Re: Object-oriented thinking in SQL context?

On Jun 10, 10:48 am, paul c <toledobythe..._at_oohay.ac> wrote:
> David BL wrote:
> > On Jun 10, 10:10 am, paul c <toledobythe..._at_oohay.ac> wrote:
> >> David BL wrote:
> >>> On Jun 10, 9:44 am, paul c <toledobythe..._at_oohay.ac> wrote:
> >>>> David BL wrote:
> >>>> ...
> >>>>> This is clearly intended to mock, and in a nasty way. It is typical of
> >>>>> BB. Perhaps BB could himself do with some suggestions for reading
> >>>>> material - books with titles like "How to win friends and influence
> >>>>> people".
> >>>>> Of course Reinier was also mocking various regulars in the newsgroup.
> >>>>> However, I found his comment quite amusing - and nothing like BB's
> >>>>> aggressive tone. BB's reaction also suggests lack of balance.
> >>>> Maybe you should count yourself lucky that you didn't get called for
> >>>> suggesting a paradox could ever be a foundation, aka DVA's!
> >>> What paradox would that be?
> >> Russell's.
>
> > I know Russel's paradox. What has that got to do with it?
>
> What I suggest you do is look for his Intro. to Math. Philo., (it might
> help you if you were to concentrate), a pretty skinny book compared to
> most in the db field. Unlike his other theoretical stuff, it was aimed
> at the the layman, in particular, various self-taught technologists and
> interested readers of the time who had been introduced such theory and
> spent their Monday evenings at the local scientific society meetings in
> pre-WWI Britain, albeit ones whose first language was English. It may
> not tell you in five words why DVA's are paradoxes, and it may not tell
> most readers why "DA's" (not data administrators, ha, ha), aren't
> paradoxes, but it may give you some practice that would help to see why
> values that refer to the context that refers to them are a distortion
> of the recursion idea. I'd guess you could read parts of it for free on
> google books.

I never suggested that a value would refer to the context that refers to it. Definately not. Why do you think that?

There is nothing self-contradictory about a relation type that uses a database type as the domain type for one of the attributes, meaning that values of that domain type are sets of named relation values.

In the same way that one can static type relations by the names and types of the attributes, one can static type databases by the names and types of the relations. This would tend to avoid recursive type definitions when using DVAs - not that recursive types should be regarded as a problem in general.

> It pisses me off that various publisers like Cambridge
> Univ. help to pay their useless bureaucrats by preserving copyrights of
> dead people. On the other hand, this keeps those off the welfare rolls.
> It is perhaps the the first book to anticipate Codd, maybe by more
> than fifty years.
Received on Wed Jun 10 2009 - 05:50:27 CEST