Re: WG ACTION: 2 weeks to discuss [LL70] DNAv4 normative or informative?

    Date:        Wed, 5 May 2004 01:26:14 +0200
    From:        Erik Guttman <[email protected]>
    Message-ID:  <a05200f07bcbdce62e305@[80.139.178.51]>

[Stuart]

  | To implement this "SHOULD NOT" requirement, an implementation has to also
  | implement [DNAv4] to determine what's a "valid routable address".
  | 
  | Hence, normative reference (or the "SHOULD NOT" prohibition could be
  | removed or refined to not depend on DNAv4).

I agree with that, it is clearly a normative reference as the text
currently stands.

  | [Erik]
  | 
  | Hmmm.  Do you realize you want to raise the bar for the publishing
  | of this document beyond what the IESG has called for.

Yes, we could cheat, but then if the DNA doc never appears, the LL spec
will be useless.

The best way to fix this is to rewrite the reference so that it is
no longer normative -- the easy way to do that would be to simply
copy the text from the DNA draft that describes what is a valid routable
address, and include it in this doc (and then add an informative
reference to note the source of the text).

Do that we get exactly the meaning the doc currently has, no indefinite
delay waiting for another WG's doc to appear, and no cheating of the
process (with everyone just looking the other way) either.

[Aside: even if the IESG allow this through, because of negligence or
collusion, the RFC editor may spot it, and hold up publication anyway,
so it needs to be fixed, one way or another].

kre