Re: <OR> predicate?

On Sep 27, 6:27 pm, paul c <toledobythe..._at_oohay.ac> wrote:
> On 27/09/2010 1:26 PM, Vadim Tropashko wrote:
>
>
>
>
>
> > Familiarity is definitely one of the major factors driving adoption.
> > Witness the C ->  C++ ->  Java programming language evolution curve. I
> > updated everything. Now at the end of database definition file we have
>
> > TABLE_DUM = R00;
> > TABLE_DEE = R01;
>
> > x<AND>  y = x ^ y.
> > x<OR>  y = (x ^ (y v R11)) v (y ^ (x v R11)).
> > (<NOT>  x )  = (R11 v x) ^ x'.
>
> > Here is DeMorgan law in shiny brand new notation
> > (<NOT>  x)<OR>  (<NOT>  y) = (<NOT>  (x<AND>  y)).
>
> > Finally, the assertion you are after is:
> > x<OR>  (<NOT>  x ) = x<OR>  TABLE_DEE.
>
> > Now I wonder if there are any other pieces left before one can claim
> > QBQL as yet another Tutorial D implementation:-)
>
> A-algebra projection notation?  Just asking, really I should say
> congrats, best excuse for to enable Java that I've seen in many years.
> Are there any caveats regarding differences in the various jvms?

R <REMOVE> A = R v (R00 ^ A)`.

R= [p r]

           0 0
           0 1
           0 2
           1 1
           1 2

R <REMOVE> [r];

Output: R<REMOVE>[r]=[p]

              0
              1

(There is one little snag to fix. Currently, the inversion operator is domain dependent. Therefore, if database file doesn't define any relation with attributes containing r, then the expression R <REMOVE> [r] would fail. As I already fixed a similar problem for the complement operator, this one would be fixed asap).

> (As much as I admire the brainpower behind TTM, much of it goes into
> aspects that I don't think are essential to elementary RT, but the tiny
> A-algebra that it all depends on is a really nice piece of thinking,
> with a 'spec' of just a handful of pages, which I wish more spec's would
> try to emulate, many others are a size which is inhuman.  I always
> thought it needed an implementation of its own, if only for learning or
> checking purposes.  From comments Hugh Darwen wrote, I believe he
> originated it and CJ Date polished the definitions, so I think of it as
> the Darwen and Date algebra, not the other way 'round.)- Hide quoted text -

I'm not sure I follow their idea that SUMMARIZE is expressed via AND, OR, NOT, REMOVE, RENAME. Emailed to Hugh once, but didn't get a responce. Received on Tue Sep 28 2010 - 18:52:16 CEST