A question of time

My first Common Lisp programming job was for Mercury Scheduling (no longer around) working on the back end of airline crew control software. The main focus was assisting crew controllers in scheduling personnel on flights. We had to satisfy literally hundreds of constraints from union regulations, government regulations to…what else?…oh yeah, the constraints of reality like not being in two places at once or starting a new shift 1000 miles away in an hour’s time (computers are stupid, you have to tell them everything!).

One thing I learned is just how complicated time can be when you are inside a computer! Timezones, daylight savings, leap years, leap centuries, leap seconds….it’s a real mess.

Well, it turns out that leap years, or more generally the difference between the calendar and the actual orbit of the earth around the sun, cause a discernible albeit insignificant effect in the monthly temperature anomalies. The actual orbitally controlled seasons are not in sync with the conventional calendar.

The issues are interesting, and Tamino at Open Mind goes into them, starting out with a paper by Cerveny et al., in an impressive level of detail, have a read.

7 thoughts on “A question of time

  1. As a bit of trivia, what President’s name used to be embedded in the UNIX time code?

    A: way back, Nixon: there was a period when Nixon had moved Daylight Savings around, so there was a variable called “nixonflag”…

    Like

  2. If you look back over the course of the last few years, is there anything you would have done differently or is there anything you wished had happened but didn’t happen?

    Like

  3. A leap-year bug in Microsoft Zunes disabled them on the last day of 2008…

    A few months ago, around the change of Daylight Savings Time, someone we were meeting missed by an hour, because between him and his secretary, one used a Mac and one Windows, and when exchanging calendaring information, something went wrong.

    Like

  4. Yes, surprising bugs abound around DST issues…

    I was using an older version of postgresql and found that I got the wrong date when asking for CURRENT_DATE + ‘interval 1 week’ if a DST change occurred at any point in that interval! That was a bugger to figure out, and was never fixed just avoided through hackery.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s