Comment to Pointer Methods in RDF 1.0

Hi all,

I am providing some further background information as a followup of my
message of Jan 22
<http://lists.w3.org/Archives/Public/public-wai-ert/2013Jan/0003.html>
which I brought about in today's conference. As I said in the call, I hold
a particular interest in that document progressing, given that I have
encountered situations when I wished that advanced use scenarios (as the
one I depicted in my previous email) could be supported. Thus, I am open
to provide more, extensive efforts if my assistance is needed.

First, regarding editor's contact info. A few weeks after I sent my
comments, I noticed Carlos Iglesias was not working at Fundación CTIC
anymore since Dec 2012. I directly approached him, who confirmed he no
longer had access to CTIC email, and thus had not received my comments
(neither directly nor through the WG mailing list, I then attached him a
copy just in case). I suggest that he could be contacted ( contact info
here <http://carlosiglesias.es/blog/contact/> ) to check the role he can
assume in the future regarding the ERWG, and specifically regarding the
Pointers in RDF note.

Second, regarding my concern that the document might have become
potentially outdated. The current Working Draft was released in early
2011, and it had indeed had very few changes since the first public
Working Draft <http://www.w3.org/TR/2009/WD-Pointers-in-RDF-20090310/> in
2009. During these four years, several results in the area have progressed
quite much, such as:

[1] For multimedia contents: Media Fragments URI 1.0 (basic)
<http://www.w3.org/TR/media-frags/> together with other companion
documents. A recommendation by the Media Fragments WG I discussed in the
previous email with comments (sorry I forgot including the link there). It
deals with a URIref-based syntax and semantics to refer to media fragments
in four dimensions: temporal, spacial, tracks, and named fragment ids
(e.g.
chapters).

[2] For paged contents: Parameters for Opening PDF Files
<http://www.adobe.com/content/dam/Adobe/en/devnet/acrobat/pdfs/pdf_open_parameters.pdf>
A note by Adobe specifying parameters that can be passed to PDF-reading
software in different ways (including URI fragments) to refer to different
parts of the document. Even though it was already public in 2009, relevant
advances on PDF have since happened. On the one hand, several new PDF
versions have been defined and they have been included in the ISO 32000-1
standard. On the other hand, PDF has gone under specific advances
regarding accessibility in the W3C, such as the submission and publication
of PDF Techniques for WCAG 2.0. In between, PDF/UA-1 has been standardized
as ISO 14289-1:2012, which defines guidelines for universal accessibility
of PDF (indeed establishing an accessible subset of PDF). Thus,
PDF-oriented pointers should not be skipped out.

[3] For text contents:
<http://www.tei-c.org/release/doc/tei-p5-doc/en/html/SA.html>. The P5
Guidelines is an extensive set of guidelines by the Text Encoding
Initiative Consortium to digitally encode literary and linguistic texts.
The P5 Guidelines were released in 2007, but they have been updated twice
a year, with a major release (P5 v2.0.1) in 2011. They are widely used in
humanistic documents, and have been adapted as well by specific
communities. Of interest to us is chapter 16 (which I link above), on
"Linking, Segmentation, and Alignment". Although it covers a document
format EARL will not usually deal with, it is the result of extensive work
by that consortium (part of which has been input indeed to other W3C
documents and the XPointer registry). Thus, it merits being taken into
account when defining a pointer format.

In addition, some W3C Recommendations, although previously existing, also
had pointer mechanisms defined and have been refined or updated since. In
particular, I am thinking of:
[4] For images:
<http://www.w3.org/TR/html5/embedded-content-0.html#the-map-element> HTML5
is now gaining widespread adoption (and has reached the Candidate
Recommendation status), having an updated definition of map and area
elements, and a processing model for image maps.

[5] For any XML content: <http://www.w3.org/TR/xlink11/> XLink, now
updated to version 1.1 defining different types of link relations for any
XML document, which we could take advantage of.
[6] ... and <http://www.w3.org/TR/xpath20/> XPath, now with version 2.0
widely supported, updated to a Second Edition, and updated in the XPointer
registry.
[7] ... and <http://tools.ietf.org/html/rfc5988> RFC5988 Web Linking
(intended to be used by the Atom format, and in the rel/rev HTML
attributes).

[8] For vector graphics: <http://www.w3.org/TR/SVG/linking.html> SVG 1.1,
now updated to Second Edition (acknowledging XLink as well), which devotes
a chapter to linking.

Etcetera.

In summary, I honestly think the Pointers in RDF document should try to:
1) take all these external developments into account, 2) procure support
for (or reference to) them whenever possible, and 3) get specific feedback
from those communities on the suitability of the result.

Regards,

Samuel.

Received on Wednesday, 3 April 2013 19:05:46 UTC