W3C home > Mailing lists > Public > www-tag@w3.org > November 2002

RE: Namespaces 1.1 Last Call -- I18N WG comments

From: Richard Tobin <richard@cogsci.ed.ac.uk>
Date: Fri, 29 Nov 2002 23:27:07 GMT
Message-Id: <200211292327.XAA09022@sorley.cogsci.ed.ac.uk>
To: Misha Wolf <Misha.Wolf@reuters.com>, Richard Tobin <richard@cogsci.ed.ac.uk>
Cc: w3c-i18n-ig@w3.org, xml-names-editor@w3.org, www-tag@w3.org

> > [Definition: IRI references which identify namespaces are considered
> > identical if and only if they are exactly the same
> > character-for-character.] This is described in greater detail, with
> > examples, in B Comparing IRI References.
> 
> The definition of "exactly the same character-for-character" does *not*
> belong in an appendix.  In particular, the definition should make clear 
> that the string comparison should be preceded by the steps described in:
> 
>    3.3.3 Attribute-Value Normalization
>    http://www.w3.org/TR/REC-xml#AVNormalize

As I mentioned in another message, not all namespace names that get
compared come from attribute values.  The description has to work, for
example, when comparing the IRI in 

 <xs:stylesheet version="1.0" xmlns:xs="http://www.w3.org/1999/XSL/Transform">

with the one specified in the text of the XSLT specification.

The issue of how you obtain the IRI from the attribute is separate from
the issue of how IRIs are compared once you have them.

> > In a namespace declaration, the IRI reference is the normalized value
> > of the attribute, so replacement of XML character and entity
> > references has already been done before any comparison.
> 
> Three problems here:
> -  it seems to be the first time normalization is mentioned
> -  it is mentioned in passing
> -  the type of normalization isn't specified

It's also now mentioned in the "Declaring Namespaces" section:

  The attribute's normalized value, must be either an IRI reference --
  the namespace name identifying the namespace -- or an empty string.

"normalized value" there is a link to section 3.3.3 of XML 1.x.

We will also be publishing an erratum to 1.0, since it incorrectly
says "value" rather than "normalized value", and has a link to the
unnormalized value in XML 1.0.

I could make the term be a link in the appendix too.

> >   http://www.example.org/wine
> >   http://www.example.org/Wine
> 
> It would be preferable to use the same examples as below, eg:
> 
>     http://www.example.org/rosé
>     http://www.example.org/rosÉ

I disagree; I think it is better to show the case difference without
confounding it with the issue of non-ascii characters.  And I don't
want *all* the examples to be IRIs that are not also URIs, especially
since we're advising people not to use non-URIs yet.

-- Richard
Received on Friday, 29 November 2002 18:27:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:13 GMT