Pascal Thubert (pthubert) <[email protected]> wrote:
    > What I have so far:

    > 1) minimal is a base that we expect will operate in many networks since
    > it appears to be needed to build the next stage where dedicated time
    > slots can be negotiated. Apparently this pleads against informational

    > 2) minimal is a recommendation for device builders, as opposed to
    > network admin. Apparently this pleads for std track rather than BCP

Not to disagree, but to clarify wording:
    Many applicability statements in for-instance email, or routing or
    enterprise-networking (INT) explain how to deploy/configure a (set of) 
protocol
    for use in a particular setting. I.e. what parameters to tweak, what
    options to turn on/off.  These are advice operational to operators, and
    do not cause the operator to change their RFP.

    Minimal is something that would appear in an RFP, and is a set of
    requirements to implementors as to which options to build in.  Given the
    constrained nature of devices, those decisions can not be left up to 
operators.

    > 3) minimal defines a way to compute the Rank that cannot be obtained
    > with a simple parameter in an existing implementation. The operation
    > SHOULD be programmed in the device for interoperation and that
    > operation is not specified in a preexisting RFC. This pleads for std
    > track

Well, this might point towards there needing to be a new ROLL document that
should be on standards track that minimal should reference.

This might still be a good idea.


--
Michael Richardson <[email protected]>, Sandelman Software Works
 -= IPv6 IoT consulting =-



Attachment: signature.asc
Description: PGP signature

_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch

Reply via email to