Spun out from #10 (point 8, raised by @mx5kevin).
Problem
EpixNet today is a closed P2P swarm. Files available on EpixNet aren't reachable from BitTorrent/IPFS clients and vice versa, which:
- Forces users to run EpixNet as a separate program alongside their existing download tools
- Makes it harder to seed the same large files across networks without duplicating storage
- Limits discoverability of EpixNet content
AnnounceBitTorrent announces to BitTorrent trackers for peer discovery, but does not bridge the protocols themselves.
Proposal
Explore protocol bridging in three directions:
1. Magnet URI / torrent import
2. IPFS CID references
- Sites can reference IPFS CIDs for large/immutable content
- EpixNet client fetches via IPFS gateway or embedded node
- Optional: publish EpixNet files to IPFS for wider availability
3. Outbound torrent seeding
- Files hosted on EpixNet can be made available as torrents
- Uses existing hashfield + push swarm as the seeder backend
Design questions
- Bundle a BT client (libtorrent) or shell out to external client?
- How do infohash -> content.json references get signed/verified?
- What's the trust model when pulling from external networks?
- Per-site opt-in or global?
Non-goals
- Not replacing EpixNet's native swarm, this is additive
- Not claiming feature parity with dedicated BT clients
Credit: @mx5kevin in #10.
Spun out from #10 (point 8, raised by @mx5kevin).
Problem
EpixNet today is a closed P2P swarm. Files available on EpixNet aren't reachable from BitTorrent/IPFS clients and vice versa, which:
AnnounceBitTorrentannounces to BitTorrent trackers for peer discovery, but does not bridge the protocols themselves.Proposal
Explore protocol bridging in three directions:
1. Magnet URI / torrent import
2. IPFS CID references
3. Outbound torrent seeding
Design questions
Non-goals
Credit: @mx5kevin in #10.