[tor-talk] Status/progress of TorRouter?
andrew at torproject.is
Thu Mar 29 14:42:14 UTC 2012
On Thu, 29 Mar 2012 15:29:50 +0200
<proper at secure-mail.biz> wrote:
> What is the status of TorRouter? Any progress on the project? I've
> been monitoring the active trac tickets and wiki sites. There are no
> changes since a long time. Is the progress behind closed doors?
> What is up with that project? Became it to big, unmaintainable,
> time-consuming? Or what are the issues, that no one is talking about
> it anymore?
Torouter is still a work in progress. There are really two projects
here, the one we want to build and ship in scale and the one the
community is building already. The goal of the torouter is to increase
the number of relays and bridges globally. The secondary goal is to
make it easy to use tor through a dedicated device. Both projects are
still so young I don't have a preference for one over the other. My
goal is to get more tor relays and bridges globally. Whichever project
manages to accomplish that first, wins. There are benefits and costs to
# Tor's ideas
The current state is we're working on plans to build our own hardware
to handle the torouter needs. The excito b3 and dreamplug hardware is
great, but costs too much and includes too many other features not
needed in a torouter. Features like audio/spdif, hdmi, 1080p HD video,
bluetooth, etc are not needed and just increase the price of the base
circuit board. We are experimenting with the Freescale iMX53 board 
to see if the current ARMv7 chips can handle being a tor bridge or
relay. It also works well as a ARM-arch build system for debs and rpms.
We're working with Bunnie of Bunnie Studios, and he thinks the
next-generation Freescale chipset and board are far better options,
because of the ARMv9 architecture and crypto offload build into the
We think this is roughly a $500k project to get the hardware designed,
manufactured, and to create a highly usable, point and click web
interface for the system based on Debian's armhf distribution. We want
to be able to offer hardware torouters, fully packaged and ready to go
for under $75, while still covering ongoing support, maintenance, and
We're currently doing a beta-test of a web interface to Tor with the
excito b3 hardware and around 20 people. I'll let Runa talk more about
how it's going and summarize the feedback from the testers.
There are lots of people excited about this project, but finding
funding to get started has been a challenge. The base hardware we
create can be re-purposed for other 'plug computing' projects beyond
the torouter. We're pondering kickstarter as well as a for-profit
subsidiary to get some funding to turn this into a reality.
And finally, torouter is a poor name. Onionbox is the leading candidate
for a public launch.
# Community ideas
The Access Labs team and others continue to work on their original
idea of a torouter based on openwrt with a luci/lua web interface for
tor. They've made great progress and have their system working on
current hardware on the market today. They are running into the
challenges of running tor in tight environments with limited cpu and
ram capacities. It's a fine challenge to solve, and they are making
progress. Having an actual embeddable tor, which is fully functional,
would be fantastic. It could open Tor up to many more devices.
It's probably at the point where they could start a larger beta test
to ship devices pre-configured with everything. The largest barrier to
adoption is integrating their tor into an existing device. The vast
majority of the world, including the technical community, wants
something that just works. Buying a Buffalo device, installing openwrt,
installing tor, configuring it all for transparent proxying, is more
work than many are willing to do. However, buying a device that is
pre-configured with all of this appeals to many more people.
The future is bright and waiting.
More information about the tor-talk