Summary

  • The IETF’s training RFP requires a self-paced course and minimum, opt-in data collection; its 31 August clarification says neither learner assessment nor individual tracking is expected.
  • Those privacy limits do not prevent accountability. The course can be evaluated through its content, accessibility, currency and correction record rather than a persistent file on each learner.
  • As of 10 October, the public record showed a status mismatch: the 6 October board report said proposals were being evaluated, while the RFP register still marked the item OPEN. That is a disclosure problem to clarify, not proof of a failed procurement.

The online course now being procured is meant to make a difficult process easier to enter. The IETF’s 12 August request for proposals describes a self-guided, on-demand programme built from its existing day-long New Participants course. It asks for six areas: why to participate, an introduction to the IETF, how participation works, standards development, Hackathons, and Internet-Drafts and RFCs. The modules should stand alone while forming a coherent course. The intended delivery is in English. The RFP does not say the course has been awarded or built.

A welcome path with a privacy boundary

The strongest design choice appears in the procurement clarification. The 31 August Q&A says no learner assessment is expected and no tracking is expected. The RFP separately calls for only the minimum data required, collected on an opt-in basis and handled under the IETF Administration LLC’s privacy practices.

That is more than a narrow technical setting. A course can explain how to write a draft, join a working-group discussion or test an idea at a Hackathon without building an account of which person watched which lesson, how long they stayed or whether they passed a quiz. The procurement documents describe an educational resource, not a participant-scoring system.

But “no tracking” cannot answer whether the resource is useful. The RFP itself recognises that material will age: it requires modules that can be updated independently, editable source files, captions and transcripts where practical, and a review-and-revision period with the LLC and relevant volunteers. The Q&A also says bidders need not add special support for constrained connectivity beyond the general delivery platforms. “Self-paced” therefore describes scheduling, not whether someone on a metered or intermittent connection can reach the material.

The documents do not publish a quantified outcome framework or say what evidence will settle whether the course succeeds.

Measure the course, not the person

The distinction is practical. A program-level record could list which of the six modules have been released, when each was last checked against current IETF process, whether captions and practical transcripts are available, what corrections were made, and who owns the next update. A voluntary anonymous feedback channel could identify confusing material without joining a response to a learner’s identity or later activity. The decision to use such a channel is not documented; it is one possible design.

These measures would not prove that a viewer became an effective contributor. They would answer smaller questions that an organisation can act on: Is every required topic present? Can a module be revised without rebuilding the whole course? Are accessible assets delivered? When a process changes, is the affected lesson updated? An explicit limit on what cannot be known is more credible than a proxy score presented as proof of learning.

The current IETF New Participants page defines a new participant as someone who has attended fewer than five meetings, remotely or onsite, and separately lists meeting resources and a guides programme. The proposed online course should therefore be read as one access layer among several, not as a replacement for personal guidance or the work of existing participants. Heng Lu’s Note 73 is a useful editorial lens here: participation can broaden understanding and surface affected interests, but attendance or training does not itself create representative authority. A course can lower the cost of taking part; it cannot certify that a process represents everyone affected by it.

The status record has its own test

The procurement timeline deserves the same distinction between evidence and inference. The RFP set 11 September for final proposals, 21 September for preferred-bidder selection and the start of negotiations, and 28 September for contract execution and work to begin. The public executive-director report for the LLC Board meeting on 6 October says the RFP had closed and responses were being evaluated at the time of writing. Yet the public RFP register captured on 10 October still showed the entry as OPEN and listed no vendor.

The divergence does not establish that a deadline was breached. The schedule is a planned sequence; the report describes a later evaluation stage; and the register may simply be out of date. Together, however, they show why a transparent process needs a dated status, not only a static schedule. “Open”, “closed”, “evaluating”, “selected” and “contracted” are different states. A last-updated date and one current phase would let readers distinguish a delay from a stale page.

For now, the public evidence supports a narrow conclusion. The IETF has specified a privacy-conscious training procurement and a review process, while leaving public outcome measures and the current vendor-selection state unclear. The online course is expected to be deployed by 31 January 2027, but the sources reviewed here do not show an award, contract, launch or learner result.

Sources