W3C home > Mailing lists > Public > public-rif-wg@w3.org > April 2008

Re: CURIE proposal ...

From: Michael Kifer <kifer@cs.sunysb.edu>
Date: Tue, 22 Apr 2008 14:46:11 -0400
To: Bijan Parsia <bparsia@cs.man.ac.uk>
Cc: Axel Polleres <axel.polleres@deri.org>, Sandro Hawke <sandro@w3.org>, Chris Welty <cawelty@gmail.com>, "Public-Rif-Wg (E-mail)" <public-rif-wg@w3.org>
Message-ID: <2681.1208889971@cs.sunysb.edu>


With your ability to pull things out of context, you should explore joining
the current political campaign down here...

Here is my position on the presentation syntax issue:

1. Ideally, the pres syntax should be a parsable language.
   We started with that years ago, but some people were adamantly opposed.
2. If not 1, then the presentation syntax should be as simple as possible,
   and its primary use should be semantics+examples.
   Some simple shortcuts are ok, but they should really be simple and not
   ugly hacks.
3. If possible, use a slightly restricted version of the presentation
   syntax to define a real language. Given the experience with 1, this
   should be non-normative.
   If this discussion opens the eyes of those who were against #1 then
   we should go back to 1.
   But we no longer have the time for that or #3 in this phase.


> On 22 Apr 2008, at 17:49, Michael Kifer wrote:
> >> On 22 Apr 2008, at 16:08, Michael Kifer wrote:
> >>>
> >>>> My point is that a halfway stop is no good. Either have a fully
> >>>> specced Presentation Syntax or use the XML directly.
> >>>
> >>> Yeah! Write semantics using XML and then come back.
> >> [snip]
> >>
> >> Dude, it's a disjunction. I know you don't like disjunctions, but
> >> that's no reason to ignore them.
> >
> > I love disjunctions. I see that you did not understand what I said,  
> > so I
> > will explain it formally.
> >
> > You: a \/ b.
> > Me:  ~b.
> 
> You may have thought you were doing that, but you weren't. Sorry. At  
> least, your implicature failed. If you wanted to say not B, then you  
> shouldn't have made it an imperative.
> 
>   In fact, afaict, you embrace ~a as well.
> 
> Plus, I never said what I embrace. I just warned that having a  
> presentation syntax raises certain expectations, however hard you try  
> to squash them.
> 
> >> My point is that if you have a presentation syntax, people *will* use
> >> it and treat it as a concrete syntax no matter what warnings you put
> >> up. Surely your experience with recent comments confirms this.
> >
> > Are you talking to me?
> 
> Yes:
> 	http://lists.w3.org/Archives/Public/www-rdf-rules/2008Apr/0031.html
> 
> """The presentation syntax, which is used there is
> for the purpose of giving definitions and semantics, since it is clearly
> not possible to use XML for that.
> 
> This does not close the door to the introduction of another human- 
> readable
> syntax, but, at least for now, the group has decided against that.   
> It is
> likely that the presentation syntax will be "slightly concretized" to
> simplify the job of writing test cases and such. But it is much less  
> likely
> that this concretized version will become normative."""
> 
> 
> >> Finally, I don't think writing a semantics using XML (e.g., if one
> >> uses RELAX NG's compact syntax) is so bad.
> 
> See below. Plus, we may be doing this for OWL. I'll let you know how  
> it works out.
> 
> > Let me repeat: come back when you are done.
> >
> >> I understand you prefer
> >> not to, which is fine.
> 
> See above.
> 
> >> But then don't bitch when people want the
> >> presentation to be a (readable) concrete syntax.
> >
> > You are utterly confused.
> 
> If by "you" you mean *you*, then I agree :)
> 
> > I was bitching?
> 
> 	http://lists.w3.org/Archives/Public/www-rdf-rules/2008Apr/0029.html
> 
> """It would have been helpful if you were a bit more specific, but  
> you should
> understand that FLD is a logic framework, and it uses notation typically
> found in any logic textbook. FLD's (and BLD's) presentation syntax is
> abstract syntax designed to describe the formal semantics. This
> presentation syntax is missing several important aspects of a concrete
> language -- on purpose. This nature of the syntax is stressed at  
> least twice
> in each document. The concrete language is XML."""
> 
> I started this conversation with a friendly observation about the  
> experience of having a so called abstract syntax for specification  
> purposes (and if you don't like giving a semantics in terms of xml,  
> try doing it in terms of rdf :)). I'm not sure why you decided to go  
> postal on my ass, but you shouldn't be surprised when people kick back.
> 
> Cheers,
> Bijan.
> 
Received on Tuesday, 22 April 2008 18:47:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:33:48 GMT