Boolean bug?

[Robert Jacobs 0]

Who could resist the phrase Boolean bug? I may have discovered one. New York City births from 1901 to 1907 are indexed at http://www.germangenealogygroup.com/NYCBirths.stm I tried to devise a filter which would find all 1901-1907 NYC births in my data:

 

(Birth/date does not come before 1901 AND

Birth/date does not come after 1907) AND

([OR terms which select for the five boroughs])

END

 

This seemed to work fine until I realized that my Aunt Sonia, born in 1907, is incorrectly excluded. OK; I changed the second term to

 

Birth/date comes before 1908

 

Now Sonia is in. So it looks as though (as the symbol <= suggests) the plain-language term should be "does not equal or come after" That's fine with me, so I tested it against "does not come before" (>=) and it works the other way. 1901 births ARE included.

 

The two terms should work the same way, and I think the meaning should be "does not equal or come after(or before)" and the language of the operator phrase changed to reflect that.

 

I hope I've explained this clearly and haven't erred, either before or after.

 

 

Robert

[Jim Byram 0]

>= is greater than or equal

 

I suspect that the operators are coming from VFP so they work like they work.

 

You might try an inclusive filter rather than an exclusive filter. And Sheila makes a good point.

[altenbernd 0]

Try putting in a full date (ie Birth/date does not come after 12/31/1907). I suspect the month and date might be defaulting to 01/01 if not entered.

 

Sheila Altenbernd

[Robert Jacobs 0]

Try putting in a full date (ie Birth/date does not come after 12/31/1907). I suspect the month and date might be defaulting to 01/01 if not entered.

 

Sheila Altenbernd

 

That's got it, Sheila -- vy good!

 

Jim -- I guess we can't add this one to the grumbles about VFP. Inclusive worked, by the way.

 

Good wishes to all.

 

 

Robert

Edited July 1, 2010 by Robert Jacobs