RE: [xri-editors] Section for XRI authority cref resolution

["Dave McAlpin" <dave.mcalpin@epokinc.com>]

Since relative cross-references caused so much internal discussion when we
talked about them on the list earlier this week, I'd vote that we ues an
absolute cross-reference in the example below. It might be less confusing to
save relative cross-references for the primer where we can explain their
intended use.

Dave

> -----Original Message-----
> From: Drummond Reed [mailto:drummond.reed@onename.com]
> Sent: Monday, November 03, 2003 10:56 PM
> To: XRI Editors (E-mail)
> Subject: [xri-editors] Section for XRI authority cref resolution
>
>
> Following is my proposed text for the brief discussion of resolving
> cross-references in XRI authorities. (As proof of why we need to add
> this, I was surprised how long it took me to think it through. It's
> really quite trivial, but the novelty of cross-references as an
> identifier construct is what could trip up implementers without simple
> guidance like this.)
>
> This would go in as section 3.2.6 right after the Examples section (sort
> of like a footnote to the Examples) and right before the User Relative
> XRIs section.
>
> Resolving Cross-References in XRI Authorities
> A sub-segment within an XRI authority segment may be a cross-reference
> (see sections 2.1.1.4 and 2.1.2). Resolving a cross-reference is
> identical to resolving any other sub-segment because the cross-reference
> is considered opaque. In other words, the value of the cross-reference
> (including the parentheses) is considered the literal value of the
> sub-segment for the purpose of constructing the Next Resolution URI as
> described in section 3.2.4.
> For example, when resolving "xri:@:a:b:(.c.d)/f", if the "b" sub-segment
> resolves to a Next Authority URI of "http://example.com/xri-authority/";,
> then the Next Resolution URI would be
> "http://example.com/xri-authority/:(.c.d)", because "(.c.d)" is the
> literal value of the entire cross-reference.
>
> Comments?
>
> =Drummond
>
>
>
>
>
> To unsubscribe from this mailing list (and be removed from the
> roster of the OASIS TC), go to
> http://www.oasis-open.org/apps/org/workgroup/xri-editors/members/l
eave_workgroup.php.