Date:        Thu, 2 Dec 2004 22:29:05 +0100
    From:        Ignatios Souvatzis <is@NetBSD.org>
    Message-ID:  <20041202212905.GE1005@beverly.kleinbus.org>

  | No... our UDP fragments at the link limit, for one-hop routes. our ICMP
  | fragments at 1280 bytes... instead of _limiting the output_ to 1280 bytes
  | which I would expect from reading RFC 2463.

Anything you've read that leads you to believe that ICMP echo/echo-reply
packets should ever be size limited (to < 64K anyway) should be reported
to the authors of the relevant RFC and the IETF working group that
produced it, as that was certainly never the intention - whatever text
suggests that (or even is capable of being read that way) should be cleaned
up in the next version of the doc.

ICMP error packets should generally be limited to 1280 bytes, to avoid
fragmentation - not ICMP informational packets, which can be as large
as they need to be - and most certainly, an ICMP echo reply must be
exactly as large as the ICMP echo packet that elicited it.

kre