W3C home > Mailing lists > Public > www-rdf-comments@w3.org > January to March 2001

Re: RDF comments from the ages

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: Tue, 13 Mar 2001 15:46:56 +0000
Message-ID: <3AAE40F0.72F63C0C@hplb.hpl.hp.com>
To: Aaron Swartz <aswartz@upclink.com>
CC: RDF Comments <www-rdf-comments@w3.org>

You clearly have been very busy :)  It's taken me most of the day to go
through this lot.

I have detailed below the actions I've taken.  An updated version of the
issues list will appear in due course.


Aaron Swartz wrote:
> Here's what I get from trawling through RDF-DEV and RDF Interest for all of
> 1999:
> Spec treats email addresses as literals, not URIs
> (probably should be added to errata)
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-03/0002.html

Whilst email addresses can be represented as URI's, I don't see why it
is an error when they are not.

> The first known instance of the set or bag question
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-04/0001.html
>     DanBri cites the use of set terminology as evidence towards a set:
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-04/0003.html
>     EricP says that in a multi-user system, it should be a bag (sorta):
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-04/0007.html
>     Bleed-over from XML-L:
>     http://listserv.heanet.ie/cgi-bin/wa?A2=ind9910&L=xml-l&D=0&P=13979
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-10/0019.html

The issue:


has been generalised to cover the absence of any formal description of
RDF graphs in the spec.

> No parseType="Resource" examples in the spec
> (another errata?)
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-06/0010.html

I have added a reference under editorial.

> First known instance of Qname->URI problems:
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-07/0012.html

This issue is already present as:


>     conformance problem with xmlns I hadn't seen before:
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-07/0013.html

I don't see a problem here.  RDF has no unique attribute requirement.

> How do you constrain members of a container in RDFS?
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-07/0015.html

This issue is already present as:


> Does a statement imply its reification?
> (I don't think so, but others seem to agree that it does)
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-07/0028.html

I believe the current understanding is clear that the presence of a
statement and its reification in a model are independent.

> Can properties have no value?
> http://www.mailbase.ac.uk/lists/rdf-dev/1999-08/0001.html
>     RRS says use a genid:
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-09/0015.html
>     Cowan says use empty literal:
>     http://www.mailbase.ac.uk/lists/rdf-dev/1999-09/0016.html

Interesting that this should come up again recently.  I'll await a
response to my response to the recent Sirpac Bug thread.

> First known aboutEachPrefix complaint
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0068.html

This issue is already present as:


> xml:lang not in the model
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0068.html

This issue is already present as:


> rdf:ID should create isDefinedBy triples
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0068.html
> Do nodes exist? Even if they aren't attached to anything?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0079.html
>     Guha: Yes, nodes exist.
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0083.html

Actually, the message doesn say 'should', it says 'could'.  Added a
reference to this message under:


> #rdfms-logical-formalism
> Guha: RDF is a very proper subset of predicate logic.
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0083.html
>     Melnik: I proved that the most general RDF model does not exist.

I think Sergey's claim is that he proved it does exist.

>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0085.html
>     Formal algebra in:
>     http://www-diglib.stanford.edu/diglib/ginf/WD/rdf-alg/rdf-alg.ps

This is further discusion about the nature of the RDF graph.

> We need a standard way to talk about the source of triples:
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0089.html
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0097.html
> http://www-db.Stanford.EDU/~stefan/updates.html

I don't see a specification issue here.  Reification can be used to
model the source of statements.

> Is pointing to a fragment of a document the same as point to that literal?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0073.html

At the RDFCore level, clearly not.  They are different triples.

>     Renato: I wanted to distinguish between identifers and reference
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0091.html
>     DanBri: It doesn't solve the issue, need more properties (rdfs:seeAlso)
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0093.html
> What does a URI mean? / Linking to resources
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0106.html
> http://www-db.Stanford.EDU/~stefan/updates.html
>     Is URI of a description the same as URI of what is being described?
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0102.html
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0107.html
>     Difference between URI as string, URI as derefed bits, URI as object
>     http://lists.w3.org/Archives/Public/uri/2000Oct/0006.html

This has been touched on in a number of issues, e.g.


This is really a web architecture issue, however, I think we'd all be
better off we had a clearer formal model and terminology for URI's and
what they name, perhaps something similar to what the topic maps folks
have done.

