A sequence of bytes identified by a hash and subsequently a CID. The hash-based identification makes a block effectively immutable. Can be exchanged using various protocols. Can be easily cached, because due to the immutability a block can never be cached in an outdated version.
CID
content identifier: a label identifying content (in particular, an IPLD block). See also: IPFS CIDs
DNSLink
Reference to (usually) a CID, implemented using the Domain Name System (DNS). Commonly used as updatable name for content addressed data. See also DNSLink Standard
DOI
A DOI is an identifier of a digital object. It doesn’t rely on content verification (as CIDs do), but instead creates a contractual framework which encourages DOI providers to properly take care of the referenced objects. It’s commonly required for scientific publication. See doi.org for general information and DataCite Docs for information about DOIs for scientific datasets.
To keep (and if required fetch) data identified by CID. In many cases pinning is done recursively (i.e. data referenced by the pinned CID is pinned as well).