Re: SUPPORT FOR DECLARATIVE TRANSITION CONSTRAINTS

On Sep 26, 5:14 pm, r..._at_raampje.lan (Reinier Post) wrote:
> Brian Selzer wrote:
> >SUPPORT FOR DECLARATIVE TRANSITION CONSTRAINTS
>
> >A proposed enhancement to D
>
> >By Brian S. Selzer
>
> >ABSTRACT
>
> >A non-procedural mechanism is proposed for declaring transition
> >constraints using set-based relational operators, without altering
> >relvar headings and independent of whether updates affect keys.
>
> [...]
>
>   Brian,
>
> You still misunderstand the fundamental objection against
> you proposal (which hasn't really changed in the time I've
> been following this newsgroup).  The objection is that
> no matter what mechanics you add within a database, there
> is fundamentally no way to ensure from within the database
> that its statements remain up to date with changes to states
> of affairs in the world.  No matter what kind of database
> I use to record information about comp.database.theory
> and no matter what kinds of constraints it will support,
> that database fundamentally has no way of knowing whether
> any given posting written by 'Brian Selzer' is written by the
> original Brian or an imposter posting under the same name.
> Adding a PERSON_ID allows you to *express* that problem,
> but it doesn't allow you to solve it.  Adding constraints
> on what transactions may do, as you propose here, doesn't
> solve it, either.  And nothing else will.
>
> --
> Reinier

Are you assuming that I claim that the system should be able to ensure correctness rather than just consistency? Let me assure you that I claim no such thing! The system can't detect lies. It can't detect mistakes that are consistent with its constraints. It can only detect inconsistencies.

No matter how you try to disguise it, the objection boils down to the objection that a child makes when he is forced to show his work rather than just giving the answer: "Why do I have to write down all the steps when I already know the answer?"

It is not necessary for the system to be able to ensure that it is up to date. It is necessary for the users to be able to keep the system up to date, otherwise the content of the database should be considered suspect. Luckily, there is a fundamental mechanism for the users to keep the system up to date. In the world of TTM, that mechanism is the multiple assignment. Received on Mon Sep 27 2010 - 05:09:10 CEST