[LEAPSECS] Look before you don't leap

[Poul-Henning Kamp]

--------
In message <05E65CAF-D064-4D4E-AA16-195FE7D15A73 at noao.edu>, Rob Seaman writes:

>On the other hand, the one thing we can be sure about POSIX is
>that it will ultimately have a finite lifespan.  But a day on Earth
>(and on Mars and Pluto) will always be a synodic (mean solar) day,
>whatever decision is made at WRC-15.

You are wrong in both claims.

The fundamental problem in POSIX is the value of the integer 'time_t'
not in how it is represented in human form.  POSIX is therefore
perfectly capable of handling any calendarial form of time you care
for, *including* UTC with leap-seconds, (with the footnote that
a few seconds suffer from an ambiguity of the same kind as the
"repeat-hour" during DST switchback).

On much of the planet, "a day" is not a "synodic (mean solar) day"
twice a year, when DST is started/stopped, and for any person
travelling from one timezone to another, the concept "day" has
nothing at all to do with the mean solar day.  So "always" ?
not even close...


-- 
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.