Re: [Anima] Adoption call for draft-liu-anima-grasp-api-06, ends Dec. 12, 2017

[Michael Richardson <[e-mail]>]

I have no objection to the API document being adopted.

I generally believe that the IETF should do more API documents rather than
less. I do find some of the ways that the IEEE tries to define
protocols via APIs (such as 802.15.4) unreasonably obtuse, and I don't
propose to go that route.  The way we do APIs is just fine for me.

Having said that, I would like the WG to either:
  1) progress this document very quickly with the intention of doing
     a -bis in two or three years.

  2) progress this document very slowly such that it won't be done
     until two or three years.

I am saying this because I believe that there are simply unknown unknowns
here, which we will only discover once people attempt attempt to build
non-trivial ASAs.   I point to PFKEY as an example of an API that was
standardized too soon, never revised, and turned out to be inadequate for
real use.  It was never revised because each vendor had already made their
own extensions and deployed them, and did not want to change, and the people
who did those extensions were not part of the IETF process, so there was no
energy.

--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-