[Freedombox-discuss] FreedomBox and Bitcoin (and the petition)

[Jonathan Wilkes]

----- Original Message -----
> From: Ted Smith <tedks at riseup.net>
> To: freedombox-discuss <freedombox-discuss at lists.alioth.debian.org>
> Cc: 
> Sent: Tuesday, November 13, 2012 8:51 AM
> Subject: Re: [Freedombox-discuss] FreedomBox and Bitcoin (and the petition)
> 

[...]

> For reference, it's really, really, really hard to take web browsing and
> make it anonymous. Forcing every connection from browsers to go through
> Tor won't help you anywhere near as much as it does for Bitcoin. Bitcoin
> has already done 99% of the hard work building a protocol that is
> totally (as far as either of us can tell) anonymous if it's _just_
> tunneled through Tor without any other filtering.

When I referred to "these" attacks going away, I was talking about implementations
of the Kaminsky attack, or efforts to link an IP address with a transaction even
without connecting to all nodes (say, making inferences from the data available at
blockchain.info).  Bitcoin behind Tor still isn't anonymous in many obvious
ways.  I refuse to respond to a phrase as fatuous as "totally anonymous", but if the
system really protected its users' identity[1] in a practical way then the "coin taint"
discussions and (unfortunately) implementations by some services would not be
possible.  Bitcoin is not Chaumian cash-- which _would_ give the user the kind of
anonymity to protect against a self-destructive coin-tainting scheme since it's
untraceable.  Whatever definition of "totally anonymous" you're using, it is false.

That is just one of the consequences of Bitcoin not being anonymous, and there are
many other problems with transactions linking to user identity.  You should have a look
at http://bitcoin.org/ if you don't understand the technology and its (non-)features, and
ask questions on a Bitcoin-specific list or forum for more information.

-Jonathan

[1] The Bitcoin-behind-Tor user is trusting that no one will care about the inferences that
can be gleaned from a database that holds every single transaction ever made.  Even in
the best of circumstances where this user doesn't leak _any_ identifying information out of
band, the "coin taint" problem shows that they cannot be immune to such an attack.
Whether we like it or not, "coin that well-known Bitcoin service foo claimed was stolen"
is mathematically provable, uniquely-identifying information that _will_ currently break that
Bitcoin's fungibility.  There have been ways proposed to address this, but it's sheer confusion
to claim "total anonymity" for Bitcoin-behind-Tor in its current state.