Re: Generalised approach to storing address details

From: Sampo Syreeni <decoy_at_iki.fi>
Date: Tue, 12 Dec 2006 01:39:12 +0200
Message-ID: <Pine.SOL.4.62.0612120123530.6483_at_kruuna.helsinki.fi>


On 2006-12-11, paul c wrote:

> I would like to know how the RM handles hierarchies, without the aid
> of a builtin such as TTM's TCLOSE that is essentially outside the
> scope of the RM (eg., it seems to me that it does a transformation
> that can't be couched in fundamental RM terms.)

The classical, Codd-style, RM does not consider transitive closures/hierarchy/recursion at all. Later thoughts by the pioneers do, but they mostly limit themselves to single table/relation recursion. (IOW tail recursion.) RM/V2 is the clearest example to me, and I seem to remember it draws from RM/T, there.

Many of the other researchers in the seventies suggested the same or more, in connection with the universal relation model, the translations from RM to hierarhical/network, and later on the many extensions to the relational model like the nested relational one(s). At one point I think even Codd considered a relational interface with sets of relations because of this precise application scenario. Such considerations have made it in some form even till today, for example in the case of second order dependencies used for data interoperability.

But historically it all didn't take, mostly because Codd's original vision of the relational model was mostly correct: sometimes you might need this sort of power, but mostly what you wanted was propositional logic calculus with huge extensions and very little intensional capability. What was left over was easily filled in by external, procedural tools and practice, which already existed as legacy from the earlier, more primitive times. The Bill of Materials problem did have to be solved in principle and provisions had to be made for it in the later relational models and languages, but that's as far as it got; in practice relational interfaces didn't consider the matter until the nineties.

-- 
Sampo Syreeni, aka decoy - mailto:decoy_at_iki.fi, tel:+358-50-5756111
student/math+cs/helsinki university, http://www.iki.fi/~decoy/front
openpgp: 050985C2/025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
Received on Tue Dec 12 2006 - 00:39:12 CET

Original text of this message