home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.databases.theory
- Path: sparky!uunet!ide!biesty
- From: biesty@ide.com (Bill Biesty)
- Subject: Re: Normalization and Relations
- Message-ID: <1992Sep9.204725.10955@ide.com>
- Sender: usenet@ide.com
- Organization: IDE, San Francisco
- References: <1992Sep9.061732.10792@constellation.ecn.uoknor.edu>
- Date: Wed, 9 Sep 92 20:47:25 GMT
- Lines: 33
-
- In article <1992Sep9.061732.10792@constellation.ecn.uoknor.edu> cdblueth@essex.ecn.uoknor.edu (Chris D Bluethman) writes:
- >What are the merits of not putting the zip code attribute
- >and a state attribute in the same table? Say an address table.
- >If you have a relation with last name, street address, phone,
- >state, and zip, is it really in thrid normal form? What about
- >placing zip in this table, and then having state somewhere else?
- >After all, the zip code does determine the state.
-
- What your leaving out of the picture is that normalization is
- not the end all of data modeling. The Information area is very
- important, and generally not available in sytactical form.
-
- So while it is quite possible to normalize the zip code into another
- entity with an attribute state, itis rarely done because such
- information is not usually interesting. However if you were the
- US Post Office, you would probably do it and have additional
- attributes like station name, street address, zip code :-)
-
- The textbook example is color. If you're buying a car, color
- is an attribute. If you're manufacturing a car, color is one
- or more entities that you use to track, hue, ingredients,
- paint manufacturer, number of coats, drying temperature, etc.
-
- Another reason is data stability. The PO changes zip codes, states
- and addresses remain much more stable. I have had trouble getting
- things mailed to me from mail order places with zip code cross
- checking (so they can get the best postal rate) when my zip code
- changed. Storing the state, as an override say, is much more
- robust that relying on an incomplete code table.
-
- --
- Bill Biesty <biestie@aol.com> (I don't read this too often)
- <biesty@ide.com> These are not the opinions of my employer.
-