Re: Possreps - still trying to see the value

From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Sun, 04 Mar 2007 21:37:16 +0100
Message-ID: <45eb2d75$0$333$e4fe514c@news.xs4all.nl>

Marshall wrote:

> Okay, we mentioned possreps in passing in a recent thread.
> I'm still unclear on their value.

[snip]
Exegesis is not my hobby, but for this I'll make an exception :-)

In Foundations for Object/Relational database: The Third Manifesto (I remember thinking for a moment that even Date had gone mad before I skimmed through this one and bought it. Later TTM editions had a different title.), Date & Darwen wrote at page 100 about scalar types:

<TTM>
... no matter how complex their actual representation might be, scalar types have /no/ /user-visible/ /components/.* (Actual representations are an implementation matter, of course, not something that is of concern to the model.) ...

---

*Do not be misled by the fact that scalar type do have "possible representations" and those possible representations in turn do have components that are user-visible ... (italics original)
</TTM>

At the earlier mention of possreps (p66, tutorial D intro) no purpose or necessity is stated.

It's purpose in the argumentation, IMHO, is to introduce encapsulation (while avoiding the term encapsulation). At page 352 (bold): "a type is encapsulated if and only if it is scalar". Received on Sun Mar 04 2007 - 14:37:16 CST