Summary

  • RFC 3869 connected the Internet’s shared infrastructure to a funding problem: no single organization captures all the benefits of research into its general foundations.
  • The IAB recommended consistent non-commercial support but explicitly did not propose that the IAB, IETF or IRTF administer that funding.

The most revealing sentence in RFC 3869 arrives after its recommendation. The Internet Architecture Board had argued that research into the Internet’s common infrastructure needed consistent, sufficient non-commercial funding. Then it drew a line around its own authority: the document was not suggesting a direct or indirect role for the IAB, IETF or Internet Research Task Force in handling the money.

That boundary turns a familiar appeal for research support into a question of institutional design. Who should pay for work whose benefits are distributed across operators, vendors, researchers and users, and who should decide which work receives support? RFC 3869 argued that the absence of a single owner created a weak funding case for research on general Internet infrastructure. A firm might rationally finance a project that yields a product advantage, but the payoff from improving shared protocols, security or reliability can spill beyond that firm.

The document’s diagnosis was that commercially supported work therefore tended toward nearer-term competitive returns, while some higher-risk or infrastructure-wide questions lacked a natural sponsor.

The authors did not present commercial funding as a villain. They acknowledged its practical role: implementations, deployment and follow-up evaluation matter, and commercial support can help move research into products. Their concern was about relying on that source alone. If the expected private return is smaller than the wider public benefit, a useful research question can remain unfunded even when many parties would benefit from an answer.

RFC 3869 placed this concern in a short history. It described early packet-network and Internet research as heavily supported by DARPA, then recounted 1980s U.S. government, NSF and other public support for research, prototypes and network operations. It described reduced U.S. government research support from the mid-1990s and a stronger commercial role, alongside other national and regional sources. That chronology is the IAB’s 2004 account, not a modern funding database. It should not be read as a current measurement or as proof that every commercial project serves only a short-term interest.

The document’s open-research list also had limits by design. Its authors called the list incomplete and illustrative, not a comprehensive grant-agency guide, a ranking of proposals, or a prescription for what a funder must choose. It could make the case that many questions remained without claiming the authority to allocate funds among them.

That restraint matters because RFC 3869 was not creating a grant program. It named no administering institution, appropriation, review process or award mechanism. Instead, the IAB supplied a diagnosis and recommendation while declining to convert technical stewardship into financial authority. The distinction is easy to miss if one compresses the conclusion to “the IAB wanted more funding.” It wanted the research to be financed, but it did not propose that the architecture or standards bodies become its fund manager.

The surrounding institutions reinforce the need to keep unlike budgets separate. RFC 2014 and RFC 7418 describe the IRTF’s research-group role and its different expectations from IETF standards work; RFC 5743 describes an IRTF publication stream that is not itself an Internet Standards Track. RFC 8712, by contrast, concerns the IETF–Internet Society administrative, legal and financial relationship after the IASA 2.0 changes. RFC 8729 and RFC 8700 recount the RFC Series and RFC Editor’s publication infrastructure.

Those documents illuminate research, standards and operating support, but they do not turn RFC 3869’s research question into an IETF operations budget question.

Seen retrospectively, RFC 3869’s contribution is as much a separation of roles as a funding plea: shared technical questions can need durable support, while the body that articulates that need need not be the body that controls the purse. The RFC says nothing about whether that institutional design was later implemented or whether the 2004 diagnosis still describes current funding. Its value here is the boundary it made explicit.

Sources