Re: Error in RDF/XML Syntax Specification?

* Graham Klyne <GK@ninebynine.org> [2005-04-06 09:19+0100]
> At 13:46 05/04/05 -0400, Dan Brickley wrote:
> 
> 
> >New version: http://www.w3.org/2001/sw/RDFCore/errata#rdf-syntax-grammar
> >
> >Incorporating Graham's qualifier, a typo fix from Pat Hayes, and a
> >closing observation I could do with someone reviewing (tried to
> >interpret Dave's IRC comment that this was really 2 issues...):
> >[[
> >Serialization of datatyped empty literals is not anticipated by the
> >RDF/XML grammar.
> >
> >This is believed by several developers and former WG-members to be an
> >omission in the grammar defined by the RDF/XML Syntax Specification: a
> >bug was reported (and acknowledged by the editor), relating to the use
> >of an rdf:datatype attribute on empty RDF properties. See the archived
> >mailing list thread for technical details. In addition to the question
> >of the RDF/XML grammar's syntactic completeness, note that this issue
> >identifies a construct that occurs within RDF graphs that cannot be
> >serialized in the RDF/XML syntax.
> >]]
> >
> >Is that last claim right? Is there a difference btw between
> >
> ><foo:prop rdf:datatype="http://www.w3.org/2001/XMLSchema#string"
> >        ></foo:prop>
> >
> >...versus:
> >
> >        <foo:prop rdf:datatype="http://www.w3.org/2001/XMLSchema#string"/>
> >
> >...in terms of this issue and the grammar productions?
> 
> From memory, I don't believe there's any difference at all.  So I think 
> that while the final sentence is technically correct, I'm not sure what it 
> adds to the problem description.  Otherwise, I think your text is fine as a 
> description of the issue raised.
> 
> Secondly, having looked at the link Brian sent, specifically:
> [[
> For the third class of change, W3C requires:
> 
>    1. Review by the community to ensure the technical soundness of 
> proposed corrections.
>    2. Timely publication of the edited Recommendation, with corrections 
> incorporated.
> 
> For the third class of change, the Working Group MUST either:
> 
>    1. Request that the Director issue a Call for Review of an Edited 
> Recommendation, or
>    2. Issue a Call for Review of Proposed Corrections that have not been 
> incorporated into an edited draft (e.g., those listed on an errata page). 
> After this review, the Director MAY announce that the proposed corrections 
> are normative.
> 
> While the second approach is designed so that a Working Group can establish 
> normative corrections quickly, it does not obviate the need to incorporate 
> changes into an edited version of the Recommendation. In particular, when 
> corrections are numerous or complex, integrating them into a single 
> document is important for interoperability; readers might otherwise 
> interpret the corrections differently.
> ]]
> -- http://www.w3.org/2004/02/Process-20040205/process.html#rec-modify
> 
> It seems that if we have general consensus that this is an error in the 
> current syntax spec, we really need a proposed amendment for community 
> review, and maybe a new test case.  (Arjohn has already supplied the 
> latter.)  I have not yet studied the actual grammar to decide an 
> appropriate fix, and I suspect Dave will come up with one more quickly than 
> anyone else.  When such is proposed, I'll try to incorporate it into my 
> parser and report back the test results (based on Dave's Raptor test suite).

So... If I'm left to do the errata on my own, it'll be a quick list of
summaries with pointers into the archive. Better than nothing, but not
as polished a job as the Process doc expects of WGs. I don't currently
have funding that gives me time to spend on this, unfortunately, though 
I'll do what I can. If a WG were to do it, we might reasonably expect 
more to be achievable. One possibility would be to manage RDFCore errata 
as a taskforce of the SW Best Practices WG Task Force. If there were 
sufficient interest in doing that, including someone to actually coordinate 
the TF (can't be me right now unless a funding angel waves a magic wand), I 
reckon we could do a more complete job. In the meantime, I'll try to at 
least note the existence of issues using the errata page or perhaps a
separate issue page, if errata.html is too formal a home for such
things. We could try using the ESW Wiki to collect raw materials, perhaps?

Dan

Received on Wednesday, 6 April 2005 08:44:45 UTC