Re: Issue with uniqueness of identification

Jos, your suggestion works only if the location of the important
document is fixed, known in advance, and is a URL. But this location might be
something like

loc(from -> "debruijn@inf.unibz.it"^^xs:anyURI
    to   -> "public-rif-wg@w3.org"^^xs:anyURI
    date -> "2009-03-26T17:59:22+01"^^xs:dateTime)

Moreover, the author might not know this in advance, so the location of the
document could be asserted via some kind of an import assertion at run time.

I realize that this is well beyond the capabilities of BLD, but I look
into the future, and I am thinking about FLD.

michael


On Thu, 26 Mar 2009 17:59:22 +0100
Jos de Bruijn <debruijn@inf.unibz.it> wrote:

> according to the definition of import directive [1], the IRI is the
> location (not the identifier) of the document, so there should not be a
> problem.
> 
> [1] http://www.w3.org/2005/rules/wiki/BLD#Formulas
> 
> Michael Kifer wrote:
> > One more issue with Ids in the annotations.
> > 
> > We never said that the Ids used in the annotations are unique within the
> > document or globally. This is especially problematic with document Ids.
> > If they are not unique globally, then the definition of import of documents is
> > not well-founded.
> > 
> > michael