Re: Running code (Was: I-D Action: draft-ietf-6man-ipv6only-flag-03.txt)

[Suresh Krishnan <[e-mail]>]

Hi Jordi,

> On Oct 23, 2018, at 2:33 PM, JORDI PALET MARTINEZ <[e-mail]> wrote:
> 
> Hi,
> 
> I don't think this can be enforced and even less for a specific ID or WG.

> 
> And this will be discriminatory also if applied to all the IETF work.
> 
> There are many authors that don't have time, resources, access to the source code from vendors, or even knowledge, among many many many other possible reasons, to implement an ID they submit.

I don’t think there is an expectation that the implementations have to necessarily come from the draft authors. e.g. look at the codestand work

https://codestand.ietf.org/codestand/

where people can request implementations. The underlying thesis is that there are people who have implementation skills who might be looking for guidance on what to implement.

> Also, there will be IDs which are not "implementable", etc., and clearly that doesn't disqualify that it may be a perfectly valid work.

Right. I am pretty sure this will be applicable only to Standards Track documents and the “not implementable” documents you mention will not be subject to this.

> 
> I guess this is a discussion that should be held at the general area list.

We have had this discussion multiple times over the past years.

RFC1264 (coincidently written by Bob) was the first document in this space requiring implementations before heading to Proposed Standard
RFC4794 obsoleted RFC1264 specifically pointing to the following text RFC2026  (which postdated RFC1264)

      Usually, neither implementation nor operational experience is
      required for the designation of a specification as a Proposed
      Standard.  However, such experience is highly desirable, and will
      usually represent a strong argument in favor of a Proposed
      Standard designation.

      The IESG may require implementation and/or operational experience
      prior to granting Proposed Standard status to a specification that
      materially affects the core Internet protocols or that specifies
      behavior that may have significant operational impact on the
      Internet.

and allows WGs to set their own rules. This is where things stand right now.

The current state of the art BCP is RFC7942 (which obsoleted RFC6982) that allows providing optional info about existing implementations and Section 4 of that document lists some of the benefits of doing so.

Thanks
Suresh