On Thu, Jun 26, 2014 at 10:44 AM, David Hill dhill@mindcry.org wrote:
On Thu, Jun 26, 2014 at 12:35:00AM -0500, Scott Bennett wrote:
jason@icetor.is wrote:
http://www.coindesk.com/adopt-node-project-aims-bolster-bitcoin-network-secu...
Assuming that the relevant bitcoin programs could be taught to talkSOCKS, then it seems that tor hidden services would, in principle if not in performance, be an ideal solution. Running those bitcoin "full" nodes as hidden services might well make them less vulnerable to being shut
bitcoind works fine with tor and has some onion full nodes.
Performance of hidden services, however, are severely constrained by the hidden services protocol, which can slow connection times enough to make one consider USnail as a possible alternative, and the need for circuits of 2n-1 relays, which makes access even slower than normal tor circuits of n relays.
Performance of hidden services is actually rather good. ymmv.
I am using btcd, an alternative full-node implementation written in golang. Find it at https://github.com/conformal/btcd. It has built in proxy support. The wallet, btcwallet, is separate. It also has proxy support, so that you may connect to btcd over tor or as a tor hidden service. That can be found at https://github.com/conformal/btcwallet.
bitcoind nodes are a nice target to look for wallets. But with btcd, I run that at home while btcwallet runs on my encrypted laptop which connects to btcd over tor. There is no wallet on my btcd node machine.