Request for Comments

A Request for Comments (RFC) is a publication series from the principal technical development and standards-setting bodies for the internet. Begun in 1969 by Steve Crocker at UCLA as part of the ARPANET project, the RFC series was originally an informal mechanism for researchers to share notes, propose protocols, and solicit feedback. The name was deliberately humble — a 'request for comments' rather than a specification or standard — and this humility shaped the internet's distinctive culture of open, incremental, consensus-based technical development.

The RFC series is numbered sequentially, and each document is never revised in place. Instead, updates are published as new RFCs that obsolete or update earlier ones. This creates a permanent, versioned record of technical decision-making that is unusual in engineering: every mistake, every revision, every reconsideration is preserved and citable. The RFC editor, currently overseen by the Internet Engineering Task Force (IETF), maintains the series and assigns numbers.

RFCs range from informational documents and best-practice guides to formal standards. An RFC becomes a standard through a well-defined process: Proposed Standard, Draft Standard, and finally Internet Standard. But many of the most consequential RFCs — including those defining TCP/IP — never formally progressed beyond Draft Standard, because the IETF eventually concluded that the elevation process added bureaucratic overhead without technical value.

The RFC process embodies a particular theory of technical governance: authority derives from running code and rough consensus, not from institutional position or formal vote. The IETF's mantra — 'we reject kings, presidents, and voting; we believe in rough consensus and running code' — is the organizational expression of this principle. It is also, in practice, a system that privileges the participants who can afford to attend meetings, write drafts, and maintain implementations.

The RFC series is often celebrated as a model of open technical governance. It is better understood as a model of how governance can function when the stakes are low and the participants share a technical culture. The RFC process works for internet protocols because protocol designers share assumptions about interoperability, because the costs of disagreement are bounded, and because implementation creates its own accountability. Whether the model scales to domains with higher stakes — AI safety standards, biotechnology governance, climate protocols — is an open question. The RFC culture's humility is genuine. Its transferability is not.