Review of http://www.w3.org/TR/2007/WD-curie-20071126/ "CURIE Syntax 1.0"

[Bcc'd to www-tag@w3.org]

I have an action from the TAG [1] to provide a review of http://www.w3.org/TR/2007/WD-curie-20071126/. What follow is intended to complete that action.

Stuart Williams
[1] http://www.w3.org/2001/tag/group/track/actions/81
--
Hewlett-Packard Limited registered Office: Cain Road, Bracknell, Berks RG12 1HN
Registered No: 690597 England


Editorial:
==========

Section 1: Introduction

The narrative presents an isbn based example. The first para after the 'box' states:

        "Yet, in the example given, the whole reason for using a QName was to abbreviate the URI..."

However, there is no URI given in the example (I'm reading isbn: as a 'prefix' rather than a URI scheme name). Suggest showing the full URI that is being abbreviated.

--

Section 3. Syntax:

What is the 'semantic' of the orange background of some statements in this section?

Also, please give the rationale for reserving the "_" prefix - I assume that its because of common use for introducing bnode ids in RDF.

--

Substantive:
============
Section 1: Introduction

"This specification addresses the problem by creating a new data type whose purpose is specifically to allow for the abbreviation of URIs in exactly this way"

I believe that to do a complete job, so that language/schema authors (language designers - the intended audience of this document) can include CURIEs in their language specification and associated schema will require an new XML Schema data type which specifies lexical and value space... say, xsd:curie. Is this need being addressed and if so where/how?

--

Section 1: Introduction:

"This type is called a "CURIE" or a "Compact URI", and QNames are a subset of this."

This is true only for the lexical space. The value space of CURIE and QNames do not overlap - and so the claim that QNames are a subset of CURIE is false and will lead to continued confusion. Please clarify or remove this claim.

--
Section 3. Syntax

"A CURIE is by definition a superset of a QName."

As above, this is misleading. Please do not make this claim (particularly in a section with normative force).

--

Section 3. Syntax

"To disambiguate a CURIE when it appears in a context where a normal [URI] may also be used, the entire CURIE is permitted to be enclosed in brackets ([, ])."

This is potentially distrubing on two counts:

1) it hints at extending the lexical space of xsd:anyURI in order to permit CURIEs in places designated for xsd:anyURI values. A separate xsd: datatype definition for CURIEs would ameliorate this concern.

2) it risks/threatens being perceived as an extension of "Generic URI Syntax" RFC 3986 or IRI syntax.

--

Section 4.1 SPARQL

The SPARQL Rec does not claim to use either CURIE or QNames. SPARQL prefixed names have some minor syntactic differences - eg. ':' is a legal prefixed name in the presense of an appropriate default prefix. There may also be detailed differences in the characters that are admissable in SPARQL prefixed names and CURIEs.

--

Section 5.2 Ambiguities between CURIEs and URIs

"For example, in XHTML the href attribute allows a URI to be specified that will be navigated on user action, but it would also be useful to be able to abbreviate this URI, using the compact syntax."

This speaks of the utility of allowing CURIEs in href values, however speaks nothing of the interoperability problem that would cause for existing browser *if* new content were deployed with such 'safe' appreviations. The remainder of the examples, fortunately, uses a 'resource' attribute (whose defn I have failed to find in XHTML 1.x or HTML 4.01) rather than href.

Suggest rewording the quoted para to avoid mention of 'href'.

--

Received on Wednesday, 13 February 2008 11:06:17 UTC