Re: Is mysql a RDBMS ?

"Mikito Harakiri" <mikharakiri_at_ywho.com> wrote in message news:dZ43b.18$84.104_at_news.oracle.com...
>
> "Leandro Guimarães Faria Corsetti Dutra" <lgcdutra_at_terra.com.br> wrote in
> message news:pan.2003.08.27.12.27.34.16318_at_terra.com.br...
> > On Mon, 25 Aug 2003 18:55:13 +0000, Anith Sen wrote:
> >
> > >> As if view updates problem is solved in the pure relational theory.
<<
> > >
> > > Who said so? This is exactly what I mentioned as being explored & lack
> of
> > > well-defined theoretical basis.
> >
> > AFAIU, D&D have actually provided a nice updateability model in _TTM_...
>
> No. Any practical implementation of relational model must admit functional
> operators. Math operators including: sin, exp, ln, etc. And with operators
> view updates is just insanely complex. Simple example:
>
> select Radius*sin(Angle) as Y, Radius*cos(Angle) as X from PolarPoint
>
> is perfectly updateable view, but the RDBMS must be aware of a lot of math
> to be able to invert it. Well in this example the math is trivial
>
> select X*X+Y*Y as Radius, arctan(Y/X) as Angle from CartesianPoint
>
> but it easily becomes graduate level math when column expressions are
> polynomials. And those are special cases only!
>
> As you see almost none of the relational operators is used in that view
(no
> join, no union, no restrict), so you simply have no way to leverage D&D
> ad-hock updateability rules.

I find it interesting that you chose the example you chose. Ironically, the example clearly demonstrates how D&D's proposed type system helps with view updatability. Obviously, the type system must have some efficient method to convert from each declared possible representation to the supported physical representations and vice versa. If one considers Point with both polar and cartesian possible representations, then the problem you mention above simply disappears. Received on Thu Aug 28 2003 - 05:45:42 CEST