Received-SPF: pass (sog-mx-4.v43.ch3.sourceforge.com: domain of gmail.com
	designates 209.85.218.53 as permitted sender)
	client-ip=209.85.218.53; envelope-from=mh.in.england@gmail.com;
	helo=mail-oi0-f53.google.com; 
MIME-Version: 1.0
Sender: mh.in.england@gmail.com
In-Reply-To: <CAACjpwKX9cwowiCruP9xw2UiqfsVXVC1TdKvA1HbQZ6UZ6qBsA@mail.gmail.com>
References: <c45a638f1e1640fe84bef01d12cda4c3@hotmail.com>
	<BLU402-EAS2546AD6C97DCED8FCE9C04CC6D20@phx.gbl>
	<CAACjpwKX9cwowiCruP9xw2UiqfsVXVC1TdKvA1HbQZ6UZ6qBsA@mail.gmail.com>
Date: Wed, 20 Aug 2014 16:37:24 +0200
Message-ID: <CANEZrP0WC2XL3Z0==BMjhWJuA8DgxBKUMKMdhh267JXduCZ0KQ@mail.gmail.com>
From: Mike Hearn <mike@plan99.net>
To: Cameron Garnham <da2ce7@gmail.com>
Content-Type: multipart/alternative; boundary=047d7bd6ac48e67bbe0501108a11
Cc: Bitcoin Dev <bitcoin-development@lists.sourceforge.net>
Subject: Re: [Bitcoin-development] Proposal: Encrypt bitcoin messages
Precedence: list

--047d7bd6ac48e67bbe0501108a11
Content-Type: text/plain; charset=UTF-8

I would be very happy if we upgraded the P2P protocol with MAC keys and a
simple home grown encryption layer, because:

   1. It's practically guaranteed that 5-eyes intelligence agencies are
   either systematically deanonymising Bitcoin users already (linking
   transactions to real world identities) or close to succeeding. Peter is
   correct. Given the way their infrastructure works, encrypting link level
   traffic would significantly raise the bar to such attacks. Quite possibly
   to the level where it's deemed unprofitable to continue.

   2. Tor is not a complete solution. The most interesting links to monitor
   are those from SPV clients connecting to Core nodes. Whilst Java SPV
   clients have the nice option of an easy bundled Tor client (er, once we fix
   the last bugs) clients that are not based on bitcoinj would have to use the
   full-blown Tor client, which is not only a PITA to bundle as Tor is not at
   all library-fied, but is a giant pile of C which is almost certainly
   exploitable. Even if it runs in a separate address space, for many
   platforms this is insufficient as a compromised Tor client could then go
   ahead and compromise your wallet app too.

Implementing a full Tor client is not a reasonable thing to ask of a wallet
developer, but doing HMAC checks and a simple ECDH exchange + AES would be
quite realistic.

--047d7bd6ac48e67bbe0501108a11
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div class=3D"gmail_extra">I would be very happy if we upg=
raded the P2P protocol with MAC keys and a simple home grown encryption lay=
er, because:</div><div class=3D"gmail_extra"><ol><li>It&#39;s practically g=
uaranteed that 5-eyes intelligence agencies are either systematically deano=
nymising Bitcoin users already (linking transactions to real world identiti=
es) or close to succeeding. Peter is correct. Given the way their infrastru=
cture works, encrypting link level traffic would significantly raise the ba=
r to such attacks. Quite possibly to the level where it&#39;s deemed unprof=
itable to continue.<br>
<br></li><li>Tor is not a complete solution. The most interesting links to =
monitor are those from SPV clients connecting to Core nodes. Whilst Java SP=
V clients have the nice option of an easy bundled Tor client (er, once we f=
ix the last bugs) clients that are not based on bitcoinj would have to use =
the full-blown Tor client, which is not only a PITA to bundle as Tor is not=
 at all library-fied, but is a giant pile of C which is almost certainly ex=
ploitable. Even if it runs in a separate address space, for many platforms =
this is insufficient as a compromised Tor client could then go ahead and co=
mpromise your wallet app too.<br>
</li></ol><div>Implementing a full Tor client is not a reasonable thing to =
ask of a wallet developer, but doing HMAC checks and a simple ECDH exchange=
 + AES would be quite realistic.</div></div></div>

--047d7bd6ac48e67bbe0501108a11--