Re: [Netconf] a couple zerotouch-21 issues

Hi Kent,

> On May 19, 2018, at 3:31 AM, Kent Watsen <kwatsen@juniper.net> wrote:
> 
> 
> Regarding what to do with the "zero touch device data model" in Section 8:
> 
> a) Everyone agrees on not delaying this draft any longer waiting for the keystore drafts to complete.
> 
> b) There is a difference in opinion as to if we should move Section 8 to the Appendix, or delete it altogether.  We don't have consensus.
> 
> c) There is new information since the above comments were made.  The new information is that the WG will likely adopt the crypto-types and trust-anchors drafts and, thus the dependencies change.  The type "cms" becomes "ct:cms", and the path "/ks:keystore/pinned-certificates/name" becomes " /ta:trust-anchors/pinned-certificates/name".  This changes the equation in that the probability is higher now that this non-normative module will remain valid longer.  Do we still think that it is "high risk" and, even if it is, is it really a problem?  Have we ever published an example module that imported modules that were works-in-progress before?  [I can't think of one, but I don't follow every draft either]
> 
> d) On the flip side, why is having a non-normative example better than no example at all, however evolved or not evolved the current example is?   There is a tradeoff between having something (which is nice) and the risk that it will become invalid (which would be a nuisance).  How nice is nice and how big of a nuisance would it be?  [it seems like nice > nuisance]
> 
> Thoughts?

My personal preference is that an example, in a appendix, which is non-normative, is better than no example.

mj as contributor

> 
> Kent // contributor
> 
> 
> _______________________________________________
> Netconf mailing list
> Netconf@ietf.org
> https://www.ietf.org/mailman/listinfo/netconf