Re: Operationalize orthogonality

From: mAsterdam <mAsterdam_at_vrijdag.org>
Date: Sun, 04 Jun 2006 12:32:28 +0200
Message-ID: <4482b64a$0$31641$e4fe514c_at_news.xs4all.nl>


Erwin wrote:
> Translate an external representation of a value to an internal one (and
> ring a bell if an external representation turns out to be "not a valid
> value of this type"), the structure of which is comprehensible to the
> dbms's storage engine ???
> Translate an internal representation of a value to an external one
> (maybe text-like, maybe a format defined by some host language) ???
> Define the API that the operator implementations must adhere to ???

API? The protocol must be defined. For that, the boundary needs to be clearly demarcated. I have not seen any example of this. Yet, it is (IMO) a hard requirement for this orthogonality to have any operational significance. Maybe somebody can point me in the right direction.

> I think it 's that kind of stuff. That is to say : if by "type
> system", you understand the same thing as I do.

Yes, I think so. Received on Sun Jun 04 2006 - 12:32:28 CEST

Original text of this message