Error 3071 for dates after July 3 2008

That's got me stumped still. The [Date Monitored] field was null in two
records, where it should have been July 4 and July 12. No clue how it got to
be that way, since it's a required field exactly so that this doesn't happen,
and on top of being a required field, it's also got a default value that
should pre-populate. Never seen it before, probably won't ever see it again,
but I've learned my lesson about checking the source data if the design
appears to be fine.
--
-Ary


"David W. Fenton" wrote:

=?Utf-8?B?QXJ5?= wrote in
:

Just to finish up the thread, I finally got it figured out. There
was nothing wrong with the query or design, the problem was the
result of a corrupted record in the source data. All that smacking
my head against a desk and doubting my designs, and I never even
thought to check the data itself!

What kind of corruption? I so seldom see any form of corruption in
Access databases that I'd probably never consider this, either.

--
David W. Fenton http://www.dfenton.com/
usenet at dfenton dot com http://www.dfenton.com/DFA/

=?Utf-8?B?QXJ5?= wrote in
:

That's got me stumped still. The [Date Monitored] field was null
in two records, where it should have been July 4 and July 12. No
clue how it got to be that way, since it's a required field
exactly so that this doesn't happen, and on top of being a
required field, it's also got a default value that should
pre-populate. Never seen it before, probably won't ever see it
again, but I've learned my lesson about checking the source data
if the design appears to be fine.

Aha. That's not "corruption" in the sense of what you fix with a
compact and repair, but corruption of the data.

I can see how the default value can be gone, but not how a required
field could end up Null. Weird.

--
David W. Fenton http://www.dfenton.com/
usenet at dfenton dot com http://www.dfenton.com/DFA/