However, in a direct peer-to-peer exchange the Cryptosphere does nothing to mask the transactions a particular is performing, as opposed to systems like Freenet and Tor which make an effort to obscure which host you're actually talking to by routing them through a chain of proxies. In this regard the anonymity guarantees of the Cryptosphere are no different from a system like BitTorrent, aside from the plausible deniability defense that comes from the fact all content is encrpyted and peers automatically provide storage service to other peers.
Instead, the Cryptosphere favors system robustness over guarantees on anonymity. Participants in the system maintain a history of their activities in the form of a long-chain certificate. You can think of this being somewhat like the BitCoin block chain, where the longest version always wins, and its integrity can be cryptographically verified. Every peer maintains its own long chain certificate of all its activities, including services requested and services completed.
Rather than verifying the integrity of a long chain based on hashes, the Cryptosphere uses public key cryptography. Peers requesting services sign off on both the request and delivery of a service (e.g. storing and serving a particular chunk of a file). While in isolation the data points contained within a particular long chain certificate are meaningless, peers can collect several of these certificates and build a database of other peers in the system, using tools like collaborative filtering to make intelligent decisions about which other peers are worth interacting with.
This is interesting. If i understand correctly, this means that given a transfer that you'd like to engage in, searching for a particular file you can trace back the provenance of that file through a network of peers who are making it available (if their transfer histories are accessible).
You can confirm any given peer is holding a particular portion of a file (sometimes called the "confirmation of file attack") however you can't prove they published it.
The peer "block chain" will only contain transfer metadata, not specifically which files are transfered. That information is only known by the peers involved in the exchange.
And that's where the anonymity of the system is definitely inferior to FreeNet: peers involved in any given exchange know exactly what was transferred and to what IP address.
I decided not to solve the anonymous transport problem because Tor, I2P etc are working quite diligently on that and it's a hard enough problem in and of itself. I think this has been a big stumbling block for FreeNet.
Instead, the Cryptosphere favors system robustness over guarantees on anonymity.
You trace back the "provenance of that file" through crytographic signatures. You could make your own throwaway identity, use it to publish something, and through the continued propagation of that data through the network its publication would no longer require your activity.
It should be considered psuedoanonymous publishing.
If this is just a distributed block based storage without an index that folks can search, then it's more akin to a way for you to ask folks to hold on to encrypted data for you for a while (without them knowing what the encrypted data actually contains).
I think it's more of a matter of how secure the block chain is. Note this part of the opening blurb:
To ensure quality service and prevent abuse, the Cryptosphere uses an integrated cryptographically secure reputation system which provides a distributed web of trust.
Holding someone else's encrypted data without it being anonymous doesn't sound like a problem that needs to be solved. Also, the UK has crazy encryption laws which could open any user to very bad incarcerations. Makes stealing music off of Pirate Bay or Napster look like Jay walking.
Instead, the Cryptosphere favors system robustness over guarantees on anonymity. Participants in the system maintain a history of their activities in the form of a long-chain certificate. You can think of this being somewhat like the BitCoin block chain, where the longest version always wins, and its integrity can be cryptographically verified. Every peer maintains its own long chain certificate of all its activities, including services requested and services completed.
Rather than verifying the integrity of a long chain based on hashes, the Cryptosphere uses public key cryptography. Peers requesting services sign off on both the request and delivery of a service (e.g. storing and serving a particular chunk of a file). While in isolation the data points contained within a particular long chain certificate are meaningless, peers can collect several of these certificates and build a database of other peers in the system, using tools like collaborative filtering to make intelligent decisions about which other peers are worth interacting with.
This is interesting. If i understand correctly, this means that given a transfer that you'd like to engage in, searching for a particular file you can trace back the provenance of that file through a network of peers who are making it available (if their transfer histories are accessible).