Re: Bob's 'Self-aggrandizing ignorant' Count: Was: What databases have taught me

George wrote:
> ...
>
> Marshall,
>
> Try my description:
> "The relational model comprises a collection of features useful for
> constructing large unpredictable state machines from small predictable
> state machines"
>
> 1. Is that statement true of false?
> 2. Is it a good descriptor?
> 3. Should I be calling other people idiot on the strength of such a
> statement?
>
> Marshall are you able to answer these questions?
> Of course anyone else should feel free to answer as well.
>

I don't speak for Marshall and I usually wouldn't reply to what appears to be an attempt at sophistry, but I would like to point out that the question seems clear baiting to me, as the original statement of history that you have mangled might well have been. If both are so, now each of us has taken the respective bait. The reason I'm replying is not because I speak for the cdt group but because I object to you throwing spurious stuff over the fence, the cdt group has bigger questions to deal with and programming technique, while important in general, is not specifically the topic of cdt. AFAIAC, any OO-RM comparison is absurd on the face of it and the OO people would be doing a service to all if they were to establish just what their theory is in terms that number a minimum and are devoid of code examples.

My answers (which I hope will be taken not as sarcasm but as honest replies to your question):

1. Unknown unless possibly, one assumes the RM includes nulls. Further, why would anybody try to make small state machines using relational operators? Besides, I object to the term relational model in the first place, my limited grasp of language suggesting that a model is a downsized replica of the real thing, with arbitrary functional aspects removed. Relational theory as a term seems to make more sense, that mentions the real thing.
2. No. I don't know where but I remember the early buzz about C++ and Stroustrup stating clearly that his first realization of it was during a simulation project. My bias is that I like C++ compilers because they are necessarily two-pass so I can make forward references in a module and localize my variables as well.
3. Ask your mother. If she's dead, ask yourself what she would have done.

p Received on Thu Jun 29 2006 - 22:26:34 CDT