RE: Elsets Epochs and Y2K

Ted Molczan (molczan@home.com)
Thu, 25 Feb 1999 18:12:43 -0500

Rob Matson wrote:

> The Y2K bug won't affect SkyMap, assuming Space Command
> sticks to 00001.0 for 1 Jan 2000.  To avoid confusion in dates,
> at least for the next 58 years, SkyMap assumes any two-digit
> date less than 57 is 20xx.  Otherwise it's 19xx.  (After all, no
> satellites existed prior to 1957).

A few years ago, I made the exact same Y2K fix in my satellite ephemeris
generator, but I created a new bug in the process. I overlooked the date
processor in the subroutine that enables selection between Epoch 1950 or
2000 co-ordinates. I did not discover the problem for more than one year,
when I found large positional discrepancies while attempting to analyse some
of Russell Eberst's observations, which are Epoch 1950.

It took me a long while to realize that my Y2K fix had caused Epoch 1950 to
become Epoch 2050! Of course, the fix was easy enough - just shift the
earliest year in this century from 1957 to 1950.

Interesting, how we are using the same rationalization for these quick fixes
that got us into this mess in the first place! I wonder how many date bugs
will suddenly surface around 2050? There was a Y2K article in a recent issue
of Scientific American, that discussed this very problem.

Ted Molczan