Review of CURIE Syntax 1.0

I have an open action to review the CURIE draft. I assert that this
message completes that action. This is a review of the 22 January 2008
Editor's Draft, the most recent I could find when I started.

First, some specific technical comments:

* In section 1, paragraph 3, that begins "In many cases, language..."

  I think the statement that "QNames are unsuitable in most cases
  because they are NOT (sic) intended for use in attribute values" is
  insufficiently motivated. In what sense are they unsuitable for use
  in attribute values where CURIEs are suitable? The most common
  problem with QNames in attribute values has to do with prefix
  mappings ("QNames in content") and CURIEs seem to inherit all of
  those problems.

  I suppose the square bracket notation might be held up as a solution
  to this problem, but I observe that it comes with problems of its
  own. In languages that expect a sequence of tokens, it's not
  uncommon to make the attribute type NMTOKENS. While QNames are
  acceptable as NMTOKENS, CURIES with square brackets are not.

* In section 3, paragraph 2, that begins "When CURIES are used in XML-based..."

  I recall that early in the development of CURIEs, I and others, expressed
  grave concerns about introducing alternate mechanisms for defining prefixes
  in XML documents. Subsequent drafts returned to the use of xmlns:

  I see that this draft (or perhaps I failed to notice it in the last draft
  I read), reintroduces the possibility of "additional prefix mapping definition

  I continue to think that this is very unwise at best.

  It troubles me particularly that these mechanisms are used in section 4.3
  immediately after the statement "documents annotated with RDFa could use
  the xmlns mechanism to define prefixes". Given that they could, I assert
  that they SHOULD, if not MUST.

* In section 5.2, paragraph 1, that begins "There will be situations..."

  The end of that paragraph suggests that the following example is about
  an email address, but that appears not to be the case. At least, I don't
  see how to interpret the example:

    <span rel="foaf:homePage" resource="http://...">home</span>

  as an email address.

In summary:

I appreciate the work that the editors have undertaken to explain the
motivatation for CURIEs and address concerns that were expressed
earlier. I'm disappointed to see alternate prefix mapping mechanisms
reintroduced to XML, but I'm not sure that's entirely unprecedented
On the whole I remain unconvinced that the benefits asserted for
CURIEs outweigh the costs.

                                        Be seeing you,

Norman Walsh <> | When told of a man who had acquired            | great wealth, a sage replied, 'Has he
                              | also acquired the days in which to
                              | spend it?'--Solomon Ibn Gabirol

Received on Tuesday, 12 February 2008 20:23:20 UTC