- From: Jeremy Carroll <jjc@hplb.hpl.hp.com>
- Date: Fri, 6 Dec 2002 13:04:37 +0100
- To: "Dave Beckett" <dave.beckett@bristol.ac.uk>, "Brian McBride" <bwm@hplb.hpl.hp.com>
- Cc: <w3c-rdfcore-wg@w3.org>, <jjc@hplb.hpl.hp.com>
> I went back to the paragraph above and added the emphasis keywords. > > [[[ > This specification allows some freedom to choose exactly what string > is used as the lexical form of an XML Literal. Whatever string is > used, it MUST correspond to an XML document when enclosed within a > start and end element tag, and its canonicalization (without > comments, as defined in [REF]) MUST be the same as the same > canonicalization of the literal text l. It is acceptable to use l > without any changes but this is incorrect if, for example, l uses > entity references or namespace prefixes defined in the outer XML > document. > ]]] > > If MUST is used above, then X14CN becomes normative. And I'm worried > when we add new normative refs. We may have to weaken that to SHOULD. > > I think > "but this is incorrect if, for example" > might have to be rewritten in the emphasis form something like: > "but this MUST NOT be used if, for example,..." > > ? > > > I just noticed > http://www.w3.org/TR/2002/WD-rdf-concepts-20021108/#section-XMLLiteral > refers to: Canonical XML [XML-C14N] (with comments). > > so is it with/without comments? > Sorry I have been sleeping ... Your text looks fine to me, with one word addition "It is acceptable to use l without any changes" ==> "It is often acceptable to use l without any changes" We use to have a normative ref to XC14N in concepts, this in turn normatively refers to C14N. So, moving the XC14N ref to syntax, and having a C14N ref in concepts is *not* adding to the normative refs burden, just clarifying its size. On the comments issue the mapping from the syntax to the lexical form was agreed as with or without comments at implementation descretion. Hence the minimal requirement is without comments, which is why your text is good. However, IMO, if the implementation chooses to keep the comments in the mapping to lexical form, then it should also keep them in the datatype lexical-to-value mapping. This last point I decided arbitrarily at editor's descretion. (invitation to challenge/discuss that judgement) I guess it would be consistent with the WG decisions to also allow that dattaype l2v mapping to be implementation dependent. I prefer to localize the implementation dependency in one spot, and the syntax doc seems the right place (sorry). Jeremy
Received on Friday, 6 December 2002 07:06:00 UTC