use of "relative URI"

[pchampin]

While reviewing #43 and #45 I noticed several uses of the term "absolute URI", which I believe is not always aligned with the intention of the author.

"Absolute URI" is defined by https://www.rfc-editor.org/rfc/rfc3986#section-4.3 as

the absolute form of a URI without a fragment identifier

(emphasis is mine -- yes, this is odd, this is mostly historical)

I don't think this is appropriate in some of the places where the term is used in #43 or #45 (and possibly other places, that's why I open an issue about it). E.g., end users are very often identified by a URI with a fragment identifiers in Solid.

The correct term would be "URI" because URIs are always "absolute". What are sometimes called "relative URIs" must strictly be called "relative URI references" -- they are not URIs. See the note in https://www.rfc-editor.org/rfc/rfc3986#page-11 .

@acoburn

[acoburn]

@pchampin this is a good clarification. I was using "absolute" to be clear that this was not a relative URI, but I did not mean to exclude the possibility of fragment identifiers. We should be able to simplify instances of "absolute URI" to just "URI" as you suggest.

[pchampin]

The text in RDF-Concepts about IRIs could be useful (replacing IRI with URI and RFC3987 with RFC3986).
https://www.w3.org/TR/rdf12-concepts/#section-IRIs