Re: 1GB Tables as Classes, or Tables as Types, and all that refuted

"Alfredo Novoa" <alfredo_at_ncs.es> schrieb im Newsbeitrag news:41a6160a.17546031_at_news.wanadoo.es...
> On Wed, 24 Nov 2004 23:03:50 +0100, "Rene de Visser"
> <Rene_de_Visser_at_hotmail.com> wrote:
>
>>"Alfredo Novoa" <anovoa_at_ncs.es> schrieb im Newsbeitrag
>>news:qbv9q0trk6nrno44af8r854bue0essjgl9_at_4ax.com...
>>> On Wed, 24 Nov 2004 21:32:58 +0100, "Rene de Visser"
>>> <Rene_de_Visser_at_hotmail.com> wrote:
>>>>> This is true, but this is not the same as you said above.
>>>>In the relational programming language AP5, a type is simply a relation
>>>>of
>>>>arity 1.
>>>
>>> Then it is a poorly designed language.
>>
>>Could you expand on why you believe that was a poor design decision?
>
> Well, I shooted too fast.
>
> I supose that in AP5 a type is DESCRIBED by a 1-ary relation value
> (but not a relation variable).
Tuple components are defined by position in AP5 (as in the original Codd paper), and not by name as in his second paper. i.e. there are no globally valid relational variables. All relational variables are local to form that uses them.

>
> The problem is how to describe such relation value and how to define
> the operators of the type.

The relation value is often described algebraically using a modal logic extended with quanfiers and relations.
(Relations are of two basic types in AP5, defined or stored). In AP5 operaters are also relations.
See my other posting for how a plus operator can be defined relationally.

AP5 is mainly declarative with add and delete to add and remove tuples from stored relations.
The rest is done with triggers and relations, and functions when the relationally defined ones aren't enough.

>
> A relation is not a type but the set of values of a type can be viewed
> as a 1-ary relation value.
i.e.
values of a type <=> 1-ary relation value

type                 <=> relation / relation definition / relation name

in AP5 relations are second class objects. i.e. there is a 1-1 relationship between relation / relation defintion and relation name. Received on Thu Nov 25 2004 - 21:25:51 CET