Re: Which normal form is this violating?
Date: 25 Apr 2002 07:10:16 -0700
Message-ID: <1c37187b.0204250610.53a06606_at_posting.google.com>
dba_222_at_yahoo.com (Roger Redford) wrote in message news:<a8c29269.0204242031.9d9964f_at_posting.google.com>...
> Hello DB Design experts, 
> 
> I'm having the usual disputes about database design issues. 
> 
> The information that my coworkers have is say, x and y.
> It has a one to one relationship.  Therefore, it 
> goes into one and the same table. 
> 
> Table_A
> Fieldx	(pk)
> Fieldy
> 
> 
> However, they are arguing that it goes into another table.  
> 
> Table_A
> Fieldx	(pk)
> 
> Table_B
> Fieldx	(pk)
> Fieldy  (not null)
> 
> (Actually, thye have "designed" a number of strange tables, 
> and then put views on top of them, to come back to the same one to 
> one relationship.  Very strange and complex. )
> 
> What normal form does this violate?  It isn't 1st, 
> 2nd, or 3rd.  Boyce-Codd maybe?  The crazy thing about
> the design texts, is that they rarely cover mistakes
> in design.  This is a common one.
> 
> Thanks
If Fieldx has many attributes, it might make sense to partition the attributes across two tables. If there are only a few attributes, they should stay in the same table. What rationale have your cow-orkers given for partitioning the data as they propose? Received on Thu Apr 25 2002 - 16:10:16 CEST
