Re: [tor-dev] Statistics on fraction of connections used uni-/bidirectionally