One man's ongoing effort to make sense of the world.

Sunday, February 12, 2006

Hack China!

In Google, China and the Cult of Mammon I identified the broader enemy, Mammon, and the specific target, the Communist regime of mainland China. I also hinted at the means of attack. Now I will go into more detail. But first, I need to define some terms.

First, the *players*. There's us - we in the free world who are willing and able to act. Then there's the enemy, the ChiCom government. There are also the people of China. Let's divide these into the active dissidents and the masses. Finally, there are collaborationists - corporations that are helping the ChiComs stay in power.

Now, the tools of the enemy. They are *censorship and fear*. We can't attack the main part of fear directly. The people of China need to overcome this themselves. Those we call "dissidents" have already done so. But we can attack censorship. (And by way of attacking censorship, we can undermine the fear a bit. To weaken censorship is to visibly weaken the regime's control, hence its prestige, hence its capacity to intimidate the populace.)

The censorship has internal and external components. *Internal* censorship is when the regime prevents its subjects from talking to each other. *External* censorship is when the regime prevents its subjects from communicating with the outside world.

The censorship has two modes: prevention and intimidation. *Prevention* is when the regime uses technical means to make communication physically difficult or impossible. *Intimidation* is fear - fear of punishment if you engage in proscribed communication. Intimidation depends on being able to spy on communication, and being able to track down the participants. When Yahoo rats out a dissident, they are collaborating in the intimidation.

The communication which censorship opposes has two *aspects*: accessibility and searchability. Accessibility means the communicator can distribute information in a way that an intended communicant can receive it if he knows how to find it. Searchability is the ability of a would be communicant to find such information without knowing in advance exactly where to look for it.

Searchability is what Google's latest collaboration threatens. In order to read something, you have to know where it is. And knowing where it is, is in itself information, and can be censored.

Communication comes in three *scopes*: public, group and private. Public communication is publishing - making information available to anyone who wants to read it. Group communication is communicating to a select group. Private communication is sent to a single individual.

The communication has *choke points*, at which it is vulnerable to censorship. These are routers, gateways, and, as far as searchability is concerned, Web search engines. When a router or gateway filters transmission, the code that accomplishes this is called a *firewall*.

There are means of working around all of these. But it's not easy. I insist that it's possible and worthwhile, but it's not easy. A future post will go into detail about the state of the art. For now I'll just name the concepts involved. Don't worry if you don't understand all the terms here. If you're curious, you can Google (with luck!)

Every choke point seizes on some identifying aspect of a message. Call these the *shibboleths*. A router or gateway uses and IP address - the address of a computer on the Internet as a shibboleth. Another shibboleth used here is a port number. The way to defeat these is to use a *peer-to-peer* network, that does not use a fixed port number. A peer-to-peer network is a network within the Internet, by which various nodes (computers on the network) find each other and exchange data.

A more insidious choke point is *content*. Google refuses searches on certain keywords. More advanced firewalls (and be sure, China has them) filter on keywords in the data being sent. The way around this is encryption. Encryption also makes private communication possible, if you use it properly.

Now what about public communication? You've got a message that you want to get out to all of China, and that means the authorities can see it to. How to do so without revealing yourself, and ending up in the gulag? What you need is an encrypted peer-to-peer network that uses *multiple hops* so that the recipients don't know where the message came from. Because, sooner or later, the authorities will discover the network and view the public information on it, by the same means that everyone else does. If the network is properly designed, they will find a *node* but that won't help them track down the sender of any given message.

So, we need a peer-to-peer network with multiple hops, encryption, and a search mechanism. We need to extend it into Communist China. We need to grow the network once it's there. It's not much use until it gets large. Ideally, every person in China who has a PC should have access to this network. In practice, the regime is going to find about this network at some point and try to shut it down. But they can't shut it down all at once because they can't find it all at once. So we need a network that will grow itself faster than they can kill it off.

When the network gets large enough, censorship in China will be unworkable. In its early days, the component that's actually in China will be small. Most of those on it will be the dissidents. Once the network has put down roots, the dissidents can spread by any way they choose. We just have to get it to them in the first place, preferably without breaking any of our society's laws. (We can work on getting the laws changed at our leisure.) One way to do this is by what I call a *benign trojan*.

A benign trojan is a piece of software to run a node of this network, plus the addresses of a few existing nodes, called seed nodes, all hidden in something else. Something the authorities won't object to their citizens downloading. Perhaps some other sort of software, or an innocent looking file. It could be as simple as renaming the file. But the dissidents who stumble upon it will be able to figure out what to do with it. They'll extract the payload, and start up the nodes. Think of it as sneaking a hacksaw past the guards in a cake. (A regular trojan takes over the computer for malicious purposes. A benign trojan doesn't take over the computer, it helps take over a society.)

Once it's there, it will spread, perhaps by more benign trojans, perhaps by other means of the people's own devising. But won't the regime notice? Won't they try to shut down or block the Web sites with the benign trojans from the free world? Won't they lean on the Mammon worshipers here to get those files taken down? Yes, they will. And the files will be taken down. But it will be too late.

What will the regime do next? Try to hunt down the nodes. And they're bound to find some of them. But if we've designed the software right, they'll never find them all. New nodes will come into existence faster than they can track down the old ones.

And then what? They can stop the external component of the communication only by disconnecting the internet from the global Internet. That will devastate their economy, thus weakening their position. Remember what happened to the Soviet Empire when Reagan got stingy with the aid?

And there will still be the internal component. The only way to shut that down is to shut down the nationwide internet. Economy weakened still further. The government looking weak. Looking desperate. The very fact of their taking such drastic actions will tell the people just what the regime most wants the people *not* to know: that the rulers are afraid.

Shut it all down, or don't shut it all down. Either way, they lose.

The hard part is designing the software right. People have tried.. But they always leave out something important. That's why I've tried to enumerate every factor here. Have I missed anything? Let's have some peer review. Be as merciless as you like in your comments, so long as you make sense. Tear my work apart, and I'll patch it back together again later. When it gets to the point that you can't tear it apart, that will the time to start writing code.





Angelfire link (turn off Javascript to avoid popups)

Freenet: SSK@jbf~W~x49RjZfyJwplqwurpNmg0PAgM/marlowe/23//hackchina.html

0 Comments:

Post a Comment

<< Home