ystem) Introduced in 2015, IPFS is a distributed storage system on the Internet that uses "content-based" addressing. The Internet file system most people are familiar with is "location-based," which points to a file on a website; for example: www.abc.com/image.jpg. If the website becomes unavailable, the image.jpg file is also not retrievable.
In contrast, IPFS files are stored on multiple nodes and are addressed by a content identifier (CID), not a physical location. The CID contains the hash of the file, which is a digest of the file's data. Given the content identifier, the IPFS finds the actual file. The hash computation ensures that there cannot be multiple files with the same content but with different file names. IPFS is popular for storing non-fungible tokens (see NFT
). See CID
Like Bitcoin, Volunteers Are Required
An IPFS is maintained by interested parties, and the more volunteers, the greater chance that the data will be available if one or more nodes go offline either temporarily or forever.
Immutable Like a Blockchain
Once files are added to an IPFS, they can be removed but they cannot be altered. However, files can be updated by creating new objects that link to the previous ones. Thus, IPFS creates its own kind of versioning system. Filecoin's blockchain storage system is based on IPFS (see Filecoin
Let us just say that the creators of IPFS have ambitious goals!
Files Are Broken Into Objects
Files are stored as one or more 256KB objects. For files larger than 256KB, the first object contains no data; only links to the other objects.