- From: Paul Grosso <pgrosso@arbortext.com>
- Date: Mon, 09 Sep 2002 17:45:07 -0400
- To: reagle@w3.org, "w3c chairs" <chairs@w3.org>
- Cc: <w3c-xml-plenary@w3.org>, "XML Signature (W3C/IETF)" <w3c-ietf-xmldsig@w3.org>
At 13:19 2002 09 09 -0400, Joseph Reagle wrote: >On Thursday 05 September 2002 09:31 am, Paul Grosso wrote: >> The XML Core WG announces Last Call of "Namespaces in XML 1.1" >> available at http://www.w3.org/TR/2002/WD-xml-names11-20020905/ . >> >> Please note the availability of the version with differences >> (with the 1.0 Recommendation) highlighted at >> http://www.w3.org/TR/2002/WD-xml-names11-20020905/WD-xml-names11-20020905 >>-diff.html for your reviewing convenience. > >> The XML Core WG welcomes comments from all during the Last Call >> period, but we particularly request that the follow WGs commit >> to providing whatever feedback they have by September 28th: >> >> * XML Schema >> * XML Query (already committed) >> * XSL (already committed) >> * DOM >> * I18N >> * HTML (already committed) >> * RDF Core > >Does the fact that xmldsig is not designated mean you are already confident >that the introduction of Namespaces 1.1 will not affect canonicalized forms >of XML? Or that new versions of those specifications (and XPath even) are >needed to accommodate the change? Do you expect NS1.1 to be used with >XML1.0, or only XML1.1? On quick review, I believe that because of the >existing normative references, and the change that a namespace also applies >to attributes (section 5.1) this would require new versions of XPath, >c14ns, and xmldsig. The XML Core WG welcomes feedback from all groups. The omission of xmldsig shouldn't be read to imply much of anything (other than I'm required to have a list of groups from which we request feedback, and that's the list I dug up from somewhere else). Namespaces 1.1 is expected to be used with XML 1.1. Other than various errata, Namespaces 1.1 really only adds the ability to "undeclare" a namespace, and I doubt it would require any change to XPath et al. I think you are misunderstanding "the change that a namespace also applies to attributes." The wording change in section 5.1 (see Erratum NE07 at [1]) clarifies that the attributes in a start tag that contains a namespace declaration are also within scope of that namespace *declaration* (aka the prefix-to-namespace name mapping). This has always been the case; this erratum is merely editorial and does not change anything about what namespace the attributes are in. Specifically, "default namespaces do not apply directly to attributes" (Section 5.2). [Perhaps section 5.1 would more properly be titled "Namespace declaration scoping" or "Namespace prefix-to-namespace name mapping scoping".] paul [1] http://www.w3.org/XML/xml-names-19990114-errata#NE07
Received on Monday, 9 September 2002 17:45:31 UTC