hmmm...
I realized I'm probably approaching the whole lot the wrong way...
After starting a new thread about some basic design problems and reading some other posts in this forum I came to the conclusion the stuff discussed here is still going way over my head. However I'm becomming a bit fed up creating databases using the trial and error approach creating more errors on a long term.
After reading the "What Data Modelers Need to Learn First" thread it turned out I don't even know most of the terms used, I just created the db's the way I saw fit at the moment but no db has turned out being "perfect" which was to be expected
I found a web site listing the "5 rules of data normalization" (http://www.datamodel.org/NormalizationRules.html) which is a good thing to keep in mind but I was unable to find more basic information. What the hack, I might as well start from the beginning again the proper way

What about starting a thread to collect some good internet information sources about how to realy start designing a db?
What do people have to know first, second, etc. ?
What would be good approaches the look at a problem, etc. ?

[QUOTE][SIZE=1]Originally posted by Prodoc
>hmmm...

>What about starting a thread to collect some good internet information >sources about how to realy start designing a db?

>What would be good approaches the look at a problem, etc. ?
An approach that you might consider would be to find a Kick-Start Data Model that is similar to what you need and then adapt and extend it as appropriate.
This approach should help you learn how to create your own using a 'Design Pattern' method.

Here is page listing over 200 Kick-Start Models to get you started :-
http://www.databaseanswers.com/data_models/index.htm

Barry Williams
Principal Consultant
Database Answers