RDF-ISSUE-127 (Comment: encoding of strings): Comment: multiple ways to encode string codepoints [RDF N-Triples]

RDF-ISSUE-127 (Comment: encoding of strings): Comment: multiple ways to encode string codepoints [RDF N-Triples]

http://www.w3.org/2011/rdf-wg/track/issues/127

Raised by: Guus Schreiber
On product: RDF N-Triples

 From: Gregory Williams <greg@evilfunhouse.com>
Date: Thu, 25 Apr 2013 17:48:34 +0800

RDF WG,

I had submitted comments on N-triples back when it was a part of the Turtle document pre-LC, and Gavin indicated that N-Triples wasn't ready for publication and that my comments would hopefully be addressed in the future. In checking in on this a few days ago, I noticed that N-Triples now seems to be Note-track ("First Public Working Group Note", though I'm not sure what that actually means). The Note includes changes to N-Triples from the previous RDF Test Cases format. This concerns me (as I have already mentioned to Gavin on twitter) as I think the path of least resistance for the WG may be to leave this Note up without requiring the process involved in a REC-track document. I don't mean to suggest that the WG is intentionally trying to avoid addressing feedback, but that it may simply be hard to motivate the required work at this point.

My main concern was in the (IMO significant) change from having a single way to encode string codepoints in the old RDF Test Cases format (the encoding table in section 3.2) to the new format where many codepoints have multiple valid encodings (direct utf8 data, \u and \U escaped forms with mixed case hex numbers, direct escaped forms like "\r", etc.). Although I haven't gone through the most recent Note text in detail, I believe my comments still stand, and would appreciate a response on them.

See: http://lists.w3.org/Archives/Public/public-rdf-comments/2013Apr/0063.html

Received on Thursday, 25 April 2013 14:19:17 UTC