Re: extra comments on test cases

Hi Ivan,

Mmm...I'm not sure what the amen is to, though.

We've never had support for the dynamic generation of bnode values
using one and the same CURIE--one CURIE always generates the same
bnode. Of course, I agree that you proposed such a thing as a
desirable feature, and I can see from Ben's email that he thinks that
this feature is already here. But I'm afraid that fact remains that it
is not in the spec, and never has been.

So, I'm not sure what vote you are proposing. It seems to me that the
vote would be around a _new_ idea, which (a) I'm reluctant to
encourage since it opens up a whole new debate, and (b) the idea
doesn't seem to have a champion, anyway!

I'll therefore leave it for now, that there is nothing to vote on, but
of course if someone discovers otherwise, we can look at this again.

Regards,

Mark

On 11/02/2008, Ivan Herman <ivan@w3.org> wrote:
> Mark,
>
> as I said before, I can live with that; this is what my submitted tests (and my implementation) does. I do not consider this issue as major. But you have to
> get an amen from the group: I was merely answering to Ben's mail!
>
> Ivan
>
> ------- Original message -------
> From: Mark Birbeck <mark.birbeck@x-port.net>
> Cc: ben@adida.net, public-rdf-in-xhtml-tf@w3.org
> Sent: 11.2.'08,  16:28
>
> > Hi Ivan,
> >
> > Just a quick comment on this...I do like your idea for generating
> > unique bnodes with the syntax "[_:]". However, as myself and Shane
> > have evolved CURIEs we've taken that approach that where there is a
> > precedent, we should follow it. Indeed, the whole point of CURIEs is
> > to make something that is already being done a little bit easier.
> >
> > And, since Turtle already uses ":" and "_:" in a specific way, we've
> > gone for having consistency with it.
> >
> > So I think we should stick with what we have, with the caveat that if
> > you think there is a possibility that Turtle might ever go in the
> > direction you are suggesting, then we could for now say that this
> > syntax is not allowed, leaving the way open for adding better support
> > in the future.
> >
> > My guess is that Turtle doesn't need it, though, since you can simply
> > use '[' and ']' to get the same effect.
> >
> > Anyway, I'll leave that to you semweb guys. ;)
> >
> > Regards,
> >
> > Mark
> >
> > On 07/02/2008, Ivan Herman <ivan@w3.org> wrote:
> > >
> > >
> > > Ben Adida wrote:
> > > >
> > > >
> > > > 88: I'm not sure about this one: I thought that about="[_:]" would
> > > > generate a *new* bnode since it's not named, thus yielding two bnodes
> > > > here, since it's used twice.
> > > >
> > > > <http://www.example.org/#somebody>
> > > >    foaf:knows [a foaf:Person], [foaf:name "Dan Brickely"] .
> > > >
> > > > If you used about="[_:a]" both times, meaning you actually gave it the
> > > > same bnode identifier, then yes that would mean one bnode in my mind.
> > > > Thoughts?
> > > >
> > >
> > > Well... that has been a discussion between Mark and me. Mark can make
> > > the argument better than I can... but I think the argument says that if
> > > you do not have a local name, then a CURIE means, essentially, the
> > > prefix part. That is why it is the same BNode.
> > >
> > > My argument was a bit different, namely that the CURIE spec is 'mapped'
> > > against the usage of BNode anyway, in the sense that we give an extra
> > > interpretation to what _:XXX means (ie, that it is a BNode), so we have
> > > the freedom to define what _: means.
> > >
> > > Either way: this must be decided. So it _is_ a good test:-) because it
> > > forces us to make a decision in one way or the other:-)
> > >
> > > Ivan
> > >
> > >
> > > > -Ben
> > > >
> > >
> > > --
> > >
> > > Ivan Herman, W3C Semantic Web Activity Lead
> > > Home: http://www.w3.org/People/Ivan/
> > > PGP Key: http://www.ivan-herman.net/pgpkey.html
> > > FOAF: http://www.ivan-herman.net/foaf.rdf
> > >
> > >
> >
> >
> > --
> >   Mark Birbeck
> >
> >   mark.birbeck@x-port.net | +44 (0) 20 7689 9232
> >   http://www.x-port.net | http://internet-apps.blogspot.com
> >
> >   x-port.net Ltd. is registered in England and Wales, number 03730711
> >   The registered office is at:
> >
> >     2nd Floor
> >     Titchfield House
> >     69-85 Tabernacle Street
> >     London
> >     EC2A 4RR
>
>


-- 
  Mark Birbeck

  mark.birbeck@x-port.net | +44 (0) 20 7689 9232
  http://www.x-port.net | http://internet-apps.blogspot.com

  x-port.net Ltd. is registered in England and Wales, number 03730711
  The registered office is at:

    2nd Floor
    Titchfield House
    69-85 Tabernacle Street
    London
    EC2A 4RR

Received on Monday, 11 February 2008 18:12:17 UTC