Re: Mixing OO and DB

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Fri, 22 Feb 2008 07:36:14 -0400
Message-ID: <47beb3af$0$4033$9a566e8b_at_news.aliant.net>


S Perryman wrote:

> Bob Badour wrote:
>

>> S Perryman wrote:

>
>>> Specifically : what "principle" is actually in effect here ??

>
>>> That if some type T2 happens to be substitutable for T1 for one
>>> particular context, that T2 should be a Liskov/Wing subtype of T1 ??

>
>>> Hardly "value-added" (wrt to the Liskov/Wing definitions) is it.

>
>
>> Actually, the principle provides a method to evaluate type systems 
>> just as normalization provides a method to evaluate logical designs.

>
> FORALL System S, FORALL Type T1,T2 IN S, the Liskov/Wing subtype
> definition certainly provides the means to evaluate the degree of
> substitutability possible for any given T1/T2.
>
> I assume the above is what you meant by "type systems" (otherwise the
> analogy with normalisation is not a good one) . If I am mistaken, please
> feel free to correct me.

You left the part about subtypes out, but otherwise, that will do as the start of a pseudo-english description of an algorithm for evaluating a system of type systems. Received on Fri Feb 22 2008 - 12:36:14 CET

Original text of this message