Re: Relational Databases and Their Guts

"Paul Vernon" <paul.vernon_at_ukk.ibmm.comm> wrote in message news:bd9ao7$13qa$1_at_gazette.almaden.ibm.com...
> "Todd Bandrowsky" <anakin_at_unitedsoftworks.com> wrote in message
> news:af3d9224.0306231111.fd73ac6_at_posting.google.com...
> > > Yes. Empty sets. E.g an empty relation.
> >
> > Ooh, I like that.
> >
> > > View updating is another issue. I (happen to) disown the concept.
Views
> are
> > > relational expressions, they are not variables and hence they are not
> > > updatable. End of story.
> >
> > I like that even better!
> >
> > >
> > > View updatability is a (somewhat) wrong headed way to provide logical
data
> > > independence.
> >
> > And better still!
>
> :-)
>
> > Of all DBMS systems I've seen, only MS Access had really good and
> > consistent view updateability, and I believe that at one point MS
> > claimed that Access was a true RDBMS for that reason. In practical
> > terms, for C/S systems updateable views are nearly impossible to work
> > because of all the transaction and connection semantics.
> >
> > Now, what in your mind means logical data independence?
>
> Strictly I think of it as:
>
> The ability to interact with a database via any information equivalent
> database schema.

That ability is trivial. How does it provide independence? How does it allow one to change the base relations to a different information equivalent schema without requiring one to alter existing applications? Your statement above only requires that one can change the application to use the new schema, and I do not think that provides any independence at all. Or am I missing something?

> > I went to the link to look for the article, and I'm going to read it
> > at home tonight. I also went out and bought the 3rd Manifesto.
>
> Yehaa!
>
> > I'm
> > really interested in seeing what he has to say about mapping objects
> > to the relational model, in particular,
> > what to do with primary keys that are ranges.
>
> Ignoring the danger inherent in any OO discussion ;-), what do you mean by
> 'ranges'? Sets of an interval type? (E.g. a set of time periods ??!)

I think he will have to pick up the book on temporal databases to see what Date et al suggest for interval types. Received on Tue Jun 24 2003 - 22:50:50 CEST