Re: [bess] Joel Jaeggli's Discuss on draft-ietf-bess-mvpn-extranet-06: (with DISCUSS and COMMENT)

[Eric C Rosen <[e-mail]>]

On 3/10/2016 10:34 AM, joel jaeggli wrote:
> I think the question comes down to, is this document adequately
> prescriptive when it comes to implementation in a network.
As far as I can tell, no one has offered any reason to think that the 
document is not adequately prescriptive.

> You see to be asking us to write the words.

No, you are just being asked to state what you think is needed that 
isn't present.  This should be stated with enough precision to enable 
the authors to know what it would take to lift the DISCUSS.

The text you quote from Sue:

> I am looking for an operator-based “abstract” that focuses the reader on the key
> points.

does not help me understand what is claimed to be missing.  I do not 
know what an "operator-based abstract" would be, or where it is stated 
that such a thing is required.

> unique RD per VRF, yes it discusses this, then brings in the extranet RD
> leakage. calling this spongy is maybe an understatement.

This statement is a good example of the way in which the DISCUSS is 
unsatisfactory.  There is in fact no such technical concept as "RD 
leakage", and it is impossible to understand from the above text just 
what objection is being made.

Procedures for the use of the extranet RD are defined in such a way that 
existing procedures using the default RD will still work.

> This question
> might be easier to discuss cogently if in fact the document were easier
> to read. It is not, so you find your CO-ADs relying on the reviews of
> domain experts, and previous discussion on the list.

Lack of familiarity with the normative references seems to be a factor 
here.  However, I am not aware of any requirement to provide an 
"executive summary" for people who are not familiar with the normative 
references.

I would like to call your attention to 
http://www.ietf.org/iesg/statement/discuss-criteria.html.  In section 
3.2, "DISCUSS Non-Criteria", it says "None of the following are criteria 
for which the IESG should DISCUSS a document", and among "the following" 
it lists:

Unfiltered external party reviews. While an AD is welcome to consult 
with external parties, the AD is expected to evaluate, to understand and 
to concur with issues raised by external parties. Blindly 
cut-and-pasting an external party review into a DISCUSS is inappropriate 
if the AD is unable to defend or substantiate the issues raised in the 
review.

Also among the non-criteria is:

Stating "I think there's something wrong here, and I'll tell you what it 
is later" is not appropriate for a DISCUSS;

I think this would rule out a DISCUSS based on the feeling that "there 
might be some mistakes in the part of the document that I don't understand".

On that same page, it lists among the valid DISCUSS criteria:

It would present serious operational issues in widespread deployment, by 
for example neglecting network management or configuration entirely.

This seems to be the main criterion being used to support the DISCUSS, 
but given the number of provisioning rules and warnings contained in the 
document, and the extensive discussion of what has to be configured in 
the PEs, I don't see how one could claim that network management or 
configuration are neglected entirely.

Thus I think this DISCUSS violates the IETF process in the following ways:

- The DISCUSS is too vague to be actionable.

- It is not based on a valid DISCUSS criterion.

- It relies entirely on an "unfiltered external party review".