Re: [dispatch] RFC 3896 and 3987 vs WHATWG URL Living Standard

[John C Klensin <[e-mail]>]

Larry, 

Without trying to answer your question, two small cautions.
While WHATWG has chosen, probably for good reason, to
concentrate on URLs, as you know 3986 is about URIs more
generally.  It includes some sweeping requirements that are
arguably unreasonable or over-constraining for non-URL URIs
generally and URNs in particular.  One of the reasons the URNbis
WG got RFCs 8241 and 8254 out and then essentially gave up was
connected to those constraints and I gather that there has been
some development and deployment outside IETF that made progress
by ignoring restrictive interpretations of 3986.  If it is worth
the sort of work you suggest to recognize changes consistent
with WHATWG work, is it not equally worth considering updates
that would create better alignment with URN practice and demands
on URN extensions?

Not directly related, but my understanding is that work
elsewhere has moved considerably beyond RFC 3987 (fwiw, a
Proposed Standard while 3986 is an Internet Standard).  It seems
to me likely that we would want to treat 3986 and 3987 quite
differently, maybe even retiring the latter in favor of a
pointer to work done elsewhere.

I don't know how to peek into a can of worms without opening it.

    john


--On Wednesday, June 9, 2021 17:39 -0700 Larry Masinter
<[e-mail]> wrote:

> The question is: is there enough interest to have a BOF at
> IETF 111 to discuss
> a plan for moving forward? I don't imagine starting out with a
> presentation of
> a solution. 
> 
>   I'd probably want to focus on "minimum viable product":
> What things could we do first with minimum effort that would
> improve the current state?  A replacement for RFCs 3986 and
> 3987 wouldn't be my first step. Maybe a short standards-track
> "UPDATES" that points to/contains deltas or a new grammar that
> is more consistent with (tested) implementations.
> And a pointer to WHATWG-URL, whether normative or not I'm not
> sure.
> 
> That's a little more than Mark's "Section 1.1.3" because it
> makes normative changes but (apparently) minimal ones.
> 
> I was encouraged to see WHATWG paying attention to feedback
> from curl and node.js as implementations of URLs, not just
> browsers.