An even ten-thousand

We note an almost unknown milestone.

Last Friday was Julian Day number 2,460,000.  The last time we hit an even ten-thousand was, of course, 2,450,000, a day more familiar as October 9, 1995.  They don’t come along often, being a little less than thirty years apart; so, for instance, none of our tutor’s students was born back then.

Julian Days were invented at the last great reform of the calendar, in the sixteenth century, as a way to avoid the irregularities and complications that come with any calendar.  The motivation was initially historical.  Working out a proper, or even consistent, chronology based on kings’ reigns is difficult, even when the historical material is available and accurate.  Reconciling the various calendars in use in different places and at different times is even harder.  So Joseph Scaliger picked a day before any likely history (in 4713 BC), when three cycles of importance to calendars coincided, and counted day by day forward from there.  With the proper tables, or some arithmetic, events can be placed properly in time.

They are now used mostly by astronomers.  Astronomers need to work out times and time intervals that do not respect calendars or even events on Earth, and it’s not always an easy task.  A large section of the Almagest of Ptolemy, the master work of ancient science, is given over to working out a chronology based on Egyptian and Babylonian rulers.  A table of Julian Days (if it had been available to him) would have made his task much easier.  An astronomer can now work out when to observe his binary star that has a period of 25.46 days. And you can easily find which calendar day is thirty or forty-five or one hundred days from this Thursday.  It’s much harder working with months and years of differing numbers of days.  (There is probably a Julian Day app to make it even easier.)

So are astronomers celebrating a ten-thousand day holiday?  Not really, even those who work with the system manually and so are aware of it.  In a way, it would be contrary to the idea of Julian Days, which is to discard all those multiples and groupings of days and simply do arithmetic.  But people like to notice round numbers, so there were messages going around celebrating the event.  Of course it had nowhere near the effect of, say, the year 2000, and both numbers are in a large sense arbitrary.

A number that may have a much greater effect, however, is coming up.  Our astronomer, trying to be clever, sent a colleague greetings for the New Year 1111110000.  That’s 2016 in binary.  It was a pure coincidence that the pattern turned up.  It will appear again in 2032, and increasingly often thereafter, until 2047, when all the register is full: ten 1’s.  For 2048 we’ll need another digit.  That would be an arcane observation, except for the fact that so much of our world is digital.  Computers store numbers in binary.  Up until 2048, ten digits suffice.  How much of our software will crash on New Year’s Day of that year?