Re: Implementation of boolean types.

Gene Wirchenko wrote:
> On 24 Jul 2005 16:31:43 -0700, "dawn" <dawnwolthuis_at_gmail.com> wrote:
>
> [snip]
>
> >Similarly, I could take a proposition like this one:
> >
> >I topped the pizza crust with sauce, mozzarella, and pepperoni
> >
> >and model it with a relation
> >
> >PizzaToppings("pizzaId", "topping")
> >
> >and then ask my model the question "What topping do I put on after the
> >mozzarella?" It is not that I am using relations that introduces the
> >flaw -- it is that I opted to model it this way, which was not useful
> >for the question I'm asking. So, it is a flaw to use this particular
> >metaphor and, in this case, the metaphor is flawed.
>
> No, the metaphor is incorrect

Which seems like a flaw, eh? But I'm thinking you have that 0,1 thing going, where I'm a bit more of a grey person.

> in this case as it does not do what
> you want. That, however, does not mean it is flawed

? puzzled look

> or that it is not
> useful.

Again, "all models are flawed, but some are useful" (G. Box IIRC) I agree that a flawed data model might still be useful. In the example given, it was not, however.

> In another case, you might correctly want that metaphor.
> Maybe, when making stew?

Yup, agreed.
>
> The RM cafe. We can SET you up with a hearty stew, or you can
> ORDER pizza. Please ignore the arguments you hear: we are still
> trying to decide whether to call our collective offerings a set or a
> list.

In the mean time, the dishes are QUEUED up to be brought to the customers at their TABLES and the dirty dishes are STACKED for washing.  We better start FUNCTIONING or we will start to have poor RELATIONS, lose customers, stop paying bills and start hearing from COLLECTIONS agencies.

smiles. --dawn Received on Tue Jul 26 2005 - 04:36:30 CEST