Just to say this at Last Call time, so it's not an IESG Evaluation late surprise ...
On Sep 29, 2017 08:06, "Denis Ovsienko" <denis@xxxxxxxxxxxxx> wrote:
---- On Thu, 28 Sep 2017 14:26:22 +0100 The IESG wrote ----
>Hello all.
>The IESG has received a request from the Sunsetting IPv4 WG (sunset4) to
>consider the following document: - 'IETF: End Work on IPv4'
> <draft-ietf-sunset4-ipv6-ietf-01.txt> as Proposed Standard
In addition to what has been said already about the present revision of the document, let me make a suggestion for its next revision if the working group decides to work on it.
What such a document could propose is that all new I-Ds have to include an "IPv4 Considerations" section, which either states that the document has absolutely nothing to do with IPv4 (and in that case the section is removed before the publication) or actually makes relevant considerations along a set of guidelines (like RFC 5226 explains how to make IANA considerations). This will help both authors and readers spend sufficient time thinking about how the document will work in IPv4-only, IPv6-only or double-stack environments, so that they can identify more real issues before the publication or the implementation.
To keep those guidelines right-sized, the document could make IPv4 considerations for recent 10-15 protocols/documents to show how it works and what problems it would help to avoid.
If we did this, I wonder if the apparently-BCP-to-be might establish a default of IPv6-only, so that drafts that aren't IPv6-only are the only ones affected. In a perfect world, we would see more specifications that aren't explicitly IPv4-only or explicitly dual-stack, so only the drafts that are going down the exception path would be affected.
We should also remember that the community has been more resistant to adding mandatory sections to all RFCs in recent years, so, there's that.
Thanks,
Spencer, speaking as an individual who's watching this discussion pretty closely ...
--
Denis Ovsienko