[liberationtech] WSJ Op-Ed: No Quick Fixes for Internet Freedom

[Roger Dingledine]

On Fri, Nov 19, 2010 at 03:44:03PM -0800, Mehdi Yahyanejad wrote:
> Most of the public justification for the Internet Freedom funding has
>been the specific cases of Iran and China (read congressional hearings
>on these issues). It makes sense to ask about the effectiveness of
>the funding in Iran and China. GIFC tools have been pretty successful
>in Iran so far.
[snip]
> Having a broader strategy for "Internet freedom" and listing priorities
>is needed but those other priorities are not as well-defined as the
>technical problems. A good thing about technical problems is that with a
>good mix of money and engineers, you can generate some results.

One of the topics that I think is often overlooked in these discussions
is the question of sustainability.

I had a good chat this summer with one of the Freegate developers, who
was complaining that $1.5M would only pay for a trivial fraction of the
bandwidth they need. I was trying to convince him that he should put
the money toward work that would continue to be useful in the longer
term. After all, if you blow $1.5M for three months of bandwidth, what
are you left with afterwards?

Is the end goal to "deliver" as many youtube views to China as possible,
as soon as possible? That's a mighty roundabout way to tear down this
wall. Now, I think number of users who can circumvent is probably a good
measure of something, and I agree that Freegate and Ultrasurf have been
doing a good job of increasing the number of users who can circumvent. But
any senator who's thinking "we just need to push more youtube videos into
China" is in for some disappointment. The reality is much more complex --
it has to do with stronger social networks inside these countries, and
giving people tools to let them change the atmosphere from within their
country in the ways they see fit, and with making sure the tools can
provide enough safety for their users. That last point could lead into
a fine discussion about metrics and how you should define "successful",
but my big question here is whether three months of increased bandwidth
is the best way for them to spend their money.

In my world, bandwidth is free. There's some initial cost of building
relationships and convincing people that your cause is worth giving
bandwidth to, but once you've done that, the incremental cost of another
month is in maintaining those relationships. There's the Tor approach of
getting lots of smaller providers to donate bandwidth (inefficient, but
useful for distributing trust over multiple places if you want to avoid
attractive centralized targets). But if you're just trying to maximize
bandwidth for effort, go to Google and AT&T and whoever and ask to use
one of their spare gigabit links for a while. Falun Gong has a lot of
volunteers; surely some of them have connections in organizations who
can be convinced to help save the world? Even when they can't get "free",
getting bandwidth at-cost or subsidized would still be better.

Then the majority of the funding can be used for results that are more
sustainable, that is, will have an impact in future years too -- software
engineering so the tools are easier to maintain and adapt, increased
developer clue about adversary capabilities and how to play the arms
race, better tool usability and community education about how/why to
circumvent and what risks to consider, etc.

Now, here's where it gets controversial (well, more controversial, insofar
as I think they have no plans to take my above suggestion). Why keep
these results secret? The Freegate developers should be asking themselves
whether sharing some of the lessons they've learned would ultimately
help them achieve their goals better. Their current answer is "no way,
that would help the censors," and my current answer is "the more people
you bring up to speed on this topic, the more help you're going to get;
don't you think the censors are already paying attention?" In a sense,
this is the "open source vs proprietary" debate yet again. Fine. But the
*funders* should be asking themselves a similar question: is funding
a few individuals to privately get better at circumvention the best
progress we can make to put the world in the position we want to be for
the future? Or said a different way, are we really at the stage of the
arms race where every necessary strategy is well-understood, and it's
just a matter of "deploy it more and bigger and then we'll win"?

There's definitely a role for diversity of solutions here -- I think
it's good that Freegate and Ultrasurf are getting more funding. Also,
not every funder needs to have the priority of increasing the community's
knowledge with the long term in mind. Maybe it makes sense for State
to leave the research and development efforts to other funders like
DARPA and NSF, and then State's strategy can be "figure out what your
priorities are, then take the best known option at the time for achieving
those priorities and deploy the heck out of it." But people on all sides
of the question need to think about their goals, what they expect their
actions to achieve, and what they're giving up (now and in the future)
by the sustainability strategies they choose.

--Roger