Internetworkers: http://www.ibiblio.org/internetworkers/

["Michael D. Thomas" <mdthomas AT mindspring.com>]

8:02 p.m. today will be an exciting minute!

In a world wrought with irregularity, that minute will be symmetrical -
20:02, 11-11, 2002.

But it's not just an exception to our usually asymmetrical world, it's
symmetrical for most of the world. Not only will the minute be
symmetrical in the U.S., but will also be symmetrical in the British/European
format (20:02 11-11 2002) and in the ISO-8601/Chinese formats (2002-11-11-
20:02).

Such a minute must be cherished. I know, it won't have the pristine,
diamond-like, 12-fold symmetry of 11:11, 11-11 1111. Still, you won't see
another minute quite like this one until 21:12, 11-11, 2112. After that, there
won't be many more such minutes. The clock is ticking.

As usual, today's symmetrical minute will get all the glory, but it's
really the day that should be.... well, taking the day.  First, 11-11 is
convenient for typists. It's one of only 12 dates that may be typed with one
number key. But it stands out from that assemblage of singularities - it's the
only one of the gang that always requires 4 digits. Give those other dates
their padded zeroes (03-03, 01-11) and they aren't symmetrical. 11-11 always
is. 12-21 almost makes the mark, until you invalidate it by writing it 21-12.
(Non-Americans should transpose the dates in the previous argument.)

 There are other symmetrical moments on other days, but they just don't have
the same global impact. Since so many date-time renderings have the date and
month in the middle, the most interesting minutes occur on 11-11 - the most
pervasively symmetrical date.

With the middle token guaranteed symmetrical, 11-11 is the day for pervasive
symmetrical minutes. The next limiting factor of finding ISO-compliant (and
American, European and Chinese complaint) symmetry on November 11: the
palindrome of the last two year digits aren't guaranteed to make valid hours.
However, all inverses of valid hours are valid years
(23<-->32, 00<-->00, 22<-->22, 02<-->20, 21<-->12, etc.). Hence, each century
has 24 years where Nov. 11 has a chance at a symmetrical minute.

In the eternal search for symmetry, this only leaves one more limiting
factor. Watch out, it's the most limiting of all. Each hour has 60 minutes
whose inverses are valid century representations. But the 10 minutes ending in
zero present a padding problem like we saw with months and dates. For
instance, the minute :10 inverses to a valid century only if you write it
"01", such as: "In 0127, Ptolemy began to formulate his geocentric theory of
planetary motion that prevailed for 1400 years."

No one writes dates like that, so we really have to eliminate those 10 minutes
and their corresponding centuries from our study. But no need to dwell on
those centuries --  they all preceded the glory days of the Byzantine Empire.
For the purposes of this discussion, those cases are ancient history.

Since we're now beyond the second millennium, all of those minutes ending in 1
have also seen their mirror century pass. So at the beginning of this year,
the minutes ending in 2,3,4,5,6,7,8 and 9 inverse in to valid centuries, which
leaves 48 centuries that will have symmetrical minutes on Nov. 11. As luck
would have it, our current century is one of those. But enjoy them while they
last -- we're in for a big dry spell after 23:52 11-11, 2532. There will be no
symmetries between that minute and the year 3000.

Date format doesn't matter when talking about the outer tokens. An ISO-8601
date-time like:

7432-11-11-23:47

is still symmetrical in the American and European format:

23:47 11-11 7432

But they look different -- specifically, they aren't both composed of the same
sequence of digit characters. This is both an aesthetic and a semantic
problem: changing the date format can vary the minute by thousands of years.
If you want the strings to look the same regardless of date format, you have
to restrict your space to palindrome years and times. Following the rule
developed above, you just look for the times that are palindromes since years
have to be valid times. It's a small set, especially when you don't want
zero-padding problems: 10:01, 11:11, 12:21, 13:31, 14:41, 15:51, 20:02, 21:12,
22:22, 23:32, 24:42 and 25:52.

This is another reason today is special: the year is a palindrome and
represents a valid time. Because of this, it is written the same regardless of
the various formats discussed. After today, there are only 5 such years -- and
thus, minutes -- left.

 Alas, after 23:59 11-11, 9532, November 11th looses its significance
altogether regardless of date format issues.  The possibly palindromic dates
1-1, 01-10, 2-2, 02-20, 3-3, 03-30, 4-4, 5-5, 6-6, 7-7, 8-8, 9-9, 10-01 and
12-21 loose significance, too. After that year the next batch of symmetrical
dates will look like 00:00 11-01, 10000 - but Americans will celebrate that
symmetry many months after it is celebrated in Britain. No longer will
date-time symmetries be such a global event and the formulas for deriving
date-time symmetry will become much more complicated. The length of the dates
will then be odd and the palindrome will then revolve around the tens-digit of
either the month or date, depending on your flavor of date format.

Sure, 11:11 11-11, 11111 will be an amazing celebration. But I'm just as sure
that, as that minute draws to a close, the partiers will reminisce about days
like today. "Symmetry was so much simpler then," they'll say.

But by then we (if there is a we) will probably have moved on to something
else with its own possibilities. As when Pope Gregory XIII decreed that 10
days must be skipped in October, 1582 to handle an 11 minute error in the old
Julian calendar, sometimes even time must change.