W3C home > Mailing lists > Public > public-rif-wg@w3.org > January 2007

Re: [UCR] ISSUE-12 and ACTION6198 (semantic web rule language)

From: Dave Reynolds <der@hplb.hpl.hp.com>
Date: Mon, 22 Jan 2007 23:06:07 +0000
Message-ID: <45B5435F.6050502@hplb.hpl.hp.com>
To: Chris Welty <cawelty@gmail.com>
Cc: Sandro Hawke <sandro@w3.org>, Christian de Sainte Marie <csma@ilog.fr>, RIF WG <public-rif-wg@w3.org>

Chris Welty wrote:
> 
> 
> At the Jan 2 telecon there was no objection to the following text for 
> addressing Issue-12:
> 
> Sandro Hawke wrote:
>> ---------------------------------
>> PROPOSED - The RIF WG will define a small number of standard dialects.
>> Although each dialect may be considered a rule language, it will be
>> designed for the sole purpose of rule interchange.  Since RIF will
>> support rules that can process RDF as data and will be compatible with
>> OWL then any or all of these dialects might be usable as a Semantic Web
>> rule language.  The Working Group does not anticipate labeling any of
>> these dialects as "standard" for the Semantic Web.
>> ---------------------------------
> 
> However, I object.  I do not want text that says we will NOT standardize 
> a SW rule language, or any softening thereof (like, we do not 
> "anticipate"...).  I am happy with text that says this is not our 
> principle goal, since it is not.
> 
> I think now I understand DaveR's goal - up until this point (despite 
> Dave being quite clear about it all along) I was assuming he was looking 
> for some indication that we WILL standardize a SW rule language, but now 
> I (finally) understand that what he's after is a clear and definite 
> statement one way or the other. 

I had tried to be clear that my aim was to reduce the ambiguity - so we 
could either start committing to doing something here or start 
clarifying expectations. Sure my initial preference had been to get a 
positive commitment but when even the watered down version of that was 
rejected at f2f3 then I fell back on just getting clarity.

> I'm not willing to make such a 
> commitment, as I see many *possible* outcomes and uses for RIF, 
> including being the basis of some future standard SW rules language.  
> However I also see the possibility that it would not be.  I would like 
> to leave both of these possibilities open.
> 
> The charter is vague on this point.  Up until I understood DaveR's 
> motivation here, I believed this issue was about making the vagueness in 
> the charter concerning this point explicit - i.e. making it clear that 
> the vagueness is intentional, while also emphasizing the possibility 
> that a standard SW rules language may come out of RIF.
> 
> Therefore I don't think there is text that I can live with that will 
> satisfy Dave's desire to have a clear statement.  Further, and more 
> seriously, I think such a statement would constitute a change to the 
> charter.

I can accept that there is no text that will satisfy both of us so at 
this point I suppose I just have to give in and drop the issue. After 
all part of my goal was to stop us replaying the same argument in 
circles and just get it pinned down, not to spawn more endless 
discussion of marginal interest to half the group. Sure failed in that one!

However, I really dispute that this sort of statement would be a change 
to the charter. The charter sketches a need for some family of rule 
interchange languages and a line of attack on the problem but leaves a 
lot of details open. Our job in developing the UCR was to pin down the 
space rather better and say what specific requirements we would address. 
Implicit in the charter is the possibility that RIF might be, or might 
lay the foundation for, some semantic web rules language(s)[*]. It seems 
to me entirely within scope that in the UCR work we could have defined 
some variant on that as an explicit requirement for RIF and clarified 
what it meant, or we could have clarified it as an explicit 
non-requirement.

To me this was entirely about what the RIF WG is committing to do, not 
what others might do with RIF in the future. At no point have I 
suggested that we say RIF *can't* be used as a foundation for some 
future semweb rule language.

Dave

[*] This reading is particularly easy when taken in the context of all 
the member submissions.
Received on Monday, 22 January 2007 23:06:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:33:35 GMT