W3C home > Mailing lists > Public > w3c-dist-auth@w3.org > April to June 1998

Re: DAV properties and advanced collections

From: Jim Davis <jdavis@parc.xerox.com>
Date: Sun, 14 Jun 1998 15:22:35 PDT
Message-Id: <3.0.3.32.19980614152235.0080a100@mailback.parc.xerox.com>
To: w3c-dist-auth@w3.org
At 01:55 PM 6/10/98 PDT, Judith Slein wrote:
>I have some general confusions about DAV: properties.
>
>1. Can a client create a DAV: property using PROPPATCH?  Could a client
>create DAV:ordering on a collection at a location that doesn't comply with
>the ordered collections part of WebDAV?  Could a client create some DAV:
>property not defined in any WebDAV specification?

Nothing in WebDAV prohibits it, so I would say yes to all three.  In
general, dav properties are read-only, but nothing says that *all* dav
properties must be read-only.  

If your concern is that a client might set a property, expecting it to be
live,  on a server that does not understand it?

>2.  Copy / Move behavior for properties in general depends on the value of
>propertybehavior in the request.  Does this apply to DAV: properties?  What
>would happen if the request specified "omit" and DAV:reftarget was present
>on the source resource, but the target location is on a basic DAV server?
>Would the DAV:reftarget property be missing from the new resource, or would
>it be present but dead?  What would happen if there was no propertybehavior
>specified? Would DAV:reftarget just become a dead property at the new
>location?

I think it would become dead.  The result would be zero-length resource but
it would presumably have both the dav:resourcetype set to dav:refmember and
the dav:reftarget set to the target.    This should work fine.

On the other hand, doing a  COPY that explicitly allowed dav:refintegrity
to be omitted would mean one was taking a risk.  No reason to forbid it
though.
Received on Sunday, 14 June 1998 18:24:58 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:43:47 GMT