Re: [Idr] I-D Action: draft-ietf-idr-large-community-01.txt

[Nick Hilliard <[e-mail]>]

Jeffrey Haas wrote:
> I'd drop the "and BGP Extended Communities".

you're correct on this on the grounds stated, i.e. there is no
widespread use of extcomms for inter-as signaling.

I added that line in there, and have an item on my todo list to get a
couple of lines of text together in the next couple of days to explain
in the Introduction section why extcomms are unsuitable for solving the
problem at hand.

The plan behind this was that when this draft goes through iesg review,
the first question that's going to be asked is: "pls explain why
extcomms won't work for your use case?", so given that that is likely to
precipitate a couple of lines of explanation, the secdir review will
then ask, "why didn't you mention extcomms in the security section when
you mentioned rfc1997 both here and in the intro?".

That, at least, was the rationale.  The thing about this draft is that
the principal is so easy to understand that a lot of people in the
review process are likely to want to bikeshed it.

So there are two options open regarding this:

1. omit everything about extcomms and face a likely barrage of nits from
post-wglc review, and also don't explain to posterity why they were
considered a non-viable solution to the problem at hand, or

2. put in an honorary mention of extcomms and in no more than a tiny
number of lines, explain why they are unsuitable.

I'd appreciate your input on this though.  If you think it's unnecessary
or counterproductive, that might save me word-whacking time which could
be productively spent tilting at other windmills.

Nick