Received-SPF: pass (sog-mx-3.v43.ch3.sourceforge.com: domain of gmail.com
	designates 209.85.216.46 as permitted sender)
	client-ip=209.85.216.46; envelope-from=etotheipi@gmail.com;
	helo=mail-qa0-f46.google.com; 
Message-ID: <4F7B62C6.6010007@gmail.com>
Date: Tue, 03 Apr 2012 16:51:18 -0400
From: Alan Reiner <etotheipi@gmail.com>
User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US;
	rv:1.9.2.28) Gecko/20120313 Thunderbird/3.1.20
MIME-Version: 1.0
To: Gavin Andresen <gavinandresen@gmail.com>
References: <4F7A1227.7070306@gmail.com>
	<CABsx9T3MQzJ5gN5xTZ9y5d-og11=mB86cM3ZP4S-fhjs1U+20g@mail.gmail.com>
In-Reply-To: <CABsx9T3MQzJ5gN5xTZ9y5d-og11=mB86cM3ZP4S-fhjs1U+20g@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Cc: Bitcoin Dev <bitcoin-development@lists.sourceforge.net>
Subject: Re: [Bitcoin-development] Signature Blocks and URI Sign Requests
Precedence: list

On 04/03/2012 02:46 PM, Gavin Andresen wrote:
> RE: signature blocks and BIP 10:
>
> We should avoid reinventing the wheel, if we can. I think we should
> extend existing standards whenever possible.
>
> So: could we encode signature blocks or BIP-10 transactions using
> S/MIME ?  Or is there a more appropriate "sign a message" standard we
> could/should use?
>
> You're glossing over little details like what character encoding is
> used for the message, but I'd rather leverage all the work already
> done by the IETF to nail down all those little details rather then
> re-discover them and come up with our own solutions.
>
I'm glossing over details because I actually have no experience dealing 
with character encodings,  or the implementation specifics of existing 
solutions (PGP or S/MIME).   That's why I'm emailing this list: I want 
to figure this stuff out, and at the same time try to converge on 
something that is efficient and can be interoperable between Armory and 
the Satoshi client (wallets, signature collection, sig blocks).

I don't go into these things solely to reinvent stuff.  My primary 
motivation for both ideas I have pitched so far (BIP 0010 and the sig 
blocks) is the versatility.  I like the encoding-independent, visual 
compactness of PGP ASCII-armored text blocks, but I don't like their 
opaqueness.  PGP vs my signature blocks basically look the same to a 
casual user, but even a moderately-knowledgeable user can appreciate the 
human-readable components of it.  You can visually identify if 
signatures are missing from sig-collection packet, or see that you 
signed with the wrong address without having to load an external program.

But that isn't a critical requirement, it's just my personal 
preference.  I'm fine with existing systems if it sidesteps a lot of 
problems and there's easy interface to it.    But, I don't want to have 
another BSDDB-wallet situation where we end up with 10x more capability 
than we need, and pay for it with 10x the complexity (at least in this 
case, using PGP is an existing crypto/security-sensitive technology).  I 
have made "simplicity" one of my goals in Armory.

Alternatively, we could change the discussion to a requirements 
discussion, to first figure out what we need in order to address 
multi-signature collection, etc.  Then evaluate competing ideas based on 
their qualities relative to the requirements.