Re: [Idr] The IDR WG has placed draft-spaghetti-idr-deprecate-8-9-10 in state "Candidate for WG Adoption"

[Job Snijders <[e-mail]>]

Dear Nick, others,

On Thu, Jun 30, 2022 at 02:51:05PM +0100, Nick Hilliard wrote:
> IETF Secretariat wrote on 30/06/2022 14:42:
> > adoption scheduled tentatively for 7/25 to 8/12
> 
> is this wise?  It could be argued that this validates the practice of
> squatting on codepoints.

The reasoning I applied when submitting this internet-draft followed
my understanding of the group consensus which led to RFC 8093 [1]

I believe deprecation of "already squatted-on" codepoints has a few
benefits:

* Deprecation aids future protocol extension proposers. The effort that
  follows from being assigned a burned codepoint can be considerable.
  A new protocol extension is unlikely to serve as a proper lever to
  'undo the use of squatted values'; and might impede on the new
  protocol extension's deployability.

* Squatting does not lead to offical recognition or acknowledgement;
  this in turn encouraging stakeholders to follow the working group
  agreed upon assignment policies & procedures. I'd argue that
  deprecation invalidates the usefulness of squatting.

* Deprecation gives clear guidance to protocol implementers (for example
  tcpdump and wireshark) on what a particular protocol codepoint means
  (or doesn't mean).

Deprecation of codepoints is our mechanism to instruct IANA to not
assign a given value, since there isn't a "even though this value is in
the available pool, please don't hand it out for anything"...

Deprecations in context of IDR are considered temporary, see John's
message https://mailarchive.ietf.org/arch/msg/idr/EMZdzGX9CsMQwLuRaeaFM_hycD8/
A deprecation can be made undone following working group consensus.

If we run out of codepoints, the deprecated ones (after thorough study)
would be the first ones to evaluate for eligibility to return to the
unused pool.

Lastly, I put this internet-draft forward to provide the working group a
tangible proposal to discuss. I'm not attached to any specific outcome.

Kind regards,

Job

[1]: https://datatracker.ietf.org/doc/html/rfc8093