Re: RDF Description Services - and WebDAV

I went back and refreshed my apparently faulty memory on the WebDAV
protocol, and need to make the following corrections to myself:

> The WebDAV protocol has rudimentary
> operators for getting resources that have properties set to a specified
> value.

Rather, you can find resources that are associated with a particular named
property, not a value.  There are several caveats to this, including access
rights and collection membership.  You can also list all properties defined
on all resources in a given collection, again with caveats about access
rights, lock status, etc.

> For example, the side-effect of doing a GET
> request is that you can also get all the properties associated with that
> resource.
> 
> Request:
> 
> ... <!- ellipses indicates omitted protocol boilerplate -->
> GET http://www.w3.org/~danbri HTTP/1.1
> ...
> 
> Reply:
> ...
>     <D:prop>
>         <dc:creator>Dan Brickley</dc:creator>
>     </D:prop>
> ...

This is wrong in several ways.  First, I can't find any proof in the
protocol spec that GET behaves in this way (that's what I get for going by
memory only).  The correct request is PROPFIND.  You either explicitly name
the resource you are interested in, or some ancestor "collection". In either
case, you can use parameters in the request to look for a specific property,
to get property names without their values, to control the "depth" of search
in the resource tree hierarchy of a collection, etc.

So my point about resources being implicit isn't quite right.  All property
requests are relative to some resource or some collection of resources. 
Still, the association of resource to property is still somewhat less
explicit than for RDF.

Perry

Received on Tuesday, 8 February 2000 15:10:02 UTC