Re: Issues subtracting Unix epoch from date
Re: Issues subtracting Unix epoch from date
- Subject: Re: Issues subtracting Unix epoch from date
- From: "Mark J. Reed" <email@hidden>
- Date: Sat, 20 Feb 2010 21:52:03 -0500
On Sat, Feb 20, 2010 at 8:01 PM, Doug McNutt <email@hidden> wrote:
> <ducking>
No need for that. You won't find any objections here.
> Sometime during WWII the US Army gave us Zulu time, now known as GMT or perhaps other acronyms that include atomic time and leap seconds.
GMT has been GMT since at least the 19th century; I don't know exactly
when the US military adopted the alphabet time zone designations, but
Z for Universal Time is the only one left these days.
The problem with "GMT" is that exactly what it means has varied over
time, which is why they officially deprecated the acronym in favor of
UT ("Universal Time").
There are various flavors of UT. UT0, UT1, and UT2 are time based on
various measurements of the actual rotation of the earth, with
increasing corrections for the vagaries of that motion and those
measurements. But the Earth's rotation is far from a constant speed,
so in UTn, the length of "a second" varies with the speed of that
rotation. What we actually use for civil timekeeping is UTC, which,
unlike the other UTn timescales, ticks at a constant rate defined by
atomic clocks. Since the Earth is a worse clock than atomic clocks,
UTC periodically has extra seconds inserted to keep it within a second
of the other varieties of UT. So UTC is the only one with leap
seconds. (There's also TAI, which is atomic time ticking along
constantly, unheeding of the Earth's rotation; it agreed with UT as of
January 1st 1958, but is now 34 seconds ahead of UTC, a discrepancy
which grows every time there's a forward leap second.)
> Astronomers, I am one, have been using Julian date/time since Hector was a pup.
Now, now, no need to exaggerate. :) The Julian calendar only came
along in 46 BC, and astronomers haven't used it that long; for
centuries they kept using the Egyptian calendar with its unchanging
365-day year - not having leap years to worry about made the math
easier. Scaliger's Julian Period dates to 1583, but the use of it to
form an absolute day count (the JD number) was an innovation due to
Herschel in the 19th century.
> UNIX has been using Zulu time when it time stamps a file on a disk since UNIX existed.
Quite the sensible decision on the part of its designers. Sadly
Windows and the original Mac didn't follow suit.
> The very idea that modern Computer Science (capitalized to indicate it's acceptance as a science) even tries to do anything else but keep Zulu time is just silly. The proper
> solution is for computer science to declare that it knows ONLY Zulu and that users had better decide that they accept it.
Computer Science has little to do with the practicalities of
timekeeping or running an operating system, I'm afraid.
I personally don't understand this obsessive need to keep the clock
and calendar in agreement with the sun and the seasons. Europeans
gave up on keeping months in synch with the phases of the moon over
two millennia ago, and their world didn't dissolve into chaos. Would
it be so terrible if the seasons moved around the calendar? It
doesn't seem to bother the Muslims much, and their calendar goes all
the way around the tropical year in only about 30 years. By
comparison, the "outmoded", "inaccurate" Julian calendar only moves
about 3 days every 400 years with respect to the seasons. The horror!
In a similar vein, even if we agree that UTC should stay in synch with
the Earth's rotation, I think we could allow a little more than one
second's leeway, given that we operate on mean time across hour-wide
time zones. More, in some cases; Atlanta is in US/Eastern but is far
enough west that we ought to be in Central. True noon doesn't happen
until 12:45pm, but hey, at least we know that mean noon in Greenwich
occurs at 7AM (8AM in the summer) on the dot, not a second one way or
the other!
--
Mark J. Reed <email@hidden>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
AppleScript-Users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
Archives: http://lists.apple.com/archives/applescript-users
This email sent to email@hidden