draft needed - fe80::/10 vs fe80::/64 (Re: Objection to draft-ietf-6man-rfc4291bis-07.txt)

神明達哉 <jinmei@wide.ad.jp> Thu, 30 March 2017 02:29 UTC

MIME-Version: 1.0
Sender: jinmei.tatuya@gmail.com
From: 神明達哉 <jinmei@wide.ad.jp>
Date: Wed, 29 Mar 2017 19:29:54 -0700
Message-ID: <CAJE_bqcc=xHkcScxuQEszhJjKFqV3bPd=eswOwUf_t9K6bi64g@mail.gmail.com>
Subject: draft needed - fe80::/10 vs fe80::/64 (Re: Objection to draft-ietf-6man-rfc4291bis-07.txt)
To: Alexandre Petrescu <alexandre.petrescu@gmail.com>
Cc: IPv6 IPv6 List <ipv6@ietf.org>
Content-Type: text/plain; charset="UTF-8"
Archived-At: <https://mailarchive.ietf.org/arch/msg/ipv6/FdX9BNbJmf8B5sVwKA-c3ei3KgY>
Precedence: list

At Wed, 29 Mar 2017 20:39:27 -0500,
Alexandre Petrescu <alexandre.petrescu@gmail.com> wrote:

> The BSD code that was shown earlier checks the src of a received ND
> packet against fe80::/10 (not against fe80::/64).
>
> And other code in same OS BSD adds fe80::/64 (not fe80::/10) in the rt
> table.

I see there's some loose area between fe80::/10 and fe80::/64, but
that does not immediately mean a "problem".  If you want to raise a
concern to be shared with the wg, we'd need to see a specific
operational scenario such as one where one implementation cannot
communicate with others using a link-local address (whether the
definition is fe80::/10 or fe80::/64).  So far no one perhaps except
yourself sees such a scenario.  I'd suggest you summarize that
scenario in a draft and submit it, and then discuss it here, rather
than make this thread any longer.  I believe a draft is better than
more messages in this thread - email exchanges tend to lack critical
details, and at this point its overhead outweighs the vibrancy of
real-time(-ish) discussions.  I also guess that was a suggestion by
Ole as a wg chair.

--
JINMEI, Tatuya

draft needed - fe80::/10 vs fe80::/64 (Re: Object… 神明達哉