IPFS primary sources — 2026-05-09

Classification takeaway

IPFS is best cataloged as content-addressed routing and retrieval infrastructure, not as a durability guarantee. The key reusable mechanism is the combination of CIDs, Merkle-DAG-linked content, peer routing, and block exchange, with persistence delegated outward to pinning behavior, service operators, or adjacent incentive networks such as Filecoin.

Primary sources reviewed

1. Main docs homepage — https://docs.ipfs.tech/

   • Best short statement of the project’s self-definition: open protocols for addressing, routing, and transferring data on the web.
   • Useful because it immediately narrows the analytical scope away from generic “storage” marketing.

2. What is IPFS? — https://docs.ipfs.tech/concepts/what-is-ipfs/

   • Important classification page.
   • Explicitly says IPFS can refer to a network, an implementation family, or a modular protocol suite, and just as explicitly says IPFS is not a storage provider or cloud service.

3. How IPFS works — https://docs.ipfs.tech/concepts/how-ipfs-works/

   • Best practical mechanism source.
   • Breaks the system into representation/addressing, routing, and transfer, then maps those responsibilities onto CIDs, IPLD, Kademlia DHT, Bitswap, gateways, CAR files, libp2p, IPNS, and DNSLink.

4. Persistence, permanence, and pinning — https://docs.ipfs.tech/concepts/persistence/

   • Most important source for understanding what IPFS does not guarantee.
   • Shows that nodes can garbage-collect unpinned data and that long-term retention depends on pinning, paid services, or Filecoin-style incentive layers.

5. Content addressing / CIDs — https://docs.ipfs.tech/concepts/content-addressing/

   • Best source for why CIDs are more than checksums.
   • Useful because it explains chunking, DAG structure, codecs, CID versions, and why the hash inside a CID need not equal a simple file hash.

6. Academic papers index — https://docs.ipfs.tech/concepts/further-reading/academic-papers/

   • Points to the original whitepaper and the later ACM measurement paper.
   • Useful for tying the current docs stack back to the original architecture and later empirical evaluation.

7. Specs repo — https://github.com/ipfs/specs

   • Confirms IPFS is maintained as an evolving protocol/specification family rather than a single product surface.

Whitepaper / technical-doc status

• The original whitepaper has been saved locally as ipfs-whitepaper.pdf.
• The docs are unusually strong for practical mechanism analysis and are the best source for the boundary between IPFS itself and adjacent persistence services.

Why it stays in the active corpus

• It is the baseline comparison class for content-addressed retrieval and naming.
• It helps avoid analytical blur between routing/discovery systems and true long-term storage or durability systems.
• It is especially useful alongside Filecoin and Archivist because it clarifies which layer supplies discoverability, which layer supplies incentives, and which layer supplies explicit durability engineering.