Since creating this web site, I have grown much less concerned about two important issues.
First, if we do miss the 2006 January 1 target, it is not a big deal. All that needs to be done is to drop 2006 December 31 (that is, rename that day 2007 January 1), and all is well. Similarly, a two-year delay could be easily accommodated by dropping 2007 Dec 30 AND 2007 Dec 31. (The new "last day" of the old year could be celebrated as New Year's Eve.) So, the argument that there is not sufficient time to implement CCC&T just does not hold water.
Second, I am now not at all sure that the switch would be very costly at all. Having thought a bit about it, it seems to me that the software changes are pretty minor. (I'd like to get email from some of you who actually carried out Y2K changes to see if you agree with me.)
In particular, all calendar software must include statements resembling these FORTRAN statements:
If the above is simply edited to read:
then the software will work PERFECTLY until June of 2009. Notice that the ONLY changes were in the numerical values in the data statements. These trivial software changes would take place at any time during the month of 2005 December.
That still leaves the software adjustment for the Newton Years. Jess Cully has suggested that we drop the concept of Newton Month, and instead simply have June be 38 days long in "Newton Years." If Jess's suggestion is adopted (I certainly have no objection) the additional software change to accommodate Newton Years will clearly be slightly easier, but anyway, we have eight years from now (2004 February) to implement them. So I really think the problem is entirely a matter of selling the idea.
To me, it is now an embarrassment to look inside the back page of the Observer's Handbook of the Royal Astronomical Society of Canada and see TWO calendars displayed. Why are we doing this to ourselves? There is NO reason for it!