On Tuesday 10 Jan 2006 9:51 am, Mike wrote: > but necessary enough. The point I was making was, what's the > difference between a primary key vs. any other unique field?
database design mandates that every row in a table must be unique. Where there is a serial data type as the primary key, this holds good even if every row was identical except for the serial id field. This would be bad design. Therefore, for every table one has to identify one row, or a combination of rows that has to be unique. There are two approaches to this problem - one is to make the unique field or combination of fields the primary key, and specify the id field to be unique. The other way is the opposite. And there is a never ending battle between proponents of 'natural' primary keys and 'surrogate' primary keys. -- regards kg http://www.livejournal.com/users/lawgon tally ho! http://avsap.org.in ಇಂಡ್ಲಿನಕ್ಸ வாழ்க!
