Re: foundations of relational theory? - some references for the truly starving

Craig Bennett wrote:
>>The basis in facts are very simple: Nested Relations offer no extra
>>expressive power, and plenty of complications.
>
> I agree they add no expressive power: converting a 3NF to a 1NF structure
> will give you the same expressive power. Could you expand on your "plenty of
> complications".
>

Can you refer me to a theory of normalization, constraints, dependency and query optimization ?

According to Lewis Bernstein and Kifer textbook, everything gets complicated.

> The pick model allows the "clustering of precomputed joins" by the
> programmer.

So what's the avdantage of doing this at the logical level rather than doing it as part of the DBAs task to define physical layout of tables.

> And modern variants (like UniVerse which is the one I work on).
> Allow you to use SQL queries against this clustered and precomputed data (it
> can also use other pick style languages and methodologies but that is beside
> this point).
>

Are they relationally complete ?
Do they support recursive queries or transitive closure ?

> UniVerse also allows you to apply strict typing to columns as per SQL 92
> (although this is an extension of Pick)..
>
>
>>It is you who should give me a good reference -- not marketing idiocies
>>about how "joins are difficult to compute" --, that you consider
>>relevant for showing the difference. And we'll see.
>
> Perhaps Mike mean't that joins can be *EXPENSIVE* rather than difficult to
> compute.
>
>

Or joins can equally be *INEXPENSIVE* to compute.

> Craig
>
>

Costin Received on Wed Oct 22 2003 - 06:00:53 CEST