Re: [OPSAWG] draft-ietf-opsawg-tacacs-06 Contributions, Status and Plans

["Douglas Gash (dcmgash)" <[e-mail]>]

Just to clarify: because the last batch of comments were on v5, we will
actually take that as the datum for next version.

Although the current upload (v6) can be folded into this process, what we
can actually do now is is respond to all comments in context of v5 and,
with ensuing discussion of those points, can hopefully get us to a
agreeable version of the document for v7.

On 14/05/2017 15:06, "Alan DeKok" <[e-mail]> wrote:

>
>> On May 13, 2017, at 3:03 PM, Douglas Gash (dcmgash) <[e-mail]>
>>wrote:
>> 
>> 
>> So rather than directly updating the doc, we¹re looking for an
>>individual
>> response to each item. That is doable, we¹ll start putting that
>>together.
>
>  The point is to explain *why* the review was accepted or rejected.
>i.e. to have a discussion around the topic.

Agreed. That should come out of this process.

>
>  From your earlier comment:
>
>>> So our response to your reviews has been to incorporate, where
>>>feasible,
>>> and where we can apply then, to the doc.
>
>  Which items were incorporated?
>
>  Which items were *not* incorporated?  Why were they not incorporated?
>
>  There is no need to respond to each item individually.  Grouping things
>together is fine.
Sure, we can group them, we have them in the order you posted them mainly
to try to make sure we don’t miss any,


>
>  But when there are questions, they should be answered.  When comments
>are rejected, there should be an explanation.


So to recap, this process will take v5 and your comments to v5, then
hopefully we should have a transparent process getting to a v7 which we
can agree on. We can leverage v6 where it aligns with this process.

>
>  My larger issue with the review process so far is that the existing
>implementors haven't reviewed the document.  So we have no idea whether
>or not it describes the protocol they've implemented, or the choices
>they've made.

So let’s work through the document and identify what the core should be.
In the end we should have a protocol which clarifies the syntax of
interoperation whilst leaving the flexibility as needed for the varying
deployment implementations. Implementors have made very useful comments
and discussions already in the WG, and hopefully will contribute to the
discussions which will be raised by the response to your comments and the
ensuing move to closure.

Regards,

Doug.

>
>  Alan DeKok.
>