If you're of the opinion that database design via forum is unwise, then
perhaps your best bet is to unsubscribe from a forum where that is the
purpose.
While people like you are giving advice on database design I hope that many
will point out your amateur errors and profound lack of knowledge. I would
have though with such a profound lack of knowledge you may wish to educate
yourself, perhaps, and sadly, not.
Learn from good "non-product" specific books about relational database
design not from dodgy websites. Check the points I raised with you in these
books and see if I am wrong.
You seemed more concerned with my motives, well my motives are transparent I
wanted to correct serious errors in the advice you were giving. I will be
around from time to time and will happily criticise points that are just
plain wrong. I gave you the benefit of the doubt on my first post, I have no
doubts about you at all now.
Relational Data Analysis requires a thorough understanding of the problem
domain not some person making wild assumptions like you have done on this
thread. Database Design by email is dangerous because you have to ensure
that the Client is disclosing all the relevant information, they may not
know that they are not. The collection of sources for the RDA process will
throw up a lot of questions that the skilled developer will ask of the
Client.
You do not have appear to have the necessary rigour to be advising people
about database designs. Throughout this thread you have failed to address a
series of salient points, I had hoped you had accepted them, but it would
appear you are not equipped to argue your point not that it would have made
sense anyway as you are wrong, and I only looked at one table and you sadly
confirmed that you would have taken no steps to void data duplication and
cared little for the data integrity and consistency of the database.
Your failings can be corrected with a little bit of education, I hope you
will consider this.
As I said at the beginning I do not want to get into the Surrogate vs
Natural Key debate, but you will find that most of the people who do
advocate the use of Surrogate Keys (in all tables) will also ensure that the
Natural Key (when available) is implemented as a Unique Non-Null Index.