W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > September 2003

Re: Review of nodeID text

From: Dave Beckett <dave.beckett@bristol.ac.uk>
Date: Tue, 23 Sep 2003 14:49:36 +0100
To: Jeremy Carroll <jjc@hpl.hp.com>
Cc: w3c-rdfcore-wg@w3.org
Message-Id: <20030923144936.23651e38.dave.beckett@bristol.ac.uk>

On Mon, 22 Sep 2003 16:49:14 +0300
Jeremy Carroll <jjc@hpl.hp.com> wrote:

> 
> 
> This is a partial review of rdf syntax 05 sept 2003.

I have made no changes since then (waiting for updated staging area script).

> Considered text:
> 
> 6.1.7
> http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030905/#section-blank-nodeid-event
> 
> 6.3.3 Grammar Action Notation
> 
> 5.2 Identifiers
> http://www.w3.org/TR/2003/WD-rdf-syntax-grammar-20030905/#section-Identifiers
> 
> ==========================================
> Overall - I think the concept works well enough, but the wording could be 
> polished.
> 
> 
> 
> Some **suggestions** for the editors consideration.
> 
> 
> In 5.2 final para:
> 
> delete
> [[
> In addition to rdf:nodeID attribute values being used to give blank node 
> identifiers, these identifiers also need to be generated when not given 
> explicitly.
> ]]
> replace with
> [[
> It may be helpful to generate blank node identifiers even when
> none is made explicit with an rdf:nodeID attribute value.
> ]]

Yes, that's better than the existing words which tends to emphasis the wrong bit
(ie not the generation).

> (This step is done by the implementation technique described in the WD but is 
> needed because of the non-normative dependency on n-triples.
> ARP for instance does not generate such IDs; whereas ARP within Jena does, as 
> does ARP when used as an RDF/XML to N-triple filter.)
> 
> Alternative wording which I like less, but is clearer than the current wording 
> might be:
> [[
> Blank node identifiers are also generated when
> none is made explicit with an rdf:nodeID attribute value.
> ]]
> 
> or
> [[
> In the grammar, blank node identifiers are also generated when
> none is made explicit with an rdf:nodeID attribute value.
> ]]

I prefer:

Blank node identifiers are also generated when
none is given explicitly with an rdf:nodeID attribute value.

with possible forward reference to generated-blank-node-id() in 6.3.3 ?

> 
> 6.3.3 no change suggested.
> 
> 
> 6.1.7
> 
> Suggest delete
> [[
> string-value
> 
> The value is the concatenation of "_:" and the value of the *identifier* 
> accessor. It is permitted to further modify this value in order to preserve 
> distinct blank node identity as discussed in section 5.2 Identifiers.
> ]]
> 
> replace with
> [[
> string-value
> 
> The value is a function of the value of the *identifier* accessor. 
> The value begins with "_:".
> The value matches the N-triples *nodeID* production.
> The function MUST preserve distinct blank node identity as discussed in 
> section 5.2 Identifiers.
> ]]
> 
> 
> i.e. replace a half implemented version based on concatenation, with merely a 
> specification of the blank-nodeid event. The line
> [[The value begins with "_:".]]
> is implied by the N-triple production, so could be deleted. I think it is 
> clearer with it in.

Slightly rewording it
 [[
 string-value
 
 The value is a function of the value of the *identifier* accessor. 
 The value begins with "_:" and the entire value MUST match the N-triples *nodeID* production.
 The function MUST preserve distinct blank node identity as discussed in 
 section 5.2 Identifiers.
]]

Dave
Received on Tuesday, 23 September 2003 09:50:10 EDT

This archive was generated by hypermail pre-2.1.9 : Tuesday, 23 September 2003 09:50:17 EDT