Re: CURIEs / QNames

>From a quick look, it appears that Curies are a superset of QNames, with a loosening of the restrictions on the "reference" or portion following the ':' so it does not have to be a valid element name. In a sense it seems like an update of QNames to reflect current practice.

Thus  the issues are mostly  the same from and XMLSec perspective and from our perspective can be treated similarly to QNames.

CURIEs also seems to introduce the notion of "a default prefix value" which can add some complexity.

regards, Frederick

Frederick Hirsch
Nokia



On Jun 15, 2010, at 1:55 PM, ext Scott Cantor wrote:

> Thomas, I'm assuming this is the spec you were referring to?
> 
> http://www.w3.org/TR/curie/
> 
> I reviewed it briefly and I guess my reaction is that it seems to be
> proposing something that's very much like QNames (in that it has the prefix
> problem), but is definitely formally distinct.
> 
> As a personal matter, my inclination is to say this is a pretty terrible
> idea, and seems to be perpetuating the problems that QNames created. But in
> WG terms, we don't have a BP saying "avoid QNames in content" at the moment,
> so there isn't any guidance on this right now.
> 
> I actually thought we did have a BP note on QNames, but I don't see one.
> Maybe we should consider adding one, but I guess the question you're raising
> is, should my proposal be expanded to address CURIEs or be limited to
> QNames?
> 
> I don't have a ready answer, since I haven't encountered them before.
> Offhand, I can't think of a major complication that would be introduced by
> including them in the same proposal, because the part that matters is just
> the prefix. If people think this has "legs", it's probably worth including.
> 
> -- Scott
> 
> 
>

Received on Tuesday, 29 June 2010 13:39:55 UTC