possible revision of RDF serialization (a/k/a concretization) vs RDF abstract syntax

[TallTed]

rdf-concepts/spec/index.html

Lines 947 to 951 in 8b566d1

	<p>An IRI in the RDF <a>abstract syntax</a>
	MUST be <a data-cite="RFC3986#section-5">resolved</a> per [[RFC3986]] and
	MUST NOT be a <a data-cite="RFC3986#section-4.2">relative reference</a>.
	An IRI MAY contain a <a data-cite="RFC3986#section-3.5">fragment identifier</a>.
	An IRI SHOULD follow rules defined by the <a data-cite="RFC3986#section-3.1">IRI scheme</a>.

I'm confused by both the old and the new text here, which resulted in the above.

I don't think we regularly use "concrete" and I wonder about that linked definition (which is unfortunately not easily resolved from just this commit).

I would like the source text to say something close to —

    When converting an <a data-lt="concrete RDF syntax">RDF serialization</a>
    to the RDF <a>abstract syntax</a>
    (for instance, for loading into an RDF store), any IRI therein
    MUST be <a data-cite="RFC3986#section-5">resolved</a> per [[RFC3986]] and
    MUST NOT be a <a data-cite="RFC3986#section-4.2">relative reference</a>, but it
    MAY contain a <a data-cite="RFC3986#section-3.5">fragment identifier</a>.

[pchampin]

Lines 947 to 951 in 8b566d1

I'm confused by both the old and the new text here, which resulted in the above.

To be clear: the old text in this commit was proposed in PR #238, then reverted to the original text.
(see the corresponding diff in #238).

See my comment requesting the reversion.
See also my response on the commit. I believe that the original text: the goal was not to describe how to parse IRIs from concrete syntaxes to the abstract data model, but to characterize IRIs in the data model, regardless of how they were produce. So -1 on the proposal below.

I don't think we regularly use "concrete" and I wonder about that linked definition (which is unfortunately not easily resolved from just this commit).

I would like the source text to say something close to —

When converting an <a data-lt="concrete RDF syntax">RDF serialization</a>
to the RDF <a>abstract syntax</a>
(for instance, for loading into an RDF store), any IRI therein
MUST be <a data-cite="RFC3986#section-5">resolved</a> per [[RFC3986]] and
MUST NOT be a <a data-cite="RFC3986#section-4.2">relative reference</a>, but it
MAY contain a <a data-cite="RFC3986#section-3.5">fragment identifier</a>.

[afs]

the goal was not to describe how to parse IRIs from concrete syntaxes to the abstract data model, but to characterize IRIs in the data model

+1

[TallTed]

the goal was not to describe how to parse IRIs from concrete syntaxes to the abstract data model, but to characterize IRIs in the data model

OK.

[afs]

#238 (comment)

I think #242 was overtaken by events. I'll leave it open for now, and probably close it based on resolution of all of the above.

@TallTed - could you close this issue please?

[afs]

Thanks!