Note to Editors (Re "glitch")

[$ bill]

For your information. I have surveyed 37 years of daily results. The glitch is only averaging about 1 incident per year. It is NOT widespread. surveyed time frames completed, so far Current to 9-6-1984, 11-20-1976 to 10-18-1971, 5-09-1944 to 12-22-1938, 12-29-1906 to 10-4-1897. (Entries have been made after I looked at some of the above dates). I don't believe I've missed many.

[John]

I think it is a rare occurance and should be even less frequent since the last update to the code. The cure is in a record locking technique so that no two processes can access the same data, but that is for the distant future.