Mike Burr - log

[BT®] flash-cdn

Tomorrow at 11:30am, your time, literally millions of host (PCs, laptops, servers...) are going to simultaneously open up and start serving some arbitrary pool of data...which almost certainly includes copywritten movies, etc. Lots of crazy illegal bits and quibits, anything!

If one of these millions fails to do this, that person will lose a large stake that was put up to assert that they will run the server...and this will be checked automatically.

So, in essence, you "know" with a large financial certainty, that at least X hundred million wallets around the world (you will have an actual number) will do this: Just run the server, collect your deposit and a small reward for running the server.

You have to keep the software running for (ummm) 72 hours in order to recover all of your stake (plus rewards). There's a cliff. If you don't comply, you get slashed depending on your level of compliance. It's severe but as long as you act in good faith for the next few days (?) you are good. You can even have five servers running simultaneously. You're only asked to run one, but redundancy adds no complexity or real cost and lowers risk.

If you download, the open-source client, do some crypto stuff, and run the client along with everyone else, at that same time, you can pocket $1000 worth of crypto (sorry future folk, that's a used moped for us; a bit over an ounce of gold.)

Would you do it? Remember, you can be identified! You are illegal server operator number 23,229,073! Go to the police station and get in line, you're busted!

The point is, this kind of thing can't be organized because before "critical mass", everyone is afraid of prosecution.

We can leverage blockchain to provide incentives to participate as described above, the only goal being to go from zero to "greater than critical mass" instantly, so if folks are enjoying the network, they might choose to leave their servers running. They might could be expected to be volunteers from that point forward. We only needed to get to safety-with-many and hopefully such a network can just continue to exist, unless we expect some random "examples" will be chosen from the N hundred million criminals, and patiently prosecuted. Even then, we could gauge the numbers pretty well: they're prosecuting less than 100 per year and there are 900,000,000 nodes running. I would risk it.

Et cetera.


Flaw: How do you get the content ready to go. The hosts are discoverable before we reach critical mass. The content needs to be available in cleartext to the node that will serve content. How? It can't be "oh just donkey porn" sitting on your hard drive until we have our hundred billion.

Flaw: The "stake deposit, reclaim deposit" obviously consumes no value, but how do we have "rewards"? Simple donations? "If this contract has N good-faith participants and M coin-per-participant in rewards. It is therefore a 'go'!"

Flaw: Proof of participation involves some kind of oracle. I. I guess. Maybe there's an orwellian system: Spy on your neighbors. If you see something say something. Get paid for narking.

Idea: Before critical mass, file-hosting-wanters add (references to) requests that they have, along with a bounty: "If you serve file xxx located at yyy, with hash zzz, I will pay you $$$ crypto coins." Mmaybe. If you serve those, as you proposed to do, placing your deposit at stake, you get the reward, or some of it. This exact fact can be confirmed: Can we fetch file with hash xxx at path yyy (probably just the hash, like ipfs), and you respond with the file along with a signature from the same wallet you used to propose, you can collect Foo! This lets you keep the "illegal" file on a spare hard drive, and only "put on the internet" the day of the "flash". [But of course you are an IP address that is serving naughty file X, and can be identified. Crypto Magic, send help!]

- 1 toast