[Bug 227] Collection state definition in conflict between BIND and RFC2518bis

http://ietf.cse.ucsc.edu:8080/bugzilla/show_bug.cgi?id=227

lisa@osafoundation.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|lisa@osafoundation.org      |elias@cse.ucsc.edu
             Status|ASSIGNED                    |NEW



------- Additional Comments From lisa@osafoundation.org  2006-02-10 17:16 -------
Wow, I think I understood the meaning of the last paragraph there for the first
time when I tried to restate it.  I think I finally got it when I imagined a use
case -- a collection might have a dynamically generated HTML child which is not
a valid WebDAV resource, but is still a valid HTML resource.   Thus I tried to
illustrate the paragraph with such a use case.

So this is how I thought to explain it:

     Collection resources MAY have internal members with mappings to 
      non-WebDAV compliant children in the HTTP URL namespace hierarchy 
      but are not required to do so. For example, if the resource X with URL
      "http://example.com/bar/index.html" is not WebDAV compliant and the
      resource with URL "http://example.com/bar/" identifies a collection, 
      then collection "bar" might or might not have an internal member with 
      a mapping from "index.html" to the resource X.  If the collection 
      doesn't have such an internal member, presumably the consequence is 
      that the "index.html" resource might not show up in PROPFIND responses, 
      might not be locked when the collection is locked, might not have 
      WebDAV properties, and so on.

It seems this wide range of behaviorts might be harmful to interoperability.
What if the server decided to list the resource in PROPFIND responses but didn't
give it properties, or left it unlisted but forbade PUT requests to the binding
segment, and so on.

Have we seen any WebDAV servers which do this, which have bindings to non-WebDAV
members?  Is it interoperable?  If not, should we discourage this behavior?

Assigning to Elias so he can schedule discussion if necessary.



------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.

Received on Saturday, 11 February 2006 01:16:54 UTC