Zed Shaw: Why I Don't Use Tor

(I'm in no way affiliate to NearMetter, it's just the best way I found to easily read a twitter conversation)

I've tried Tor in the past and I stopped because:

* It's really slow.

* It's the chatroulette of really questionnable material. You stumble into some shit and think WTF?!

* The amount of traffic it generated caused my shitty router to slow down significantly or crash completely.

It's certainly true that humans have all manner of interesting behaviors owing to the fact that we're smart apes with huge numbers of survival heuristics. I would pause before taking a sandwich from Hitler, because I'm human, but it's not pertinent to the question of whether the sandwich is any good. (Except in as far as you think it more or less likely that the sandwich is poisoned etc.)

So I find the whole first half of the text to be a flabby way of saying that the arguments of dishonest people need to be evaluated more critically than those of honest people. But I find that the arguments of honest people need to be critically evaluated too. I think that the authors of Haystack were honest, but their assertions turned out to be dangerously wrong. (Which, by the way, we know thanks to Mr Appelbaum.) So, as a guide, the motives of the author don't seem to be very useful to me.

Then, in the second half, we find a mixture of arguments that I find valid, and many that I don't. A sense of vertigo at the amount of trust that we have to put into software is justified. It is possible to hide major bugs in code and we're standing on a stack of hardware, kernel, and userland which is incomprehensible to any one person these days.

It's also true that there are some fairly effective attacks against Tor for the capable opponent. It's a real-time mix-net, with all the tradeoffs implied and it generates a lot of research. I recommend reading some papers of the papers, I find them often to be very good.

But accusing the Tor people of being NSA agents because they once got funding from the navy doesn't hold water. The Internet was an ARPA funded project. Military spending has subsidised much of the modern world.

Many people have read through Tor's source and evaluated the protocol etc. Of course, all those people could be NSA agents too, publishing fake papers. You could, in fact, be in The Matrix. But you probably aren't.

Some, likely massively exaggerated, secret project might be monitoring every ISP on the planet and thus able to break any real-time mix net, but they probably aren't.

Likewise, all the Tor node operators that I have met might all be NSA plants, but they probably aren't.

And finally the author picks out Mr Appelbaum for special criticism because he connects him with Wikileaks. I think his assertion that the goals of Tor and Wikileaks are in conflict is wrong, but we could go around all day trying to pin down the goals of Wikileaks so that's probably not fruitful. But it does seem ironic that the author voices support for Wikileaks right after asserting that such supporters are not to be trusted.

So, while the stack of software is, indeed, large, Tor remains a reasonable tool to use. If the author is so concerned with the human aspect, the Tor authors are make regular appearances at conferences and are wonderful people to meet. So do, and are, node operators in my experience.

Also, on top of Tor, there's a fair chance that the author is using a browser who's network and SSL stack I've had a hand in. And who knows what kind of person he's taking a sandwich from now?