[LEAPSECS] The relation between calendars and leap seconds.

[Poul-Henning Kamp]

In message <alpine.LSU.2.00.0811111200170.23184 at hermes-1.csi.cam.ac.uk>, Tony F
inch writes:


>> As we saw a couple of years ago, the 400 year leap-year role is slightly

>> above the level of complexity humans can deal with, a significant

>> fraction of "people who should have known" didn't.

>

>I did think of that, but we've just been discussing some much more complex

>examples of arithmetic calendars. I couldn't argue that there had been a

>historical trend towards simplicity as well as predictability. But I agree

>that simplicity is an important requirement.


I think we have to qualify predictability here, and it ties into
simplicity somewhat.

If we have "permanent predictability", as for leap years, then utter
simplicity is less important, because the rules, semicomplex or
not, can be encoded once and for all, and it is perfectly fair to
ridicule anybody who gets it wrong, having had a generation or
more of notice.

If we only have "temporaray predictability", ie: we predict the
outcome for some particular timehorizon, as currently with leap
seconds, or with the Ramadan depending on who spots the moon first,
then simplicty comes very much to the foreground.

If nothing else because a public awareness campaign must be carried
out, and you really don't want to newspaper adds to say things like
"Adjust your timekeeping according these trignometric formulæ".

The fact that a large fraction of people, at least in countries
where the "spring forward, fall back" mnemonic sentence does not
translate, have trouble figuring out which way to adjust their
clocks on DST changes, should be a stern reminder that the general
population, and by inference, a lot of computer programmers, are
horribly incompetent when it comes to timekeeping.

Another issue is that exception rules should not trigger so seldom
that people forget about them.

Again the 400 year rule is exhibit number 1, and exhibit number two
could be how Sweden failed to follow their own established rules,
trying to recover from the old to the new calendar.

And exhibit number three could be the Indonesia Tsunami.

Most people have, at one point or another in their life, been told
that the correct response, if the water suddenly disappears from the
beach, is to run like hell inland.  But since most people never
experience this in real life, they don't when they should.

Poul-Henning

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.