- From: Larry Masinter <masinter@parc.xerox.com>
- Date: Fri, 28 Aug 1998 10:59:52 PDT
- To: "Webdav" <w3c-dist-auth@w3.org>
- Cc: "Keith Moore" <moore@cs.utk.edu>, "Patrik Fältström" <paf@swip.net>
The 'model' in draft-ietf-webdav-protocol-07.txt is ambiguous as to whether each resource is required to have a single URL which locates that resource. At the working group meeting, there seemed to be some disagreement about this point; Jim seemed to think (emphatically) that it was required that the mapping between 'resource' and 'URL' was 1-1, but others seemed to assert that it might be 1-many. However, the interpretation of the model of the DAV spec depends significantly on this choice. The language of the spec is consistent with '1-many', but in that view, DAV is missing a critical operation: obtaining (one or more of) the parent container(s) for a resource. If the interpretation is that DAV requires 1-1 mapping of resources and URLs, then it is necessary, I believe, to have an operation in DAV to obtain the "DAV-canonical URL" for a given resource, and to insist that DAV clients always first translate the resource they are asked to act on to the DAV-canonical one before performing any subsequent operations. I think it is imperative that this ambiguity in the specification be resolved, and that either 'fix' to the protocol be applied. Notes that in the 1-many interpretation, it should be made clearer that properties are associated with resources rather than with their URLs (e.g., that a PROPPATCH on one URL associated with a resource will affect the PROPFIND returned by the other URL). Also, some of the language about DELETE with depth would need to be clarified, since it's written in a style that assumes that there is only one path to a given resource. If the model requires 1-1, this should be spelled out as well, and some explanation of how canonicalization is applied should be supplied as well. (For example, a server that is case-insensitive probably needs to automatically canonicalize the URLs of the resources that it is asked to author, since the 1-1 model requires the client to do string manipulation of resources' URLs to determine whether a collection is or is not a parent of another resource. Larry -- http://www.parc.xerox.com/masinter
Received on Friday, 28 August 1998 14:00:17 UTC