- From: Felix Sasaki <fsasaki@w3.org>
- Date: Wed, 25 May 2005 21:16:22 +0900
- To: w3c-svg-wg@w3.org
- Cc: public-i18n-core@w3.org, Felix Sasaki <fsasaki@w3.org>, eb2m-mrt@asahi-net.or.jp
Dear SVG Working Group, Considering our first comment on SVG Tiny 1.2 [1], we had some feedback from Murata Makoto, I have attached his mail below. In my (personal) opinion, you should make it a requirement for the validation process of SVG Tiny 1.2. that the processor relies on a data type library which implements XML 1.1, see below "(implementation) Jing (2) attribute values or element contents". Best, Felix. [1] http://www.w3.org/International/2005/05/svg-tiny-review.html >Felix, > >Thank you for your mail. Here is my understanding of the >current status. > >--------------------------------------------------- >1. XML1.1-names in RELAX NG schemas > >Spec: Not allowed. > >Implementation(jing): Not allowed. > >2. XML1.1-names in XML documents > >Spec: Not allowed > >Implementation (jing) > >(1) element or attribute names > >Jing assumes that the XML parser tests well-formedness of element and >attribute names. Thus, <anyName/> matches every XML 1.1 names. > >(2) attribute values or element contents > >Jing simply uses implementations of XML Schema Part 2. Thus, ><element name="foo"><data type="NCName"/></element> matches ><foo>・</foo> if the underlying WXS2 implementation >allows ・ as a name character. > >----------------------------------------------------------------- > >I do not think W3C would like to publish schemas containing non-ASCII >names. If this assumption is true, I think that all what you actually >need is implementations of XML Schema Part 2 that allow XML-1.1-names. > >If we would like to write RNG schemas using XML1.1-names, we have >to create a new version of RELAX NG and change implementations >accordingly. > >Hope this helps. > >Cheers, > -- MURATA Makoto <EB2M-MRT@asahi-net.or.jp> >
Received on Wednesday, 25 May 2005 12:16:33 UTC