W3C home > Mailing lists > Public > public-xg-webid@w3.org > November 2011

Re: VOTE CONCLUSION - was: 2 resolutions, please vote

From: Stéphane Corlosquet <scorlosquet@gmail.com>
Date: Tue, 22 Nov 2011 11:51:40 -0500
Message-ID: <CAGR+nnHrwDKxFbZER4-yn28N3YGGWr=mHqdnQ9Ny=TLMmw+RWw@mail.gmail.com>
To: Henry Story <henry.story@bblfish.net>
Cc: WebID XG <public-xg-webid@w3.org>
Hi Henry,

On Tue, Nov 22, 2011 at 4:24 AM, Henry Story <henry.story@bblfish.net>wrote:

> Thanks for the votes.
>
> So there is pretty unanimous support for both proposals.
>
> 1) xsd:hexBinary
> ----------------
>
>  has been argued in detail so I think that is established. It also fits in
> nicely with the
> XML dsig crypto usage -  http://www.w3.org/TR/xmldsig-core/  - which uses
> binaries. That means our ontology could be thought of as the
> GRDDLable ontology of that spec. We use a hex encoding of binary but they
> use a base64 one.
>
> QUESTION: is it on the cards that the SPARQL working group add an
> identification of base64 and hex binaries? (It need not be this year, it
> can be in the next years) Ie: that a query made in one encoding should be
> true when made on a graph with the same binary encoded in the other
> encoding? This would be very handy. Can someone take on the task of passing
> that need on to that group?
>
> 2) Merge RSA into cert ontogy
> -----------------------------
>
> This had less debate here so I think there are a few questions I have on
> that subject, before I proceed. The reason I put them in different
> namespaces was that I felt writing
>
>  key  rsa:modulus ...
>      rsa:exponent ...
>
> made more sense than writing
>
>  key cert:modulus ...
>
> but perhaps that is more visible with DSA and their one letter names p q
> ...
>
>  key dsa:p "...";
>      dsa:q "...";
>
> immediately explains things whereas cert:p, cert:q a bit less...
>
> Now was the reason that people did not want these to be in the same
> namespace simply that as we know having too many namespaces
> would be annoying to large providers?


Web developers get confused with multiple namespaces. It's best to keep
them to a minimum. Having just one namespace allows to use new features of
RDFa 1.1, namely @vocab and the initial context, which defines prefixes for
the namespaces used in the spec like foaf, owl, rdf are all in the RDFa
Core Initial Context [1] so they do not need to be declared in the HTML
document. This is how you could write WebID with RDFa 1.1, using the vocab
attribute which defines the default namespace for terms:

<div vocab="http://www.w3.org/ns/auth/cert#" about="#me"
typeof="foaf:Person">
  <span property="foaf:name">Joe</span>
  <div rel="key">
    <h2>My RSA Public Key</h2>
    <div typeof="RSAPublicKey">
      <dl>
      <dt>Modulus (hexadecimal)</dt>
      <dd property="modulus" datatype="xsd:hexBinary">
        00cb24ed85d64d79...
      </dd>
      <dt>Exponent (decimal)</dt>
      <dd property="public_exponent">65537</dd>
      </dl>
    </div>
  </div>
</div>

of course you can alternatively define cert: and use CURIEs all the way.


>  Is that does solved with RDFa profiles?
> http://www.w3.org/TR/2011/WD-rdfa-core-20110331/#s_profiles ?
>

RDFa profiles were dropped from the RDFa spec.

Steph.

[1] http://www.w3.org/2011/rdfa-context/rdfa-1.1.html


>
> I'll write out the ontology in more detail for RSA and DSA today ignoring
> the details in the meantime.
>
> Henry
>
>
> On 21 Nov 2011, at 17:13, Henry Story wrote:
>
> > The WebID minutes have been published here:
> >
> >  http://www.w3.org/2011/11/21-webid-minutes.html
> >
> > So we have two RESOLUTIONS, but we would of course like some wider
> approval by implementors
> >
> > -  use XSD:hexBinary for modulus
> >    + this makes it possible to express the verification in one very
> efficient SPARQL ASK query
> >    + it is a well known schema widely supported by all the tools
> >    + the UI part can be dealt with using javascript : if someone comes
> up with some nice javascript to make it look good,
> >      I would be very tempted to add it to the spec
> >
> > -  merge rsa into cert ontology
> >    background:
> >     if we use xsd:hexBinary then there is only one relation that is
> widely used in the cert ontology the cert:key relation
> >     the rsa ontology has only two relations that are used
> >   => so we might as well merge rsa into the cert ontology
> >
> >  Do we have wider support for this?
> >
> >       Henry
>
> Social Web Architect
> http://bblfish.net/
>
>
>
Received on Tuesday, 22 November 2011 16:52:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 22 November 2011 16:52:11 GMT