Re: RDFCore 2nd last call

Hello Brian,

Many thanks for your mail.

At 17:16 03/08/22 +0100, Brian McBride wrote:

>Richard, Martin,
>
>After some discussion today, whilst no formal decision was taken, the 
>RDFCore chairs indicated they "were minded" that there should be a second 
>last call on the RDFCore specs.
>
>We are aware that the I18N are unhappy with the current RDFCore design for 
>handling lang tags in xml literals.

Do you have any idea of what the RDF Core WG might be going to do with it?

Btw, a page with the documentation of our formal objection is taking shape
at http://www.w3.org/International/Group/2003/rdf.html. [This is currently
member-only.]


>Are there any other aspects of the current design as expressed in the 
>current editors drafts, for example the relationship between plain and xml 
>literals, that you would you like RDFCore to address.

We will check during our teleconference tomorrow (Tuesday).


>It would clearly be helpful to us to be able to address any such concerns 
>before publishing new documents.

Here is what I managed to remember over the weekend, and what I'll
bring into our teleconference tomorrow:

- There was the issue of casing in language tags. Tex submitted a comment
   from us, you clarified the spec, and we accepted the clarification
   and told you so, so this should be okay.

- I read the Primer, and handed my well-marked copy of it over to Eric.
   He has confirmed that to a large extent, my comments were addressed
   in recent edits. Eric plans to bring the marked-up copy back to MIT,
   which will allow me to check the details. Some more work on some details,
   in particular i18n-related examples (which were extremely sparse in
   the last call version) may be needed, but I don't think this should
   hold you up in any way.

- I remember having seen a grammatical mistake (wrong gender) in a
   German example in one of the documents, but can't recall just now
   which document it was. This is easy to fix.

- There was the issue of always escaping non-ASCII characters in N-triples.
   We have sent that to you very late, and you have accepted it only as a
   post-lastcall comment. I don't remember just now how much you have
   done on this issue, but given the circumstances of our last comment,
   it's probably not much. I'm personally not inclined to resubmit this
   issues, but I know that others in the WG may feel otherwise (and this
   is an issue that's much easier to understand than than xml:lang on
   XML Literals).

- There is the issue of equivalence between plain literals and XML
   Literals. On this issue, we have discovered that this equivalence
   doesn't hold rather late in the process, and have therefore not
   insisted on it (apart from the confusing equivalence with octets
   that turned up at some point). I think that a lot of progress has
   been made on this issue, so it could be left as is, but with a new
   last call, we may want to consider whether we should bring it up
   and ask for full resolution.

- We may have to think about whether we should register an issue
   with regards to the treatment of XML Literals as types. When
   I read some of the last call documents, I found this treatment
   odd, but thought that it may be okay as a technical device as
   long as it didn't have any undesirable side-effects.


I hope that this is the kind of information you were looking for.
If not, please feel free to tell me.


Regards,    Martin.

Received on Monday, 25 August 2003 18:42:48 UTC