Re: RIF: soundness and semantics --> functional requirements?

Paul Vincent wrote:

>The *requirement* is interchange, and a *solution* to achieve that would be a 
>declarative semantics interchange format. 

"Interchange" unqualified is insufficient to imply the "functional requirements"
we were talking about.  I would more completely state it as:

  Meaning preserving interchange of rules for at least three families of rule
  languages: Logic Programming, Condition Action, and First Order Logic.

Meaning preserving => formal semantics
Coverage (which could have been stated wrt rule types instead of languages) => 
 multiple semantics

>In my systems engineering days, we used the term "functional requirements" in 
>order to specify the design constraints deduced from the requirements to guide 
>the implementation. Generally these are the consensus / obvious deductions from 
>the (business or high level) requirements - such as if a data store is required 
>then the functional requirement is to use a database. Perhaps we need candidate 
>functional requirements listed like this, separate from the critical success 
>factors + requirements?

Agreed.  This thread is about determining some of the "functional requirements"
or constraints on the language/format design.  

-Evan

Received on Thursday, 18 May 2006 14:19:42 UTC