Re: [spring] We don't seem to be following our processes (Re: Network Programming - Penultimate Segment Popping)

[Fernando Gont <[e-mail]>]

On 6/12/19 16:42, [e-mail] wrote:
> 
>> While I may agree with you that is an attack on process here – and you may even find consensus on that statement – I am far from convinced you would find consensus on the question of which group is conducting the attack on process.
> 
> From https://mailarchive.ietf.org/arch/browse/ipv6/?qdr=m&so=frm
> 
> The last month's 10 top posters are:
> Fernando: 25
> Ole:      17 (including this message :-))
[....]

I don't think there is much room for interpretation here, but anyway I
should ask: are you suggesting that I have attacked or been attacking
the process? -- that's quite a statement from a wg chair to a wg
participant. And if not, I wonder what you are trying to illustrate with
the posted stats.


While one may or may not like loud folks, I'm not sure there's much
process specified in that respect. We do have, however, a process that
involves things like "doing a wg call for adoption in order for a wg to
work on a document", or even "make decisions on the mailing list".

What triggered the subject of this thread was, indeed, the fact that you
were implying that 6man had made a decision to work on EH insertion,
without a call for adoption of the relevant document, and without the
list given the chance to have an opinion.

Similarly, we have IETF consensus on what's in RFC8200. You may like it
or not. It may be good or not. But that's the consensus that we have.
But then you start with your rationale about limited domains (for which
there's no consensus) as an argument for folks being able to violate our
*That* is what I would deem as an attack of the process.

The above, that is, not having the wg make the decisions, and finding
ways to violate existing spec without following the due process, is what
would seem to me as an attack to the process.

As someone that is very involved in 6man, that tries to follow the
process as closely as possible, of course I've been loud on the topic.
The above has a lot to do with the fairness environment in which we are
operating, and remaining silent about it would be part of the problem.

You are a 6man wg chair. What I have been trying hard to is for the
decisions to be taken by the wg in a fair way (e.g. when it comes to a
decision to work on a document), and for the existing specs to be
respected and complied with unless there's formal consensus not to
change the specs.

Is being loud about that "attacking the process"? Or is that actually
that should simply happen without any participant asking, anyway?

Other than the theories of "proxy wars" do you've mentioned... have you
considered that there might be a conflict of interests going on, here?
e.g. between the working group consensus and a vendor's agenda?  As
chair, it is the wg that you serve.


We have been a long way when it comes to SRv6. It can be summarized to
include things like:

* When it comes to EH-insertion, rather than come up with a proposal to
say "hey, we want to change the standard, and this is the reason why",
there have been all sorts of manipulations of different kinds.

*  The first one being the claim that there was ambiguity in RFC2460
related to EH-insertion. Which, of course, everyone knew wasn't true.

* But, given the discussion, a number of us suggested that it should be
made even more crystal clear. Even after the insistence of many of us,
6man managed to ship rfc2460bis with such "ambiguity" in, simply because
a large number of folks (from the same vendor with a strong interest in
doing EH insertion), popped up on the list to move things forward "as
is". I did note this to Suresh at the time.... but the document shipped,
anyway.

* An heterogeneous group of folks (Jinmei, Enno, myself, Mark, and
others) weighed in during IETF LC. And then finally the corresponding
"clarification" (which I think was clear to everyone, anyway) was added
to what would eventually become RFC8200.

* During IESG review, it was somehow suggested by a folk (from the same
vendor as in other cases of the previous bullets) that before proceeding
with the publication of this document, we should essentially wait for a
proposal for EH-insertion to be worked out. Fortunately, RFC8200 was
approved and published, anyway.

* spring seems to have continued operating as if EH-insertion was
allowed. -- what looks to me a bit like "cook things up far from 6man",
which is of course unacceptable, since it is 6man that's in charge of
IPv6 standardization.

* When this was raised on 6man, and pointed out, then  we now heard
(after the said failure of "ambiguity in rfc2460" try), that it's okay
for them to violate RFC8200 and do EH-insertion, because they are
operating in a "limited domain" (when the IETF has consensus even about
what a "limited domain" is).

* Then you noted that 6man had decided to work on EHs at Singapore. And
when I polled you about that (how come?), you essentially asked me to go
look at the videos and minutes. But it turns out that there was no call
for adoption of any eh-insertion document (either in favour or against),
and obviously the list wasn't polled about it, either.


As noted, I have simply been trying hard for the decisions to be taken
by the wg (as opposed to "magical consensus"), and for the existing
specs to be respected and complied with unless there's formal consensus
not to do so.

Is being loud about that "attacking the process"? Or is that actually
that should simply happen without any wg participant asking, anyway?

Thanks,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492