Re: [DNSOP] what i said at the mic (re: dnssec-key-timing)

Jelte Jansen <> Tue, 17 November 2009 20:44 UTC

Return-Path: <>
Received: from localhost (localhost []) by (Postfix) with ESMTP id 8CD563A6989 for <>; Tue, 17 Nov 2009 12:44:42 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -2.288
X-Spam-Status: No, score=-2.288 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HOST_MISMATCH_NET=0.311]
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id H36mAWgUI7Su for <>; Tue, 17 Nov 2009 12:44:41 -0800 (PST)
Received: from ( [IPv6:2001:4f8:3:bb::5]) by (Postfix) with ESMTP id 851B53A67A1 for <>; Tue, 17 Nov 2009 12:44:41 -0800 (PST)
Received: from [] ( []) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by (Postfix) with ESMTP id 33C9EE6026; Tue, 17 Nov 2009 20:44:38 +0000 (UTC) (envelope-from
Message-ID: <>
Date: Tue, 17 Nov 2009 21:42:57 +0100
From: Jelte Jansen <>
User-Agent: Thunderbird (X11/20090817)
MIME-Version: 1.0
To: Johan Ihren <>
References: <> <>
In-Reply-To: <>
X-Enigmail-Version: 0.95.7
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 7bit
Subject: Re: [DNSOP] what i said at the mic (re: dnssec-key-timing)
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: IETF DNSOP WG mailing list <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Tue, 17 Nov 2009 20:44:42 -0000

Hash: SHA1

Johan Ihren wrote:
> I don't remember exactly what I said, but what I meant was not that
> "double signature" is sufficient to accomplish an algorithm rollover,
> only that it was needed as part of one. So I think you and I agree hee.

I think so too, mostly :)

I would dub what I proposed for algorithm rollovers 'reverse double signature'

>> (btw i agree with olaf that some form of collaboration between these
>> documents
>> might be nice)
> Of course there will be, although I'm still skeptical about circular
> dependencies.

Of course, though i rather see a circular dependency than an inter-rfc
inconsistency. I was thinking more of collaborative documents :)

> This falls down into the question about whether to cover all the
> rollover "methods" or make a recommendation and only cover that
> alternative. My take away from the WG was that the key timing doc should
> not make recommendations, but describe all the alternatives.

Well, since I am the proud owner of the humble opinion that algorithm rollover
requires a substantially different method, I would personally like to see that
method covered as well.

>> I think a reference to 4641bis and a scheme to match the text there
>> would be nice
> My primary issue with that is that as soon as we refer to 4641bis there
> is a circular dependency and then the two documents must be published
> together. My secondary issue is that I maintain my position that the key
> timing document is "theory" while 4641bis is "practice". Therefore, as
> the key timing document covers a more narrow topic in greater depth,
> there is just no way to avoid references from 4641bis to us.

Ah, now there may lie the (a?) problem. I was thinking of 4641bis as the
defining rfc (ie. what to do), and key timing as a mathematical description of
(some of) the methods mentioned in the former (ie. how to do it).

Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla -