Summary

  • Revision -06 of draft-knodel-nomcom-gender-representation, uploaded on 7 October, raises the proposed maximum guarantee for a voluntary opt-in pool from three seats to five.

  • It remains an active individual Internet-Draft, not an adopted IETF rule. Datatracker says individual I-Ds are not endorsed by the IETF and have no formal standing in the standards process.

  • Self-declarations are not verified. RFC 8713's challenge process for the selection remains, but the draft says a challenge cannot be based on a volunteer's declaration.

  • If the opt-in pool contains five or fewer people, all are selected subject to the existing affiliation limit. The draft explicitly recognizes the resulting incentive to declare, including for organizations seeking seats.

  • RFC 8713 and the current 2026–2027 NomCom page specify ten voting volunteers, while -06's rationale and worked example use eleven voting seats. The available sources do not explain the difference.

A stronger proposal, not a change to IETF rules

Mallory Knodel uploaded revision -06 of draft-knodel-nomcom-gender-representation, co-authored with Tara Tarakiyee, on 7 October. The IETF Datatracker entry classifies it as an active individual Internet-Draft and states that individual I-Ds are not endorsed by the IETF and have no formal standing in the standards process. No RFC stream or intended RFC status is listed, and the IESG state is “I-D Exists.”

That procedural boundary matters. The 7 October change is a revision to an individual proposal for NomCom selection, not a change to an adopted IETF rule. The frozen record provides no basis to say the proposal has been adopted, endorsed, scheduled for implementation or assigned to an IETF working group.

The Datatracker version history shows revision -05 on 3 October and -06 four days later. The new version increases the maximum guarantee from three seats to five, makes explicit the proposed duty of the NomCom chair to name the “dominant gender” in the call for volunteers, allows an erroneous opt-in to be corrected before the general pool list is published, and adds a worked selection example.

A self-declared pool without verification

The proposal creates an opt-in pool for eligible volunteers who declare that they do not identify as members of the “dominant gender” named by the NomCom chair. Volunteers are not asked to state their gender.

Nor is the pool to be constructed by inference. The draft says gender must not be inferred from pronouns, mailing-list discussions, meetings or Datatracker data. Public gender or pronoun disclosure in Datatracker remains voluntary, as does gender information collected through meeting registration; neither may be used to construct the pool.

The declaration itself is not verified. Instead of establishing precise rules for checking it, the draft says the mechanism relies on participants' honour and integrity. Control over whether to opt in therefore rests with the volunteer rather than with an external process for determining gender.

That is separate from a challenge to the selection. The challenge period under RFC 8713 remains, but the draft says a challenge cannot be based on a volunteer's declaration. Correcting an accidental opt-in, challenging the selection and publicly verifying that the selection mechanics were followed are distinct controls.

How five guaranteed seats change the odds

Revision -06 defines the guarantee as:

Here, p is the size of the opt-in pool and r is the guaranteed number. If three people opt in, the guarantee is three; at five or more, it is capped at five.

If p = 0, there is no guarantee. The draft says the community must then be notified that all voting volunteers may share one gender for that reason.

The rule does not create additional NomCom seats. It calls for one selection under RFC 3797 using the published general pool list. The guarantee changes an ordinary random result only when that result would seat fewer than r opt-in members. Existing affiliation constraints and fallback rules still apply.

The incentive effect becomes particularly strong when the pool is small. The draft says members of a small opt-in pool are much more likely to be seated. If five or fewer people opt in, all are selected, subject to the affiliation limit.

The draft explicitly identifies an incentive to declare and includes the case of organizations seeking seats. It also says the affiliation limit constrains how much any single organization can gain. Those are incentives created by the proposed mechanism, not evidence that any volunteer or organization has acted on them.

Independent verification requires a durable disclosure

The design seeks to make the selection independently checkable. A single RFC 3797 selection is conducted against a published general pool list that identifies opt-in members. Revision -06 says this makes the outcome independently verifiable.

The same text identifies the privacy cost. The list is archived, membership in the pool can be compiled across years, and for some volunteers the declaration may disclose more than they otherwise made public. Volunteers must be told about that consequence when declaring.

The correction window ends before publication of the general pool list. What happens afterward is different: the list becomes a persistent public record, is archived, and can support repeated cross-year compilation. RFC 8713's challenge process and the affiliation cap perform still other functions. Neither is a mechanism for reversing an already public disclosure.

The proposal therefore avoids building the pool by privately inferring gender from other information, while requiring enough public information to let outsiders reconstruct and check the selection. Verifiability gains an evidentiary trail, but the disclosure cost falls primarily on those who choose to opt in.

A small pool could also produce repeated service

Higher selection odds may have consequences beyond a single NomCom cycle. The draft warns that a small opt-in pool could result in the same volunteers serving repeatedly.

That can impose a cost on the volunteers themselves, but it can also affect the candidate pool. Under RFC 8713, sitting NomCom members are ineligible for positions that NomCom fills. Repeated service by the same people could therefore both place recurring demands on those volunteers and remove them, while serving, from the pool of candidates for NomCom-selected positions.

This is a possible second-order effect identified by the proposed mechanism, not evidence that repeated service will necessarily occur or that any particular volunteer has behaved accordingly.

Ten voting seats in the current rules, eleven in -06

A numerical discrepancy remains unresolved. RFC 8713 specifies ten voting volunteers for NomCom candidate selection; the chair, liaisons and advisors do not vote on candidate selection. The official 2026–2027 NomCom page likewise lists ten voting members.

Revision -06, by contrast, says a cap of five approaches parity for a NomCom with eleven seats. Its worked example also models eleven voting seats.

The six primary sources in the record do not explain the difference. Because the eleven-seat assumption forms part of the numerical rationale for the five-seat cap, it merits clarification in a later revision. The mismatch alone does not establish that the draft is invalid, legally defective or motivated by any particular intent.

Context figures do not measure the eligible volunteer pool

The draft says five of the twelve NomComs seated from 2015 through 2026 had no women among their voting members. That figure is a claim made by the draft and should be treated as such. It does not establish that a committee's composition caused any particular leadership-selection outcome.

Separately, the IETF 2025 Community Survey reports that women were 9.82% of regular participants in 2025 and under 10% of all survey respondents.

Those figures describe populations measured by the Community Survey, not the eligible NomCom volunteer pool. They therefore cannot be used directly to estimate the future value of p, the number of opt-in declarations, or how often the five-seat guarantee would alter a random selection.

Sources