[rpd] A question for the PDWG

[Andrew Alston]

Just as a note on RFC7020 -

I looked at this document and noted that it is information and replaced a BCP - which was interesting in and of itself (that normally doesn't happen), but yes - it is purely an informational document.

The question of if a new BCP around the functioning of RIR's is needed - is an interesting question, and if there is sufficient interest in doing an updated 7020 either as a BCP or as an informational document - I'd be happy to work on something with those interested.  It has also been raised with me about writing a corporate governance for RIR's document under the auspices of the IETF, either as an IETF document or an ISE document.  On the former, I think there is merit in additional work here - on the latter - I am far from convinced that writing RIR governance documents through the IETF - either via informational track, bcp track or ISE track would be a good idea - in fact I can see distinct problems with it.

If there is interest in doing an updated 7020 though - I'd happily contribute - to the point where I'd consider AD sponsoring such if the need arose and the contents of the document were sane (provided I wasn't an author on it as that would create a distinct conflict that I could not be party to)

The question on this though is - what are the areas of 7020 that would need to be fixed/changed - and do we have sufficient people who are wiling to take the pen to do this work.  The other question of course would be if we wish to take this to a BCP or another informational document.  Curious to hear thoughts.

Thanks

Andrew 


-----Original Message-----
From: David Conrad <drc at virtualized.org> 
Sent: Wednesday, May 25, 2022 7:45 PM
To: Fernando Frediani <fhfrediani at gmail.com>
Cc: rpd at afrinic.net
Subject: Re: [rpd] A question for the PDWG

Fernando,

On May 25, 2022, at 8:44 AM, Fernando Frediani <fhfrediani at gmail.com> wrote:
> I think you missed my main point about this topic.

I don’t believe so. My understanding (please correct me if I am mistaken) is that you would like there to be (or rather, that you believe it illogical for there not to exist) a governance hierarchy that sits atop the RIRs such that misbehavior by an RIR can be directly addressed. I am merely pointing out that while this may (or may not: what happens when the parent is captured by “the bad guys"?) be a good idea, it does not reflect current reality as defined by existing arrangements between network operators, the RIRs, and ICANN. Despite your view about the logic, ICANN has no mechanism to “de-recognize” an RIR and even if it did, it’s wildly unlikely the RIRs or network operators would care.

> That must be a counter balance measures for the rest o Internet community to stop bad actors, including entire RIRs to do things that may affect the stability of the Internet.

There is: that power is vested in the respective RIR communities.

> Regarding RFC 7020 I personally hope it gets fixed at some point. I don't think that is all bad, but it lacks a fundamental point to any system like this: double degree of jurisdiction which must exist in any administrative and legal system. It simply makes any RIR administrative final.

Again, RFC 7020 merely documented the system existing at the time of publication (it might have evolved since then). It is descriptive, not proscriptive. The Internet, including the administration of numbering resources, is _decentralized_. This has both positive and negative implications. For example, you can’t "appeal to authority" since there isn’t one.

Regards,
-drc