public distributed filesystem 4864

I like the feedback guys-gals. It seems like there is at least a general interest for this concept.

So far these are the points that have been raised: * file security - The consensus is that the files will be encrypted. Possibly ahead of time so that this is the job of the user and not the p2p filesystem(lets call it p2pfs for now).

* Data persistance reliability - Obviously, any p2p network is ad-hoc in nature and the protocol must be capable of dealing with this. - redundancy is necessity - possible implementations include par2 and-or RAID architecture - files are split into blocks and distributed across the community - 1 suggestion was to take a block ratio approach to prevent leeching(or dumping in this case) - a ratio system would require a valid verification method such as some type of reliable checksum approach. e.g. cksum every 5 bytes for a random block of data

Things to think about: - Since this is p2p, there should be no choke-points. i.e. eliminate the need for dedicated servers. - what existing open source projects can we build on to solve our current points above? (no reinventing of the wheel) - How to evaluate supernodes of the p2pfs community? Or is everyone a supernode? - The checksum system would require a vote within the community; how is this announced-scheduled?