- From: Dave Reynolds <der@hplb.hpl.hp.com>
- Date: Tue, 04 Dec 2007 15:00:44 +0000
- To: Christian de Sainte Marie <csma@ilog.fr>
- CC: RIF WG <public-rif-wg@w3.org>
Christian de Sainte Marie wrote: > I assumed that the specification for Const with the XML:lang tag would > look something like (in the style of the definition I used in PRD): > > The `Const` element has a required `type` attribute and an optional > `xml:lang`attribute: > - The value of the `type` attribute is the identifier of the variable > symbol space. It must be an absolute IRI > - The value of the `xml:lang` attribute is the identifier of the > language for the presentation of the Const to the user. It should one of > the two-letter codes defined in ISO-639. The attribute MUST be omitted > if the type of the Const is not rif:text (alternatively: a compliant > implementation MUST ignore the xml:lang attribute...; or anything to > that effect). > The content of the `Const` element is the constant's name, which can be > any Unicode character string. > > XML Syntax (in pseudo-BNF): > > <Const type="IRI" xml:lang="language identifier"> > Unicode character string > </Const> > > WHat would be wrong with something like that? Quibbles aside that's broadly what I'd expect. However, first that's not compatible the proposed resolution is it? That supposes that the lang element is a separable attribute of the const whereas the proposed resolution seems to say that it is part of the const string itself. Second the phrase "Const's name" is confusing. This is about how to encode text literals, the string content is not it's name it is its value. [This does relate to the notion of presentational names. For example for rules I would expect the rule name metadata to permit multiple rif:text values to support l10n (and I'd recommend using rdfs:label as the metadata property URI). ] Dave -- Hewlett-Packard Limited Registered Office: Cain Road, Bracknell, Berks RG12 1HN Registered No: 690597 England
Received on Tuesday, 4 December 2007 15:01:05 UTC