W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2002

Re: Syntax Doc

From: Jeremy Carroll <jjc@hpl.hp.com>
Date: Tue, 29 Oct 2002 14:47:01 +0100
To: w3c-rdfcore-wg@w3.org
Message-id: <BHEGLCKMOHGLGNOKPGHDAEILCAAA.jjc@hpl.hp.com>

Jeremy:
> Section 2 must have one of the following changes:
> - clearly marked as informative

Dave:
> No,

Clearly a matter of disagreement.

Jeremy
> - moved to being an informative appendix

Dave:
> No.  People needed examples and there were definitely a need for them
> plus an explanation of how the syntax works - this document isn't
> just for parser writers.  I've already had positive feedback on this.

Non-sequitor. A non-normative appendix is still part of the doc, still
available for the reader. The primary role of a normative spec is to
specify. Other docs can give examples.


Jeremy:
> - deleted
Dave:
> Definitely not.  See above feedback.

This is still a non sequitor, but I won't push this point. (Useful material
can be published independent of a spec. A clear recommendation can only be
published as part of a standards process).

Jeremy:
> As a taster:
> [[
> 2.8 "beginning a:Collection"
>
> The example should be change to delete the whitespace between the end of
the
> ex:prop start tag and the beginning of the a:Collection start tag, so that
> the comment cannot be misinterpreted as indicating that the whitespace is
> insignificant.
> ]]
> I think my other corrections are at a similar level of pedantry.

Dave:
>I didn't mention whitespace in the text, but pointed to the
>concepts doc section on XML Literals.


The concepts doc on XML Literals normatively defers to XML Canonicalization,
which I think normatively defers to the XPath Nodeset, where I believe you
find the whitespace as significant (or you may need to go further back). A
reader faced with a *normative* section that says that the object node
labelled with XML content beginning a:Collection, and faced with some
RDF/XML that delivers "\n     <a:Collection ..." may be forgiven for not
following the refs and simply adding a spurious trim() function to their
understanding of how RDF works.



> Would the phrase "object node" would make more sense?

Yes, you used that a few times, it's better but still awkward though. There
is little clarity in section 2 as to when you are talking about XML and when
you are talking about RDF. ... I'll try and be constructive in my detailed
comments tomorrow.

Jeremy
Received on Tuesday, 29 October 2002 08:47:06 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 3 September 2003 09:52:33 EDT