Re: How to normalize this?
Date: Tue, 7 May 2013 00:51:01 +0200
Message-ID: <518833d5$0$6355$e4fe514c_at_dreader35.news.xs4all.nl>
On 2013-05-06 21:44:04 +0000, compdb_at_hotmail.com said:
> On Monday, May 6, 2013 8:18:38 AM UTC-7, Jan Hidders wrote:
>> On 2013-05-06 12:26:10 +0000, Erwin said:
>
>>> How can deliberate admission of "information differences" coexist with>
>>> > "aims of being lossless" ?
>
> It is called a "lossless" decomposition because if the components are
> projections of the original then they join to the original. Nb this
> assumes certain predicates for the components in terms of the original.
> (It has nothing per se to do with other predicates.) Nb this does not
> deal with appropriate constraints on the components or the original.
>
> Compare to additionally being "dependency-preserving" where if given
> dependencies hold in the components then they hold in the original. So
> enforcing them in the components enforces them in the join. Although
> moving from 3nf to bcnf is not dependency-preserving, one can introduce
> suitable inter-component constraints. Nb other constraints are not
> dealt with by normalization.
>
>> Well, "lossless" means nothing is lost, but, yes, something is allowed> >> to be added, i.e., the new schema might be able to represent> >> information that the old one could not.
>
> Normalization does not "allow" this. What happens is people normalize,
> then they notice they really wanted a schema allowing more states (ie
> predicates that the normalization-generated predicates imply), which
> they didn't need to normalize to find, then they change to new the
> predicates for the components, giving a new predicate for their join,
> BUT in such a way that just happens to leave the new components
> properly normalizing the new join. They confuse this process with
> normalization and do not understand normalization's role in it.
I'm afraid you are losing me here. You seem to be unilaterally redefining the notion of normalization and then accusing other people of misunderstanding what they are doing just because they are using a different definition. That's not going to lead to a productive discussion.
- Jan Hidders