Re: problem with historical data and referential integrity

I got the first hardcopy edition with the CD -- I hate having to download stuff in ZIP and tar files.

Rick was on ANSI X3H2 and was responsible for the temporal stuff we have. He proposed more, but then withdrew it when further extensions were becoming too complex for a general database language.

The sales on the book were not that good and neither were the sales Chris Date's temporal queries book. My TREES book did better in four months than either of them. You might want to steal both books from the library :)

In the case of Date's book, he invents his own database language with temporal extensions based on the "chronon" model of time.

Jan Hidders is more up to date than I am, but the two major schools of thought are "discrete points (chornons) in time that make sets to be treated as atomic values" and "the continuum model" using half-open intervals to be treated as atomic values. Date is a Chronon person and the majority of the temporal database research community are interval people.

The thing that gets weird with Date's book is that he uses one of his "make a set from these discrete values" operators on a set of salesmen's identifier numbers. Try as I might, I cannot mentally picture a continuum of salesmen. Aleph Null spam, perhaps.

He then made the political mistake of attacking some of the people in the temporal community on a website. This did not win friends. Received on Wed Dec 08 2004 - 05:09:11 CET