Swarm

Use case

In general, think static web-page and digital asset downloads. Like FTP but with files specified by hash rather than by server & path.

Specs

Reverse hash-lookup (given hash it finds the data) with:

• Low-ish latency Lowest likely download rate of asset should be reasonable, even (especially) for assets of < 1MB.
• High throughput Average download rate of asset should be minimal.
• Private Plausible deniability against data publication or subscription.
• Lossy No implicit guarantee that other nodes will maintain data.

Existing solutions

• BitTorrent:

Basic Design

Uses the "bzz" protocol string of ÐΞVP2P.

Rest coming soon, once I've finished prototyping. Gav.