- From: Jim Whitehead <ejw@ics.uci.edu>
- Date: Thu, 19 Feb 1998 15:22:10 -0800
- To: "'WEBDAV WG'" <w3c-dist-auth@w3.org>
After considering the issues, the WebDAV Design Team has decided the best technical approach for achieving extensibility and avoiding collisions in WebDAV's use of XML is the W3C XML namespace proposal. The current name space proposal can be found at: http://www.w3.org/TR/1998/NOTE-xml-names A note on extensible languages by Tim Berners-Lee and Dan Connolly can be found at: http://www.w3.org/TR/1998/NOTE-webarch-extlang-19980210 The requirements of the WebDAV Distributed Authoring protocol for XML extensibility are: * WebDAV clients (i.e., multiple stakeholders) must be able to add XML elements to property values, where the XML elements have guaranteably unique names. * Future WebDAV clients and servers need a mechanism to add new XML elements to XML found in the protocol stream. To avoid collisions, these new elements must have guaranteably unique names. * The unique name is used by the stakeholder to identify the semantics of the XML elements for which they control the semantics. * Any stakeholder must be able to assign their own unique names without contacting a central registration authority. * These unique names should not dramatically increase the size of XML documents which use them. The XML namespace mechanism addresses the requirements nicely, since each stakeholder / extension author will have a URL namespace that they own, and can use to label the XML elements under their control. The namespace directive allows URLs to be used without dramatically blowing up the size of the XML documents. Alternative technical approaches to meeting these requirements typically involve extending a DTD with the new elements. This is less desirable than the namespace approach, since use of DTDs requires a more heavyweight XML parser, and because it implies the new elements are being added to an existing DTD, rather than being used from another DTD. Furthermore, the collision behavior of alternative approaches is not as good as for XML namespaces. One concern raised on this list is the dependency WebDAV has on XML namespaces, and the problems this raises with WebDAV referencing a non-Recommendation status W3C document. There is currently a flurry of activity working on XML namespaces, and very strong support within the XML community for XML namespaces. I have been in contact with Jon Bosak, the Chair of the W3C XML Working Group, and he writes: > I can report to you that the XML WG has set the following schedule for > staging the namespace discussion: > > 2/19 First group of issues to the SIG for discussion > 2/25 First group of issues decided by the WG > 2/26 Second group of issues to the SIG for discussion > 3/11 Second group of issues decided by the WG > 3/12 Third group of issues to the SIG for discussion > 3/18 Third group of issues decided by the WG > > 3/21-22 Decisions finalized at XML WG f2f in Seattle > > 3/25 WG review of first working draft > 3/27 Rollout of WD at XML Developers' Day (Seattle) > > 3/27-4/25 Four-week public review period, including WWW7 > > 4/27 Release of Proposed Recommendation > > Note that this does not constitute a guarantee that the work will > proceed as scheduled. Some of the issues to be decided are quite > complex, and it would be unproductive in the long run to put in place > solutions that have not been carefully considered. But this does > capture our intent. I have no objection to your posting the schedule > to the WebDAV list as long as you include this disclaimer. Based on this, we feel very confident that XML namespaces will become a stable reference in time for the WebDAV Distributed Authoring specification to use it as a dependency. Since there is *substantial* support for XML namespaces, with many products and standards efforts depending on them, WebDAV will reap many benefits from supporting standard XML namespaces. - Jim
Received on Thursday, 19 February 1998 18:44:54 UTC