Re: THe OverRelational Manifesto (ORM)

U-gene wrote:
> The expession...
>
> CREATE CLASS SHIPMENT
> {
> No INTEGER;
> WareFrom WAREHOUSE;
> Items SET OF //-- and this is a set of invoice lines
> {
> Article STRING;
> Pieces INTEGER;
> }
> }
>
> ...just other way to define these two variables, which you believe to
> be "better design". But you cannot understand at all, that "CREATE
> CLASS SHIPMENT" define a set of relation, once again - SET of
> relations but your...
>
> >>Shipment base relation...
>
> is not a SET - it is just single relation. Do you understang the
> difference between single relation and SET of relations? It is the same
> as difference between single relation and DB.
>
> And (I'm sure now) that some persons here need to get a lessons of
> reading. My question is not "how to describe SHIPMENT variable in D
> language" at all. I know it very well. I speak about the type
> specification and type implimentation, about polimorphous names, about
> binding etc. and show with example how easy TheORM allows to realize
> some things. My question is "How D language can realize these things?
> if it can do it at all..."

It can. As relation valued attribute(s) representing sets of items in a relvar. Just because it can do it doesn't make it a good design. Inheritence is also covered in TTM. They don't exclude it they just don't have a model for it (yet), but neither do you in your model.

> >>3. scalar type = object class
> Where does this idea appear from??? This is true of course, but if you
> think that TheORM is about this you really don't undertand what TheORM
> is about.

Is there anyone other than you who knows what it is about? Anyway, the acronym is already taken.

--
David Portas