RE: [xri] New RFC 2396bis draft

["Drummond Reed" <drummond.reed@onename.com>]

Since I've been studying 2396bis closely as we draft the XRI spec, a few
notes about key points I've noticed:
 
1) On URNs and persistence, read this text from 1.1.3 - especially the
second paragraph below.
 
"A URI can be further classified as a locator, a name, or both. The term
"Uniform Resource Locator" (URL) refers to the subset of URIs that, in
addition to identifying a resource, provide a means of locating the
resource by describing its primary access mechanism (e.g., its network
"location"). The term "Uniform Resource Name" (URN) refers to URIs under
the "urn" scheme [RFC2141]
<http://www.apache.org/~fielding/uri/rev-2002/#RFC2141> , which are
required to remain globally unique and persistent even when the resource
ceases to exist or becomes unavailable. 
An individual scheme does not need to be classified as being just one of
"name" or "locator". Instances of URIs from any given scheme may have
the characteristics of names or locators or both, often depending on the
persistence and care in the assignment of identifiers by the naming
authority, rather than any quality of the scheme."
 
This seems to acknowledge the reasons we ended out supporting both
persistent and reassignable segments anywhere within an XRI. They still
characterize persistence ("name") or reassignability ("locator") as
something that applies to a URI as a whole, rather than something that
applies to every segment of a URI and transitively ends out applying to
the URI as a whole (following the simple rule that only if all segments
are persistent is the entire URI persistent).
 
2) The BNF has some substantial changes. Most importantly, the old
"opaque" branch of "hier-part" in the original 2396 is gone. This was
the branch for non-DNS-or-IP based schemes like URNs. Now those schemes
belong in the new "rel-path" branch of "hier-part", which seems a little
odd because they are not at all relative, they just use different
authorities than DNS or IP. In any case, in XRI the abstract context
symbols will now fall in the "rel-path" branch.
 
3) Encapsulating an IPV6 address in square brackets in the "host" option
of the "authority" segment is now directly supported as specified in RFC
2732. This is excellent as we can now inherit this directly from 2396bis
BNF instead of having to write our own BNF to incorporate it from 2732.
It does set an obvious precedent for XRI parenthetical cross-reference
syntax too.
 
=Drummond 
 
 
 
-----Original Message-----
From: Wachob, Gabe [mailto:gwachob@visa.com]
Sent: Tuesday, June 03, 2003 12:37 PM
To: 'xri@lists.oasis-open.org'
Subject: [xri] New RFC 2396bis draft
 
Just a heads up - Roy Fielding has published a new draft of RFC 2396bis
(the update to the URI definition RFC). 
 
http://www.apache.org/~fielding/uri/rev-2002/issues.html
<http://www.apache.org/~fielding/uri/rev-2002/issues.html> 
http://www.apache.org/~fielding/uri/rev-2002/rfc2396bis.html
<http://www.apache.org/~fielding/uri/rev-2002/rfc2396bis.html> 
 
I haven't reviewed it in detail, but it does appear that the BNF has
been simplified a bit, and simplified in an interesting way. I'll let
you form your own opinion. This document is worth a read.
 
    -Gabe