- From: Lisa Martin <lmartin@ca.ibm.com>
- Date: Wed, 3 Mar 2004 09:14:07 -0500
- To: jwrobie@mindspring.com
- Cc: public-qt-comments@w3.org, W3C XML Schema WG <w3c-xml-schema-wg@w3.org>
Hi Jonathan, Regarding comment 1b, we do not believe it's an editorial comment. The specific substantive request being made is that QT consider using schema context paths as the (no longer implementation-dependent) type ids for anonymous types. Thanks - On behalf of the XML Schema WG, Lisa Martin Jonathan Robie <jwrobie@mindspri ng.com> To Lisa Martin/Toronto/IBM@IBMCA 02/16/2004 05:49 cc PM public-qt-comments@w3.org, W3C XML Schema WG <w3c-xml-schema-wg@w3.org> Please respond to Subject jwrobie Re: [XPath 2.0] XSCH-XPATH-001 Lisa Martin wrote: >Dear Colleagues, > >This 2-part comment pertains to the Nov. 12 2003 version of XPath 2.0 [1]. > >[1] http://www.w3.org/TR/xpath20/ > >Lisa Martin, on behalf of the XML Schema Working Group >---------------------------------------------------- > >Section 2.1.1 Static Context > [Definition: In-scope type definitions. Each named type definition > is identified either by a QName (for a named type) or by an > implementation-dependent type identifier for an anonymous type. ... ] > > a. The use of "Each named type definition is identified either..." >implies that anonymous types are considered "named type definitions" in >this specification. Is this correct? If so, then constructor functions >are defined for anonymous types - was that intended? If not, the first >use of "named" in the definition should be dropped. > > Hi Lisa, I think the placement of parentheses is significant here: > [Definition: *In-scope type definitions.* Each named type definition > is identified either by a QName (for a *named type*) or by an > implementation-dependent > <http://www.w3.org/TR/xpath20/#dt-implementation-dependent> type > identifier (for an *anonymous type*). The in-scope type definitions > include the predefined types as described in *2.4.1 Predefined Types* > <http://www.w3.org/TR/xpath20/#id-predefined-types>. ] So there are two kinds of types - named types and anonymous types. >b. WRT implementation-dependent type ids for anonymous types, we note that >elsewhere (schema context path) QT defines names for anonymous types. Would >it be appropriate to mandate their use in this case? We also note that >discussions are ongoing between two WGs about harmonizing schema context >paths and SCDs. > > These are not "names" in our terminology - they are expressions used for matching types. Is there a specific request here? For now, I am classifying this comment as editorial. Jonathan
Received on Wednesday, 3 March 2004 09:14:43 UTC