On Tue, Nov 19, 2013 at 01:55:15PM +0100, Mike Gogulski wrote:
> Reject transactions from Tor nodes and they'll be posted on USENET, the
> Bittorrent DHTs, I2P, Freenet, blog comment spam, DNS TXT entries,
> Tahoe-LAFS shared directories, pastebins, piratepads, whatever. In any
> case, it doesn't matter where the tx is generated, as long as there's at
> least one node willing to ignore the "reject from Tor" rule and relay to
> the rest of the network. You can't stop the signal...
>
> It's very likely that this is NOT a false flag. People have been coding
> toy assassination markets for a decade or more, it's just that only
> recently has the digital bearer currency (Bitcoin) component slotted
> into place to complete the puzzle.
>
> Heck, there's someone on this very list (or at least directly connected
> to several of us) who related this anecdote early this year or late last
> year:
>
> "So, I was at a conference a few months ago and a guy sitting next to me
> was writing code for an assassination market. Right there, before my eyes."

There's also alp's work on implementing external state contracts:

https://bitcointalk.org/index.php?topic=260898.0

It's based on preliminary work done by Mike Hearn, with subsequent
improvements by myself. The basic idea is to rely on a third-party
oracle to attest to the validity of statements; the oracle releases
specific private keys depending on the outcome. I even came up with a
way to make the transactions completely private and unknown to the
oracle in the default case: essentially the transactions can be done in
such a way that with the consent of both parties the oracle's keys are
never used and never touch the blockchain. Only if both parties choose
not to co-operate do the oracle's keys (or keys derived from them) need
to be revealed.

A final touch is to make it possible for the oracle to profit from their
predictions: you can do this by selling the revealing of the specific
keys as a service, and to prevent reuse, use a shared secret securely
sent to the buyer encrypted to their pubkey. The buy knows the secret,
and thus knows the pubkey the oracle claims to reveal is in fact real,
but the buyer can't prove that to anyone else because the contents of
the message are malleable by both buyer and oracle.


One of the reasons why this assassination market may perhaps be a false
flag is because it's nowhere near as secure as it could be. A weak
reason, but worth thinking about.

Also, I'll make the public statement that I may create a fake anonymous
assassination market in the future for the express purpose of ripping
off any individual who wishes to use political violence to advance their
goals.

--
'peter'[:-1]@petertodd.org