Previous topic - Next topic


We all own two-way hardware: computers and phones that send and receive data. These devices could relay data if there was a way of coordinating the supplies/demands of relay-resources (in particular, solving the free rider problem -- as was brilliantly good-enough-solved by Bittorrent, and somewhat solved in TOR -- ).

Bitcoin, with cheap, private, micropayment channels, is the perfect solution to this coordination problem. From there, we're just a few market-heuristics (phones automatically [1] estimating their ability to provide a 'useful relay' and [2] auctioning off these relays) away from a complete solution.

If Bitcoin could do this, it alone would enable a cheap, but more importantly, private/censorship-proof internet. Anyone who wanted to use this network would need (a small quantity of) Bitcoin.

So I have long believed that this 'global data transfer market' is the true "killer app" for Bitcoin, which would give it undeniable intrinsic value and force it to inevitably take over the planet.

Someone claims to be working on that solution: Of course, they haven't actually submitted anything (as far as I can tell).

I certainly hope that someone works out these details soon.
Nullius In Verba


Paying more to download data faster is illegal under net neutrality. We can not legally run the system you are describing in USA.

Users will have to pay in both directions.
1) The people downloading data pay to download. They can pay more to download faster.
2) The people uploading data pay to make sure their data stays available for a period of time.
For popular content, I expect servers to serve for free in the hopes of making a profit on downloads.

Vitalik used math to solve related problems.  When people publish content to the internet, they want that content to stay view-able to others, even if their own computer turns off. They want the data to be stored redundantly, so even if some nodes go offline, the data is still available. They don't want the data to be too redundant, because it takes up too much space. Vitalik shows how if you make your data just 10-20% bigger, you can gain all the redundancy you need.

IPFS (interplanatary file system) is a related project: They are trying to be the easiest distributed file system for building stuff like this on top of.

storj is trying to solve the subset of this problem where you are serving files to yourself. Like storing a backup of your hard drive.

You can learn more about the payment channels Paul mentioned:
Since these payments happen off-chain, there is no transaction fee. Which lets us afford to make very small payments for kilobytes of data at a time.