Re: [TLS] Consensus Call on MTI Algorithms

Nico Williams <> Thu, 02 April 2015 19:09 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id 3CC431A1B78 for <>; Thu, 2 Apr 2015 12:09:10 -0700 (PDT)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.666
X-Spam-Status: No, score=-1.666 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, IP_NOT_FRIENDLY=0.334, RCVD_IN_DNSWL_NONE=-0.0001] autolearn=no
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id gQ0xXCBHMM5y for <>; Thu, 2 Apr 2015 12:09:09 -0700 (PDT)
Received: from ( []) by (Postfix) with ESMTP id 06D901A1A70 for <>; Thu, 2 Apr 2015 12:09:09 -0700 (PDT)
Received: from (localhost []) by (Postfix) with ESMTP id C9930554060; Thu, 2 Apr 2015 12:09:08 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha1; c=relaxed;; h=date :from:to:cc:subject:message-id:references:mime-version :content-type:in-reply-to;; bh=6/rbw5v/0BBgMD mpfOS0CMNs0JM=; b=HAImksyn2dWtDBlw3eNiBzZ4sgV1wdFkG1hfNfspViR2NN 8VDlDCYq0a+GjT7iM4GIEIe7gMHYOC7htguV1RGW5G9MWRdxWzPxCDLqtXNNn7DU Q9l2NiMaY7me5DD1DH/vkocbukUu7sElR/mjc/2nHtzhTSZmm64C3SSlSv3uA=
Received: from localhost ( []) (Authenticated sender: by (Postfix) with ESMTPA id 56A8055405F; Thu, 2 Apr 2015 12:09:08 -0700 (PDT)
Date: Thu, 02 Apr 2015 14:09:07 -0500
From: Nico Williams <>
To: "Salz, Rich" <>
Message-ID: <20150402190905.GH10960@localhost>
References: <> <20150402184849.GF10960@localhost> <>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
In-Reply-To: <>
User-Agent: Mutt/1.5.21 (2010-09-15)
Archived-At: <>
Cc: "" <>
Subject: Re: [TLS] Consensus Call on MTI Algorithms
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: "This is the mailing list for the Transport Layer Security working group of the IETF." <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Thu, 02 Apr 2015 19:09:10 -0000

On Thu, Apr 02, 2015 at 06:59:11PM +0000, Salz, Rich wrote:
> It's reasonably easy to add ciphers.  Especially since most TLS 1.3
> implementations will be built on a 1.[012] codebase. 
> It is nearly impossible to remove ciphers.

If *all* clients and servers implement the same required N>1 algorithms,
then removing one is possible when you still have at least one of those
remaining.  Things get dicey when you're out of required algorithms and
you need to prefer less-universal recommended algorithms, or if not all
implementations adhere to the requirements.

Now, if requiring 3 algorithms turns out to be a joke because some
implementors will ignore one, and others another, then 3 would be a very
bad number indeed; 4 would be worse.  We do have to consider how likely
this is; 3 will be more costly than 2, no doubt, and that may be enough.

> Because of this, I am strongly in favor of JUST TWO.

I can get on board with two, but then I'd want AES-128-GCM and
ChaCha20-Poly1305.  I think having two very different algorithms is a
big plus.