From rusty at rustcorp.com.au  Mon Mar  7 23:38:42 2016
From: rusty at rustcorp.com.au (Rusty Russell)
Date: Tue, 08 Mar 2016 10:08:42 +1030
Subject: [Lightning-dev] We don't need R-Value,
	how OP_CODESEPARATOR saves the day
In-Reply-To: <CA+1nnrkXG35rMNcxOdNgQvFYj-qBHdL2hFdmiYmyKri4Yb86aA@mail.gmail.com>
References: <CA+1nnrkXG35rMNcxOdNgQvFYj-qBHdL2hFdmiYmyKri4Yb86aA@mail.gmail.com>
Message-ID: <87a8m9oo3h.fsf@rustcorp.com.au>

Nicolas Dorier <nicolas.dorier at gmail.com> writes:
> I'm amazed as I found a useful case for OP_CODESEPARATOR.

Me too!  This is an awesome hack.

There is an important caveats: You must make sure you never ever create
a commit tx which is identical a previous one.

There's also a requirement that you store all previous signatures,
avoided by both the original draft scheme (using BIP32 and handing over
old private keys) and the deployable lightning chain-of-hashes.

Since these scripts will hardly ever hit the chain, I think this a
premature optimization.  But maybe I'm still a bit stunned by this use
of OP_CODESEPARATOR :)

Great work!
Rusty.