LibraryStorage and networks2014Design paperCorpus record
IPFS - Content Addressed, Versioned, P2P File System
IPFS. Juan Benet.
Name a file by its hash. Retrieve it from any peer who has the bytes. Version history is a graph of hashes, not a folder on one machine.
A reading of the public document. Not a copy of it, and not a claim about a later network that reused the name.
A project that says 'stored on IPFS' has named an address. It has not named a retainer.
The five-minute read
The defect
The web addresses servers. A file that moves servers changes its name. A file you wanted can be replaced at the same name.
The rule
Name a file by its hash. Retrieve it from any peer who has the bytes. Version history is a graph of hashes, not a folder on one machine.
How it is put together
Content addressing means the name is the data. Peers exchange blocks. There is no promise that someone is storing the data tomorrow.
Where the claim stops
IPFS is not Filecoin. It does not pay anyone to store.
One action, walked through
- Chunk a file.
- Hash the chunks.
- Ask the network for the hash. Any peer with the chunk can answer.
- Who is paid, if anyone, to keep the bytes?
The argument, unpacked
Why it is still on the desk
A project that says 'stored on IPFS' has named an address. It has not named a retainer.
After the text
Filecoin later added payments and proofs on top of this addressing. The 2014 draft is the retrieval idea.
What has to be true
- IPFS is not Filecoin. It does not pay anyone to store.
- A hash does not prove availability next week.
- The draft is not the current implementation spec.
What happened after the paper
Filecoin later added payments and proofs on top of this addressing. The 2014 draft is the retrieval idea.
What to check before you use the idea
- Who is paid, if anyone, to keep the bytes?
- Is the name a hash of the content?
- What answers a fetch when no peer has the chunk?
Terms
- Content address
- An identifier computed from the bytes.
- Pin
- A node's decision to keep a chunk. Not a global guarantee.
The problem the paper names
The web addresses servers. A file that moves servers changes its name. A file you wanted can be replaced at the same name.
What the design proposes
- Content addressing means the name is the data.
- Peers exchange blocks.
- There is no promise that someone is storing the data tomorrow.
How the mechanism is specified
- Chunk a file.
- Hash the chunks.
- Ask the network for the hash. Any peer with the chunk can answer.
What this page does not treat as proven
- IPFS is not Filecoin. It does not pay anyone to store.
- A hash does not prove availability next week.
- The draft is not the current implementation spec.
Why a venture studio still reads it
A project that says 'stored on IPFS' has named an address. It has not named a retainer.
This is Blockchain Lab's reading of a public design paper. It is not the paper, not a copy of it, and not an offer of tokens, equity, custody or a partnership. Later network behaviour can diverge from the text. Nothing here is investment, legal or technical advice.
Research status: Design paper. Last reviewed: 1 October 2026. This is a reading of a public paper, not investment, legal or security advice.
