Re: RIF UCR REVIEW

Dave Reynolds wrote:
>
> Adrian GIURCA wrote:
>>
>> Paula-Lavinia Patranjan wrote:
>
>>> IMO, the followed approach for gathering requirements is a well suited
>>> one. I don't think that concrete rule language should directly pose
>>> requirements on RIF. For this, we have the requirement on rule language
>>> coverage, which acts as an umbrella requirement for the ones coming 
>>> from
>>> RIFRAF.
>>>
>>>   
>> Concrete languages will participate in interchange so, I believe 
>> requirements are derived from them. Otherwise requirements are not 
>> connected with the scope of RIF  i.e. interchange. 
>
> I think this may partly be a terminology problem.
>
> The "use cases" in the UCR document are not really use cases in the 
> full sense of RUP, they do not directly drive design. Here they have 
> two main functions. They motivate high level requirements such as 
> "interoperation with OWL". They communicate the sorts of things RIF 
> might be useful for.
Well, I suppose that use cases provides requirements, but now I 
understand. They are motivation for high level requirements.
>
> The detailed design decisions are supposed to driven by consideration 
> of concrete languages (hence RIF/RAF and the list of languages 
> represented by the group) and by the more concrete use cases. Remember 
> that the "use cases" in UCR are abstractions of the larger number of 
> more grounded use cases in the Wiki.
>
>> By the way, many questions in RIF/RAF are meaningless for interchange.
>
> Possibly so, would you like to give more details on that?
I have in mind the following  languages: Prolog, F-Logic, Jess, Jena2, 
SWRL, JBoss Rules, Oracle Business Rules.
The terminology used in RIF/RAF 
<http://www.w3.org/2005/rules/wg/wiki/Rulesystem_Arrangement_Framework> 
is pure logical. In some languages it is not straightforward to express 
these questions.

Lets look now to some questions:
1.1.1. For me this seems to don't apply for too many languages.And how 
helps us the choice: "Only single occurrence of variables per predicate"
1.1.2 The choice "Head-only variables allowed"
1.3. Do you allow this in Jena2?
2.5 What is the meaning of this question in JBoss Rules, Oracle Business 
Rules for example? But in Jena2?
2.7 "...Unlabeled Clauses"  Which languages have clauses?
I don't want to extend the discussion now.  This is not my goal. I 
accept the document as it is. However, I suppose it is better that the 
RIF/RAF to be public and then its results to be used against 
requirements. Otherwise the Core may not really use the UCR and/or 
RIF/RAF document as an input.
>
> Dave
>
>
All the best,
Adrian

Received on Monday, 26 February 2007 11:40:06 UTC