RE: [KCFusion] tuples, keys, and argumentative university academics!

[Ryan Hartwich]

Title: Message

Adaryl,   I have no idea from a 'book' sense.  And I would say that you COULD say that the masterID and userName are both unique, but heck, you could also SAY that the fName and/or the LName uniquely identify the record.  You could just as easily say the password is unique if you have determined that through either business logic or SQL uniqueness that the password is the key value.   While either side may be correct from a purely book sense, I think it is more important to consider what the use of the data will be.  There may be a benefit to separating the first and last names from the username and password.  U&P both indicate a login method, the first and last name are part of the individual's contact information and would probably make more sense to be stored with addresses, phone numbers, etc.  You would then use the masterID to tie the two tables together.   Ryan   -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Adaryl Wakefield Sent: Friday, September 20, 2002 4:40 PM To: [EMAIL PROTECTED] Subject: [KCFusion] Huge ruckus broke out in one of my project groups yesterday. It was over the definition of Boyce-Codd normal form. I know the definition is every every determinate is a candidate key. We were going round and round about what constitutes a determinate. I said if the table looked like this masterList(masterID, fName, LName, userName, password) we would be ok. Another guy said we need to break up the table because userName and masterID uniquely identify the tuple. I said that it only uniqly identifies the tuple if we say so. This went on for about half an hour with many valid points on both sides and wound up a jump ball. Anybody care to settle this bar fight? A.