Re: [v6ops] Call For Adoption: draft-ietf-v6ops-ipv6-deployment

Hi Nick,
Thanks for sharing your opinion.
First of all, let me clarify the scope of this document since it seems you missed it. An RFC on the global deployment status of IPv6 is in line with the V6OPS charter and this draft aims to update old IETF RFCs on IPv6 deployment such as RFC6036, RFC7381, RFC6883. 
In this regard, the surveys from network operators and enterprises are reported in order to identify common open issues and challenges and therefore, make the community aware of them and stimulate more inputs. 
We know that the document can be improved but it can be considered a good starting point.
I would really appreciate more constructive comments.

Please find my specific answers inline as [GF].

Regards,

Giuseppe

-----Original Message-----
From: v6ops <v6ops-bounces@ietf.org> On Behalf Of Nick Hilliard
Sent: Tuesday, June 8, 2021 11:37 AM
To: Ron Bonica <rbonica=40juniper.net@dmarc.ietf.org>
Cc: v6ops@ietf.org
Subject: Re: [v6ops] Call For Adoption: draft-ietf-v6ops-ipv6-deployment

Ron Bonica wrote on 02/06/2021 19:11:
> This message initiates a call for adoption on 
> draft-ietf-v6ops-ipv6-deployment. Please post messages expressing your 
> position on adoption and rationale by 6/16/2021.

I'm struggling with this document. It looks like an aggregation of three separate things: a copy / rewording of chunks of ETSI White Paper 35, an operator survey with some results and the "Call for action" section, a set of suggestions on how to improve matters.

[GF]: As co-author of ETSI White Paper 35, let me disagree indeed there are a lot of differences. There is no survey in the ETSI WP 35. The ETSI WP 35 reports the statistics on the transition technologies based on public research. In this draft these statistics are not copied but referred. There is a big difference between "copy" and "reference". The polls among operators and enterprises and the relative findings are reported only in this draft. 

Regarding the survey, no methodology was provided and the sample size was too small to infer statistical significance.  Survey results are potentially interesting, but the results in this case are anecdotal.

[GF]: Regarding the methodology for the survey, it is just a questionnaire (see Appendix of the draft) and we can surely add more details on this. Regarding the sample size of the survey, note that we are considering 50 relevant operators. If we calculate the total number of customers served by these operators, you can immediately realize that the results are not anecdotal. We will add this information in the next version.

The Call for Action consists of a disparate list of technical and policy issues, ranging from "Government and Regulators" to "Oversized IPv6 packets".  There is no clear indication of how any particular topic made it into this list, or why other topics may have been excluded.

[GF]: We are thinking to change the title from "Call for action" to "Common IPv6 challenges". I agree that the list looks disparate, but our idea is to make this section more effective and highlight only the key message for each topic. Note that this list comes from the inputs from the IETF community. If you look at all the versions of this document, you can see how the Call for action section was extended over time due to the comments received.

Overall the document doesn't sit together happily; it draws conclusions in different areas based on insufficient data; many topics are presented without enough depth to provide sufficient justification for the claims that are made; there are a lot of unattributed statistics which could be misinterpreted as data.  I would be concerned that if it were published, many of the statistics could be requoted elsewhere using this as a primary document, and as it stands, the data quality isn't good enough to justify this.

[GF]: The data are collected from relevant references (e.g. ISOC, RIPE, APNIC,...). We tried to avoid data that cannot be verified. Could you please be more specific so we can improve?

The "Call for action" section has potential promise and would benefit from being moved to a separate document which discusses considerations for / impediments to IPv6 adoption.

[GF]: As said, this section will be reviewed to be more effective. Agree, it can be considered a starting point and inspire even more than one document on the different topics.

The operator poll needs examination in a separate document. There's insufficient detail at this point to work out whether there's enough material for an ID.

[GF]: An operator will not give you a lot of detail. There are some information that operators do not share. And again, note that the sample size is representative. We can make the calculation of the total customers served by the operators/enterprises which participated in the poll.

The ID would be better to settle on a cohesive and well-defined set of goals and then cover those comprehensively rather than providing a sparse overview of a wide selection of disparate topics.  There is material in there which is potentially interesting, but there is no clear path from where the document currently is to making a publishable rfc.  So at the moment there isn't a basis for the WG to adopt the ID.

[GF]: I agree that the document needs to be improved but I disagree that this is not in scope of V6OPS WG. As a matter of fact there was already a relevant consensus in the previous call for comments/adoption (please see the mail archive) and a large majority finds this work useful.

Nick

_______________________________________________
v6ops mailing list
v6ops@ietf.org
https://www.ietf.org/mailman/listinfo/v6ops