Re: OR DB design

you're new to this stuff aren't you ?

"Matthew Beale" <mixonic_at_synitech.com> wrote in message news:16297e95.0108151132.5cee5698_at_posting.google.com...
> Why arent Object relatoinal Databases better accepted by the
> commercial community?
>
> I am participating in a project which is attempting to use a OR
> database as a AI enviroment. The AI can "walk" the DB, traverse up of
> down the heirarchy. the trick is, the heirarchy branches both up and
> down, leading to differant relationships. We then grant the AI the
> capability to create its own links by repeatedly traversing a path.
> we add that path the the AI's memory. the more times they traverse,
> the more longer the memory lasts. the communication is based on data
> trees. We take any given scentence and break it down into componant
> parts, then use our grammer engine to reconstruct the scentance in a
> proper format for entry into the DB.
>
> In my mind, and OR DB acts much like a quark. It is not binary, like
> much of our world, but many-leveled. Because of the self-referencial
> and recursive properties of an OR, you CANNOT know exactly where you
> are at any given moment. the time to traverse the recursive
> references will be literally eternity. this is why we let the AI
> create its own "map" in its memory. We grant it pattern recognition
> so it can decide on its own when a certain path will always lead in
> the same sequence, and can recognize the simplest path to any given
> branch.
>
> I was raised on Object-Relational systems, my first DB was a OR
> design. My terminology in this area is weak, but i welcome comments
> and conversation. The OR model lends itself to a myriad of
> applications, and so i hope to encourage the construction of a OR
> databasing community. It is more than a Databasing method, it is the
> closest we can get to true selfreferencial computer applications. It
> test the limits of logic.
>
> -Matthew Beale
> mixonic_at_synitech.com
Received on Fri Aug 17 2001 - 04:57:19 CEST