Re: Question about schema

3) The usual way to do this is to have a names table, an addresses table and an intersection table for names and addresses. But, as they say, circumstances vary. It partly depends on what kinds of queries you'll be running, how often names and/or addresses change or get deleted, what your hardware looks like, etc. Don't think you'll find too many people suggesting a million tables to handle addresses.

Van

"Clark Snowdall" <clark_at_e247.com> wrote in message news:3A170759.8DA61243_at_e247.com...
> I have a quick question to pose regarding database theory. Imagine if
> you will a company that needs to keep track of 1 million users' address
> books. Is it best to:
>
> 1) Have a master table listing all the users' principle information
> (succh as username, password, real name, address, preferences, etc.),
> and then also have another table listing all of their address entries.
> This means of course that if each of the 1 million users have 10 entries
> in their address book the second of the two tables can reach 10 million
> records.
>
> OR 2) Again have a master table listing all the users' principle
> information, but instead have a new table for each user that only lists
> the address book entries for that user. This means of course that there
> would be 1 million tables, each with 10 entries each.
>
> Of these two scenarios, which would the most efficient for searching,
> which would be easiest for updating, and which would be the easiest to
> administer (providing that the application can perform the simple
> administrative tasks such as creation of deletion of these tables).
>
> Thanks for any info on this issue,
> Clark
>
>
Received on Mon Nov 20 2000 - 01:17:58 CET