Discussion: #rdfms-identity-anon-resources

All,

This note is my distillation of some points that are germane to the 
anonymous resource issue.  I plan to maintain a document that draws 
together elements of the discussion and any resolution that we may 
reach.  Below is a snapshot of that document.

In the first instance I would like to focus on the questions raised 
below.  In particular, it is not clear to me from the current M&S that 
(so-called) anonymous resources are anything more than a purely syntactic 
issue -- that a parser resolves by supplying a unique "genid".

#g
--

Issue statement
---------------
http://www.w3.org/2000/03/rdf-tracking/#rdfms-identity-anon-resources


Goals
-----
The following goals/purposes have been suggested for the use of anonymous 
resources:
- Avoiding the need for users to make up names for "intermediate" resources 
(e.g. for structured resource values).
- As a general expression of existentially quantified variables.
- As variables in query statements.

Discussion points
-----------------
RDF M&S clearly allows un-named resources in the XML serialization syntax 
for RDF. There is some discussion about whether it also specifies that 
resources in the abstract graph model may be anonymous, or whether the lack 
of a name in the XML syntax merely allows/requires the parser to make up a 
name (for which "genid" or "hash" values have been proposed).

QUESTIONS
---------

1. Are anonymous resources allowed in the abstract graph syntax?

2. If the answer to 1 is YES, what do they mean?

3. Can a graph containing an anonymous node be the same as some other graph?


Proposed resolution
-------------------
[[[TBD]]]


References
----------

[1] Issue statement: 
http://www.w3.org/2000/03/rdf-tracking/#rdfms-identity-anon-resources

[2] Arguments against digest URIs 
http://lists.w3.org/Archives/Public/www-rdf-interest/2000Jan/0001.html

[3] A discussion of anonymous resources: 
http://ioctl.org/rdf/discuss/anonymous

[4] DanC on resources as existentially quantified variables:
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0497.html,

[5] Pat Hayes' comments on the use of anonymous resources and queries: 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0522.html.  See 
also the ensuing (linked) messages in this thread.

[6] Frank Manola comments that the anonymity is not, of itself, necessarily 
introducing an existentially quantified variable: 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0326.html.   (He 
shows a similar structure for a named resource that uses an existential.)

[7] GK suggests that the use of anonymous resources/existential variables 
in a query can be viewed as a computational rather than a logical 
distinction: 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0474.html

[8] Frank Manola: what makes use of an existential into a query? 
http://lists.w3.org/Archives/Public/w3c-rdfcore-wg/2001Jun/0555.html.
DanC agrees. Frank follows up with the point that in the context of a 
query, the recipient of the query needs to be able to recognize the subject 
of the query ("do you have something that matches 'X'?"  It may matter 
little if 'X' is a Skolem constant or an existential variable - it is a 
distinguished part of the query).



------------
Graham Klyne
(GK@ACM.ORG)

Received on Thursday, 5 July 2001 10:26:58 UTC