AW: Action-1053 (Review UCR)

Dear Stella,

 

Thanks a lot for your quick review. I already addressed most of your
comments. Please find my comments below. 

 

> 

> 

> Dear WG,

> 

> 

> These comments are on UCR as of Sept. 19th; I know there are still some

> updates planned to switch the order of sections 4&5 and add additional

> analysis about implementability in RIF for some use cases.

> 

>  

> 

> Stella 

> 

>  

> 

>  ---------------------------------------------------

> 

>  

> 

> -- I think section 3 should be deleted because it doesn't really fit well

> in

> this document and the information in section 3 is covered in the RIF

> Overview document. The Venn diagram from section 3 could be moved to the

> RIF

> Overview document. 

 

I did not change it yet, so that we can discuss it today in the telecon. I
see benefits of having this short summary on the existing dialects here in
UCR as it helps to understand the RIF use cases and why some of them are not
covered by the current RIF dialects.

 

 

>To go along with that the purpose statement in Section

> 1,

> 3rd para, 1st sentence would need to be rewritten (it should be updated

> anyway, I think, to say that the purpose is to describe the use cases that

> guided the design of RIF and to document the design requirements that were

> derived from both the use cases and from the basic goals of RIF). 

 

I changed the wording of this sentence.

 

And also

> to go along with it the 2nd sentence in the last paragraph of Section 1

> would need to be deleted.

 

If we remove section 3 today I will delete this sentence.

 

> 

>        

> 

> -- The last sentence of section 6 seems to conflict (or rather to imply

> that

> this requirement wasn't met) with the 2nd part of requirement 5.1.6. The

> explanation might be that 'should' is used instead of 'must'  ...but

> anyway,

> in other documents Core is described as enabling limited rule exchange

> between logic rules and production rules, so I would reword or delete the

> last paragraph of section 6.

 

I have deleted this paragraph since this is already discussed in requirement
5.1.6.

 

 

> 

> -- It could be confusing that the end of use case 4.2 says "This used case

> is not adequately supported by the current RIF dialects as they do not

> specify explicit constructs for integrity constraints " and then the 3rd

> paragraph of the conclusion in section 6 says "For example, the invariant

> meaning of a set of integrity-constraint-enforcing rules would be

> represented within the appropriate RIF dialect... "

> 

 

Since Leora wanted to work on all discussion comments I have not changed it.
But I agree, we should make it more explicit that expressive integrity
constraints and deontic norms such as must, never both (mutual exclusive),
obliged cannot be adequately represented.

 

 

>  

> 

>  -- Some statements saying "currently" or "the group expects" could be

> updated to reflect the state that this phase of the WG is completed:

> 

>       -- Section 6, 3rd para: "The working group expects that RIF...will

> consist of several dialects"

> 

>       -- Section 5, 1st para: "The Working Group has currently approved

> the

> following requirements. "

> 

>       -- Section 5, 2nd para: "...which need to be fully covered by the

> currently specified RIF dialects

> 

>       -- Section 4, 2nd para: "guide users to RIF's currently specified

> dialects..."

> 

>       -- Section 6, 2nd to last para, 1st sentence:  "must be" --> "has

> been"

> 

> 

 

I changed all of them

 

 

 

> 

> -- Section 1, last para, 1st sentence: remove the statement about

> "...several accordant critical success factors for RIF"  because section 2

> now just talks about goals, and only in one sentence mentions an

> implication

> of a goal, but I don't think anyone would relate that to the previous

> statement about critical success factors.

> 

 

Right, since we deleted the CSFs, I deleted this statement about CFS, too.

 

>  

> 

> -- Section 5:  I would consider not having a sub-section "Basic

> Requirements" and just have those at the top level of the section and then

> continue to call out the "General Requirements" in a sub-section if

> needed.

> It's hard to believe that requirement 5.2.1 isn't a General Requirement. 

> 

 

The reason for having 5.2.1 as a basic requirement was that for some
dialects it will be quite hard to develop a fully compliant RIF
implementation, e.g. most RIF BLD engines will not implement full logical
equality. I think the current distinction in general and basic requirements
is useful. Let's discuss it today in the telecon.

 

>  

> 

> -- Section 6, 2nd to last para:

> 

>          I would delete the last sentence of this paragraph. ("Other

> requirements...")

> 

 

deleted

 

>  

> 

>  typo: Section 4.3, last sentence: engergy --> energy

 

corrected

 

 

 

Thanks a lot for the review,

 

Adrian