[core] Re: I-D Action: draft-ietf-core-uri-path-abbrev-04.txt

[Christian Amsüss <[e-mail]>]

Hello CoRE,

On Thu, Mar 19, 2026 at 03:49:20PM -0700, [e-mail] wrote:
> https://datatracker.ietf.org/doc/draft-ietf-core-uri-path-abbrev/

this very late update of -04 includes editorial changes that Carsten
recently sent (IMO the only possibly contentious item being that "this
document updates [...] to ~~alter~~++repair+ the behavior of"), but also
contains a change worth elaborating on:

I missed that as a consequence of -03, a further simplification around
the Proxy-Uri handling from "Interaction with other options" could be
made.

-03 already forbade Uri-Path and Uri-Path-Abbrev mixes resulting from
the use of Proxy-Uri (to be consistent with the easier and
well-established case of Proxy-Scheme processing). In terms of odd
combinations, this only leaves Proxy-Uri (without an actual path
segment) to be processed together with Uri-Path-Abbrev. But what I
should have simplified that already in -03 is this: The servers that
actually see this are exclusively some weird hybrid of proxy and server,
never plain servers. Thus, what used to be a SHOULD for every server
(and a MUST for some proxies) is now a MUST only for that small
proxy-and-server category. It was a SHOULD before because it was
unrealistic in the general case, but it is doable in that very special
case (if that case is even actually realized; I don't think it is).

I'll be presenting the updated -04 version as a delta from -02 in the
full meeting, hoping to meet the WG's approval for a WGLC on it.

BR
Christian

-- 
You don't become great by trying to be great. You become great by
wanting to do something, and then doing it so hard that you become great
in the process.
  -- Marie Curie (as quoted by Randall Munroe)