Re: Handles and PURLs

Hi Folks!

Butler, Mark wrote:
> 6. The concept behind PURLs and Handles is good, i.e. when a resource moves
> you don't need to worry about it. DNS already has a level of indirection
> built in, so why not do this for retrievable resources? This is discussed in
> the Stone paper cited above.

JSE: This is the concept behind PURLs and is the simplest application of 
Handles, but it is NOT the concept behind Handles.

Handles are about unambiguously and persistently naming aggregations of 
properties. Period. Take the first property in Handle record, TYPE it as 
"URL" and assign a URL as the value, and you have the simple indirection 
that PURLS give you. Alternatively, enable the various TYPEs in a record 
be Handles themselves, and the values of properties to be literals or 
Handles, and you have an infrastructure for representing e.g. arbitrary 
graphs --- and certainly arbitrary representations of digital objects, 
which was BobK's point when he conceived the HS as a fundamental step 
toward realizing the Kahn/Wilensky architecture.

Couple the capabilities described above with HS's built-in mechanisms 
for administrative access control and authentication of assertions 
within the system, and it seems to have interesting potential as a 
robust metadata dissemination infrastructure.

John

Received on Thursday, 22 May 2003 08:16:52 UTC