Re: the relational model of data objects *and* program objects
Date: 19 Apr 2005 06:20:57 -0700
Message-ID: <1113916857.303949.119360_at_l41g2000cwc.googlegroups.com>
mountain man wrote:
> "erk" <eric.kaun_at_gmail.com> wrote in message
> news:1113833874.524704.128900_at_o13g2000cwo.googlegroups.com...
> > mountain man wrote:
> >> "erk" <eric.kaun_at_gmail.com> wrote in message
> >> news:1113399979.268750.316550_at_l41g2000cwc.googlegroups.com...
> >> > I would infer that the languages used for the program objects
would
> >> > benefit from relations as first-class entities.
> >>
> >> Such as that found within SQL?
> >
> > SQL does a rather poor job of it, so I wouldn't say SQL is a good
> > "model."
>
> I dont have any problems with SQL, but I am
> still not clear on what you mean by the expression:
> "benefit from relations as first-class entities" but suspect
> it is a formalism used in set theory and/or db theory.
I just mean relations as actual language-level entities, just as classes and objects are in Java, for example. Even PL/SQL, for example, does a poor job of it, with too many types of structures and incoherent support for relations separate from stored data.
> The programs and data are --- in a certain sense --- like ying
> and yang. You cannot have one "operational" without the other.
> You cannot have a database system (in use) without an
> application running on that dbms.
Sure, although reports and queries can be seen as applications in their own right.
> So I am expecting in the end, a model to reflect
> (and address) this issue.
But "reflect (and address) this issue" gives nothing to go on. There's no coherent response to that, at least that I can formulate. What would you expect to see, even in vague terms? You've not given any indication what you're looking for, just the "issues" the model should "address," but those words have little meaning in the context of a formalism like a model.
> We make a definition and define the term "organisational
> intelligence" [as related to computer systems] to be the sum of all
> the organisational data plus the sum of the all the organisational
> programs (in production use) ----- ie: SOURCE CODE.
> (Irrespective of the language!)
Various mathematics formalize "the sum of" operators and numbers and types, but do so in terms of axioms and basic definitions. You're apparently seeking unifying axioms for both data and programs? A model should be minimal, but I'm not sure that it would capture what you're talking about so I'm thinking "model" might be the wrong word for it.
> So the (evolved) model should be not for the data, but for
> the [computerised] "organisational intelligence".
If you're looking for organizational models, have you investigated industrial engineering? Given its focus on formalizing and optimizing processes, it might be closer to what you seek. I'm not sure a model of computation that includes all organizational factors is of any value.
- erk