[The main issue, in its discussion, and rightly so, is the "futile"
uses of DISCUSS - my favorite example being 2929bis « Domain Name
System (DNS) IANA Considerations », blocked for many months by
iana.org vs. ietf.org. But my message is about the "examples" RFC such
as 2606, 3330, 3849 or 4735.]
On Tue, Jun 17, 2008 at 09:42:18AM +1200,
Brian E Carpenter <brian(_dot_)e(_dot_)carpenter(_at_)gmail(_dot_)com> wrote
a message of 62 lines which said:
In my opinion that is a source of technical confusion and possibly
of unwanted traffic.
So I think there is a strong argument that RFC 2606 values SHOULD be
used whenever reasonably possible.
However, there are many cases where it is not reasonably
possible. Steve Bellovin gave good examples for RFC 2606. RFC 3330 has
similar problems. Only a /24 is reserved which means that, for BGP
tutorials, you need to announce /25s and /26s, with a long prefix in
common, which is pedagogically bad. RFC 3849 has a similar problem,
without even the excuse of address scarcity.
Many people who use IP addresses in documentations do not use RFC 3330
because it is too much to ask the readers to see at a glance the
difference between 192.0.2.1 and 192.0.2.129. They are too similar.
So, I agree with you, a RFC 2119 "SHOULD" is OK, and therefore should
not block the RFC, just trigger a discuss (lowercase, a real discuss,
not a blocking vote).
IETF mailing list