- From: Geoffrey M. Clemm <gclemm@tantalum.atria.com>
- Date: Thu, 8 Oct 1998 15:23:51 -0400
- To: francis@netscape.com
- Cc: w3c-dist-auth@w3.org, meta2@net.lut.ac.uk
- Message-Id: <9810081923.AA13672@tantalum>
<geoff> The first is: Should there be an XML DTD for WebDAV properties? I believe the answer is "yes", and not only should complex properties be describable in XML, <john> WebDAV properties are already represented in XML (when they go over the wire, at least Yes, the prop XML element defined in section 11.11 ... <john> --there's an ongoing, nicely irrelevant, dispute over whether properties *are* XML :-). I *think* the dispute was on the update protocol for this information (which is a relevant dispute) rather than on the storage format of properties on the server (which would be irrelevant in a protocol discussion). I'll admit that it was at times hard to tell ... (:-). <geoff> but an entire object, even a collection object, should have an XML representation. <john> What advantages would this representation have over the existing MIME representation? It makes it easier to do incremental update in a simple uniform way. Without the XML DTD for collections, it would not be feasible to use an XML-based incremental update protocols for incremental update of collections. I'd guess the same is true for searching protocols, but I'll let the DASL folks comment on that. Let me re-emphasize that I'm not suggesting that we replace Mime with XML, but rather that we define the appropriate set of XML elements so that various XML-based protocols (such as incremental update and search) can be applied to collections. <geoff> The second is: Should the only way to update a WebDAV object be by specifying the full XML description of the new object? <john> Incremental updates would be nice, but we don't need them in the base protocol, since they can be replaced (less efficiently) via a sequence of LOCK/PROPFIND/PROPPATCH/UNLOCK. I'm not as concerned at what level of the protocol they are defined (base or otherwise), as I am that they *are* defined, since incremental update is needed for the protocol to be useful for large collections. And this leads to the desire to do incremental update in a uniform fashion (similarly, for searching). Cheers, Geoff
Received on Thursday, 8 October 1998 15:23:55 UTC