Is Time Constant?

Okay so this might be far too technical and of no interest what so ever to many readers of this blog, but hey… I’m going to write it anyway.

As you will probably already know 2005 was a second longer than 2004 and a second longer than 2006 will be. At the stroke of midnight on the 31st of December a second was added to co-ordinated universal time (UTC). This is not the first time that a leap second has been added to a year, actually since 1972 there have now been 23 seconds added to time.

These seconds have been added in order to keep UTC in step with solar time or the earths rotation. Time has to be added due to the slowing of the earths rotation due to tidal friction and other planetary effects.

This raises two questions in my mind… The first being, the fundamental building block of all scientific understanding is that time is a constant. Is this really the case? If one takes solar time as being the constant then the slowing of a planets rotation indeed shows that time is not actually a constant but slows over time.

The other question it raises in my mind as a programmer is how computers deal with time. In programming we calculate times in seconds since the epoch, in the case of UNIX this is Jan 1st 1970 00:00:00 GMT

Since there have been 23 leap seconds since 1972 and a computer calculates 86400 seconds per day (60x60x24) this must mean that these calculations are now at least 23 seconds out. This of course does not matter for many applications as they are not time critical, and of course this is not a hard thing to correct, knowing how much time has been added or synchronizing with time servers, but I wonder how many time critical applications are now 23 seconds out. Yet one more thing for us poor programmers to remember!

Is Time Constant?