Re: slight change to http://www.w3.org/TR/rdf-primer/#rdfmodel

Eric--

Regarding the comments you made about the RDF Primer, I need to decide 
what to do about them, and for that I need some further information from 
you as I've indicated in my reply to your message (reproduced below). 
In particular, do you understand my replies and do they make sense?  How 
important do you think this change to the Primer is?  If it's 
sufficiently important, I will raise it as an issue for the Working 
Group to consider, but it's going to be hard to decide without your 
further input.  Thanks.

--Frank

Frank Manola wrote:

> 
> Eric Prud'hommeaux wrote:
> 
>> The RDF primer model example [1] makes an assertion that has
>> [[
>> a subject http://www.example.org/index.html
>> a predicate http://purl.org/dc/elements/1.1/creator
>> and an object http://www.example.org/staffid/85740
>> ]]
>> I request changing http://www.example.org/staffid/85740
>> to
>> http://www.example.org/staffid#8570
>> per the concepts document [2].
> 
> 
> 
> Eric--
> 
> Thanks for the comment.  I'm reluctant to make this change, for several 
> reasons (which I'll describe below).  I'm also a bit unclear as to the 
> background behind the request, so I need a little more information from 
> you about this.
> 
> I'm reluctant to make the change because, even if we restrict the change 
> to the staffid/85740 usage, this doesn't just appear in one example; 
> it's used throughout the text of Section 2 (it also appears in Section 
> 4), so the changes required would be rather extensive.  (If the idea 
> behind this is that *all* usages of plain URIs should be replaced by 
> usages of URI#fragment, the changes become even more extensive, 
> involving much of the Primer).
> 
> Such changes would also, in my opinion, be somewhat misleading.  RDF 
> URIrefs can contain fragments, but they don't need to, and the Primer 
> deliberately has a mixture of usages of both kinds of URIrefs for that 
> reason (a number of the examples in Section 3 use URIrefs that include 
> fragments, as do examples in Section 5).  The Dublin Core property 
> URIrefs, for example, don't use fragments, and we can't really change that.
> 
> I'm not sure what to make of your reference to the Concepts document. 
> That section of the Concepts document describes a way of interpreting 
> RDF URIrefs that *do* include fragments, but I don't see it as mandating 
> that they *have* to contain fragments (you might also look at what the 
> Primer Appendix A says about fragments).
> 
> I'm also not sure what to make of your comment below.  What ambiguity of 
> URIs (without fragments) are you referring to?  It seems to me that a 
> URI is unambiguous (at least as an opaque name, the way RDF uses it). 
> Also, if a URI without a fragment is ambiguous, how can it become less 
> ambiguous by adding a fragment to it?
> 
> --Frank
> 
> 
>>
>> While the exact semantics of URI#fragment are still subject to debate, I
>> believe they are considered less likely to be ambigous than URI alone.
>>
>> [1] http://www.w3.org/TR/rdf-primer/#rdfmodel
>> [2] http://www.w3.org/TR/rdf-concepts/#section-fragID
>>
> 
> 


-- 
Frank Manola                   The MITRE Corporation
202 Burlington Road, MS A345   Bedford, MA 01730-1420
mailto:fmanola@mitre.org       voice: 781-271-8147   FAX: 781-271-875

Received on Friday, 28 February 2003 10:49:17 UTC