Re: data & code

"mAsterdam" <mAsterdam_at_vrijdag.org> wrote in message news:40b529c5$0$33919$e4fe514c_at_news.xs4all.nl...
> Dawn M. Wolthuis wrote:
>
> > mAsterdam wrote:
> [snip]
> >>BTW I am still looking for 'type'.
> >
> > For the definition of type?
> >
> > Def: The type of a variable v is the set of all valid values for v.
>
> So, a synonym to domain? Sounds ok with me. However, I've also seen
> people include implied valid operators as part of the type,
> some even including ad hoc operator definitions as being part of the
> type. What is your take on that? (To me this has a very artificial
> ring to it, but that just may be a matter of taste).

I have seen those claims too, but this is an area where I agree with Date -- that type and domain are the same thing. I can also see equating class and type, but classes do include the valid methods (operators) for values (objects) of that class -- but probably best to leave the OO terms out of the def of Type, eh?

If we want to be able to describe a Type as composite/non-scalar or scalar then that can only be discussed with respect to the operators packaged with the DBMS for that Type, so I'm starting to talk myself into tossing the operators into the definition -- perhaps a) & b) definitions. If we need a word for the Domain + Operators, then I'd rather use "Type" than make up a new one.

> > Note: The "type" of a value only makes sense to me as short-hand means
of
> > discussing the type of the variable whose value is this. Values outside
of
> > the context of a variable have no type as far as I'm concerned, even if
they
> > have a representation that looks like representations of some type.
>
> Sharp :-)

well.thanks -- Cheers! --dawn Received on Thu May 27 2004 - 02:05:55 CEST