I'm going to seek some advice on this one.

> How do we distinguish between resource entity and resource content?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0107.html
>     One suggestion:
>     <q cite="http://logicerror.com/semanticWeb-long">
>     Common practice for giving something a URI is to create a web page that
>     describes the object and explains that the URL of that webpage
>     represents it. For example, I created a URI for my copy of Weaving the
>     Web. Now I have said that that specific URI no longer represents the web
>     page you get back when you visit it, but instead the physical book that
>     it describes.</q>
> With RDF services, how do we distinguish between service itself and the data
> it returns?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0107.html
> What about URIs with more than one meaning? How do we deal with it?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0107.html
> What is mapping between URIs and resources? (DanBri)
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0025.html
>     WebDAV defines redirect resources -- are we happy?
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0125.html
> Is the U in URI for Uniform, Unique, or Universal?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0146.html
> Is there something special about "virtual resources" -- i.e. those described
> by RDF and which don't return a description of themselves?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0107.html
> #rdfms-literals-as-resources:
> Raised: Jonas Liljegren (jonas@paranormal.o.se)
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0107.html
>     also mentions:
>     - all dereferencable URIs should be treated as literals
>     - mime types should be subclasses of rdfs:Literal
>       (seeAlso #rdfms-literal-is-xml-structure)
>     genids should be used for literals:
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0120.html

The above are all about the nature of URI's.  See above.

> What's the skolem function?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Nov/0092.html
> http://www-db.Stanford.EDU/~stefan/updates.html
> http://www-db.stanford.edu/~melnik/rdf/api.html#digest

I'm going to check with Sergey on this one, and ask him to raise it
again on www-rdf-comments if he wishes it included on the issue list.

> Do anonymous resources have a URI?
>     DanBri: URI is just one bit of info about a resource
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0025.html
>     Two kinds of URIs (private and public):
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0028.html
>     Melnik: Agreed-upon URIs for anonymous resources are essential
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0032.html
>     DanBri: You can't do that -- you need more info:
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0033.html
>     DanBri: anonymity is not a property of a resource, but of a mention:
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0035.html
>     Melnik: Good summary of problem/solution:
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0037.html

Added reference to Sergey's summary.

> Are triples independent of the model (graph?) that they're in?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0041.html
> How do you assert something, but give the resulting triples a URI?
> (i.e. assert and reify at the same time)
> http://logicerror.com/decentralizedRDF
> http://lists.w3.org/Archives/Public/www-rdf-interest/2001Feb/0161.html
>     This makes time difficult to represent in RDF
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0149.html
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0150.html
>     Gabe: That's not true -- use rdf:ID
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0151.html
>     Another Issue: Wait -- the EBNF doesn't allow this!
>     http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0152.html

The possible grammar resttriction raised by Gabe is covered in:


> Does a rdf:Statement represent a stating or the abstract statement?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0068.html

I've added a reference to this message under


> Can't define acceptable properties for a class:
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0094.html

I'm not sure I see an issue here, given that anyone can say anything
about anything.

> Are properties inherited through subclassing?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0096.html
> Are properties of a class inherited by instances of a class?
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0097.html

I've added pointers to these to the schema editorial section.

> RDF mapping into Borges
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0137.html
>     Just kidding. ;-) Anyone still paying attention?

A new issue :)


> RDF model is property-centric, but syntax is resource-centric
> http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0148.html

An interesting observation, but I'm not sure I see an issue here.

> Issue for Brian: Can you make the issue IDs links to their URIs? Like:
>     Issue <a
>     href="http:...#rdfms-uri-for-graph">rdfms-uri-for-graph</a>: blah blah
>     blah...
> Well, that's enough issue trawling for me. That should cover you all thew
> way up to the beginning of 2000. Thanks for the blast from past -- it was
> quite interesting to see how far RDF came in a year, and how quickly it's
> been gaining steam. Yet, it's very odd to travel an entire year in one
> sitting!
> I hope these issues will be useful, and added to the list. Let me know if
> you have questions with what I wrote.
> --
> Aaron Swartz <me@aaronsw.com>|       The Info Network
>   <http://www.aaronsw.com>   |     <http://theinfo.org>
> AIM: JediOfPi | ICQ: 33158237| the way you want the web to be
Received on Tuesday, 13 March 2001 10:46:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:15:13 UTC