[thelist] DB design question

[Ed McCarroll]

At last, a thread I can understand well enough to argue foolishly
about...

> >Point taken. Now, tell me how many people know their "11-digit ZIP"?
> >Nobody. "ZIP" defines 5 digits, "ZIP+4" is the consumer-use one.
>
> No one does, but Mike did mention that the USPS makes the appropriate
> conversion services available for a fee. Something to think about if you
> had a budget and wanted to take advantage of the 11-digit zip.

11-digit ZIPs would work, so long as no data entry errors are made, the
post office never changed anyone's 11-digit ZIP, and you never attempt
to list a property outside the US.  (I have only had them change my
5-digit ZIP once.)

When this sub-thread took off, the issue was what to use as a unique
identifier for a property.  I'm having trouble understanding why so
many folks here want to jump through complicated hoops just so they
can derive that unique identifier from the same info that the real
world uses to identify the property being represented.  Is there an
underlying assumption that an important purpose of the primary key
is for searching?  I understand why fields that might be searched
should usually be indexed, but what's to be gained by composing a
primary key such that someone who knows the postal address can (with
enough difficulty to invite frequent error) figure out the primary
key?  Why not simply search the address field?

> By the way, Mike H - thanks for posting that info. That was absolutely
> fascinating.

Me-too!

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Ed McCarroll                          MailTo:Ed at ComSimplicity.com
ComSimplicity                                      (310) 838-4330
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -