Re: Why skolemization?

Steve Harris wrote:
> On 2011-03-27, at 21:02, Sandro Hawke wrote:
>> On Sun, 2011-03-27 at 18:15 +0100, Steve Harris wrote:
>>>> Well, yeah, I figured the system doing the generation could freely
>>> do
>>>> either:
>>>> or   
>>>> ... based on whether it wants to support deference or not.
>> I accidentally left out the other http example, to avoid a 303 when the
>> thing is not an information resource:
>> That makes it harder to give good error information, though.
>>> That would be fine, I thought you were advocating always using HTTP
>>> URIs.
>>> Magic URI substrings still don't quite sit well with me though.
>> I understand.  To me, they are a bit like 303s; not a solution I'm
>> particularly proud of proposing, but ... it seems like it will work, and
>> I can't think of anything better.
> Right, that's kind of where I am too.
> I guess the only questions in my mind are; do you need to identify skolem constant URIs, and is tag:{something} better than bnode:{something}.

say _: is a prefix in the syntax, map _: to _:base#

@base <> .

_:b12 = _: .

producing a new constant symbol for the existentially bound variable 
which is disjoint from the set of RDF URI References.


say _: is a prefix which always maps to <#_:>

_:b12 =

but then every existentially bound variable would also be an RDF URI 

(thinking out loud)

Received on Monday, 28 March 2011 00:43:58 UTC