W3C home > Mailing lists > Public > www-svg@w3.org > October 2008

Re: [1.2T-LC] Comments on Last Call WD of SVG T1.2 ISSUE-2130

From: Chris Lilley <chris@w3.org>
Date: Mon, 27 Oct 2008 14:41:49 +0100
Message-ID: <1366921263.20081027144149@w3.org>
To: www-svg@w3.org

Hello Cyril, www-svg,

This is the official response to your issues which have been discussed by the SVG WG.


> * Section 4 "Basic Data Types"
> The id and xml:id attributes use the <ID> type in the spec but the <ID> or <NCName> types in the attribute table. 

Yes. Following resolution of another issue about xml:id and RelaxNG,
http://lists.w3.org/Archives/Public/public-svg-wg/2008OctDec/0225.html

it has been resolved that the SVGT1.2 RelaxNG should say that xml:id is of type NCName. This is because RelaxNG does not alter the infoset, and xml:id processing is independent of the schema.

So the attribute table (which is produced from the schema) is correct. The Spec is correct, in the case where either id or xml:id (but not both) are specified on an element. The spec says what happens after xml:id processing.

In summary, yes, they say slightly different things; one tells you what the schema says, and the other tells you what the types
 
> In the definition of the <QName> type, a sentence says: "the default namespace is *not* used for unprefixed names". Could ou provide an example showing what this means ?

This is the usual processing for QNames in attribute values, and follows that of XML Namespaces for qualified and unqualified attributes. Unlike element names, the default namespace is not used for unqualified attributes. Since the one use of QNames in attribute values, in the SVGT1.2 spec, is the 'attributename' attribute, this is appropriate.

SVG WG follows the TAG finding on QNames as Identifiers in XML Content
http://www.w3.org/2001/tag/doc/qnameids.html#bindings

which states

  Specifications that use QNames to represent {URI, local-name} pairs MUST describe the algorithm that is used to map between them.

The SVG WG has reviewed the sentence in the spec and reworded it slightly to improve clarity. It now says:

"If the <QName> has a prefix, then the prefix is expanded into a tuple of an IRI reference and a local name, using the namespace declarations in effect where the name occurs. Note that, as with unprefixed attributes, the default namespace is not used for unprefixed names. "

We hope this is clearer. Please let us know if this does not clarify the issue for you.
 
> The spec text gives the following definition for the target attribute of the 'a' element:
> target = '_replace' | '_self' | '_parent' | '_top' | '_blank' | "<frame-target>" 
> but the attribute table says:
> target = '_replace' | '_self' | '_parent' | '_top' | '_blank' | <XML-Name>

frame-target is of type XML Name. We have corrected the spec to say XML-Name and added a link to the types chapter.

> The spec text for handler says:
> ev:event = "<string>"
> but the attribute table says:
> ev:event = <XML-NMTOKEN>

> The spec text for listener says:
> event = 'Event Identifier'
> but the attribute table says:
> event = <XML-NMTOKEN>

We have changed both of these to XML-NMTOKEN in the spec, and linked to the definition in the types chapter.

Please let us know whether these corrections address your concerns.

This relates to ISSUE-2130
Basic Data Types section needs clarifications  

-- 
 Chris Lilley                    mailto:chris@w3.org
 Technical Director, Interaction Domain
 W3C Graphics Activity Lead
 Co-Chair, W3C Hypertext CG
Received on Monday, 27 October 2008 14:42:01 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:40 GMT