Re: Normalizing Every Thing with XDb2

From: Neo <neo55592_at_hotmail.com>
Date: 28 Jul 2004 10:49:23 -0700
Message-ID: <4b45d3ad.0407280949.7d6fb9d1_at_posting.google.com>


> > create *.
> > create * cls mass.
> > create * quantity 1.
> > create * unit kg.
> >
> > create box1 weight (mass)(1)(kg).
>
> When you insert 100 boxes with different quantities, you need to assert in
> two different places that each number is a quantity AND a measure for some
> box.

Are you in effect saying that a color needs to be "asserted" in two different places each time a car is added to T_Car below (assuming each car is a different color)? If so, what is the significance?

T_Color
Name
red

T_Car
Name Color
Corvette ->red

> Also I suggest you get rid of those parentheses
> [ie (distance)((quantity)1)((unit)cm)]

Doing so wouldn't indicate the same thing, just as (a)((b)c)((d)e) doesn't equal (a)(b(c))(d(e)) or (a)(b)(c)(d)(e) or a b c d e, etc. Each set of parentheses is approximately equal to a level of indirection (similar to joins in RM) as shown below:

T_Color
Name
red

T_Car
Name Color
Corvette (red)

Removing parentheses would have about the same effect as removing them from LISP. Received on Wed Jul 28 2004 - 19:49:23 CEST

Original text of this message