driving from Holbrook to Flagstaff can require resetting your watch six times:


since the Navajo observe daylight saving and the Hopi and the state of Arizona do not. Timezone decisions are made locally in the U.S.

If adding layers of complication to the current timezone system is conceived as a solution to some political or engineering problem, doesn't it seem reasonable for a document to be written describing how this would work and what the local issues are likely to be? Might it not be pertinent to cite a compendium of current local timezone rules and of the processes that created those rules? The Olson DB attempts to capture the what, where and when of timezones. Where is the corresponding description of the who, how and why?

ITU RA-12 requested studies of issues pertinent to the question of redefining UTC. If timezones are really deemed pertinent to this question, then information regarding timezone policies are an obvious area for study. Who has undertaken this research project and when might we anticipate reading the work products?

> How so. I don't understand this at all. You add an hour or not after the difference gets large enough to care, and that process operates on a time scale of decades or centuries.

Who's "you"? These decisions are controversial and subject to buyer's remorse:


resulting in the need for:


> I said Atomic Time can be the basis for Civil time. Universal Time never entered into it, other than being displaced as the basis for Civil Time. It would exist, unchanged, like it has existed in all its various flavors.

That is not the proposal. The ITU proposal as it stands would redefine Coordinated Universal Time and severely damage the current concept of Universal Time, likely beyond recovery.

> My basic argument is that we can coordinate on atomic time, base civil time on that and synchronize to the sun with an offset to the time zones that sovereign states use today to determine what they want their people to use.

Sovereign states, except that you are requiring every sovereign state in the world to shift its timezone every few hundred years. Including, for that matter, sovereign states as yet unborn.

sov-er-eign (noun):
1 a : one possessing or held to possess supreme political power or sovereignty
b : one that exercises supreme authority within a limited sphere

These diverse jurisdictions will not actually possess the power to set their own clocks, but will only be held to so possess it since action will be forced on them by a sort of temporal Mandamus:


Their spheres will indeed be limited, and not only metaphorically.


More information about the LEAPSECS mailing list