Summary

  • The current Procedures for Selecting Trustees page says it was modified by Board Resolution 2026-13. The current Travel Policy for Trustees says it was amended by Board Resolution 2026-13 on 6 July 2026.
  • The formal register for unanimous written consents identifies Resolution 2026-13 as approval of amendments to the Trustee Travel Policy. The 10 June meeting page records approval of the trustee-selection amendments but still displays Resolution 2026-XX.
  • This proves a conflict in the public decision identity, not that either policy is invalid, that either vote failed, or that the Board's internal records reuse the number.
  • Internet Society should publish one canonical receipt per Board decision, linking the unique identifier, action, approval route, dates, approved policy version, source record, public projections and correction history.

One number, two answers

Ask the Internet Society's website a simple question: what did Resolution 2026-13 approve? The answer depends on the page.

The page containing the Procedures for Selecting Trustees begins, “As Modified by Board Resolution 2026-13.” Those procedures govern how Organizational Members and Chapters elect trustees, how candidates enter the process, how petitions work, how ballots are cast and how challenges are handled. The bylaws make the connection explicit: those elected trustees are chosen according to procedures set forth by Board resolution.

Open the Travel Policy for Trustees and the same identifier appears again. This time the header says the policy was amended by Board Resolution 2026-13 on 6 July 2026. The Board's register of resolutions approved by unanimous written consent agrees. It gives a title—“Approve Amendments to the Trustee Travel Policy”—describes revised approval and receipt rules, says the policy became effective immediately and records unanimous written consent on 6 July.

The public record for Board Meeting No. 195 supplies the third piece. On 10 June, the agenda included approval of revisions to the Procedures for Selecting Trustees. It publishes the whereas and resolved clauses, but the heading still reads Resolution 2026-XX. Meanwhile, the general resolution index presently exposes meeting resolutions only through Meeting No. 194, whose 2026 sequence ends at 2026-12, and points separately to the e-vote register.

The records therefore do not merely omit a detail. They make the identifier that should connect decision to policy branch in two directions.

A public collision is not a legal verdict

The cautious conclusion is narrower than the visual contradiction.

The reviewed pages do not prove that the Board failed to approve either set of amendments. Meeting No. 195 publishes resolution language for the trustee procedures. The written-consent register publishes a precise entry for the travel policy. Nor do the pages prove that the same number exists twice in the Corporate Secretary's internal books. A web header may contain a copy error, a placeholder may have survived publication, or one projection may not have been updated after a final number was assigned.

It would also be wrong to declare either policy void. Legal effectiveness may depend on approved minutes, signed consents, corporate records and applicable law—not on an analyst choosing the web page that looks most formal. The evidence supports a reporting statement about the public authority interface: an outsider cannot use Resolution 2026-13 as a unique join key today.

That distinction matters because restraint is part of accountability. An exaggerated charge would ask the public record to prove facts it does not contain. A silent shrug would ignore the facts it plainly does contain. The accurate middle is that Internet Society can probably resolve the issue with a small correction, but only Internet Society can identify the source record that authorizes that correction.

Resolution numbers are infrastructure

A resolution number looks like clerical metadata until another document depends on it. Then it becomes infrastructure.

Policy owners use the number to explain why the current text exists. Trustees use it to trace prior decisions. Candidates and voting constituencies use it to understand the rules of an election. Auditors and counsel use it to join an operative document to minutes or written consents. Future Boards use it to amend or reaffirm the right instrument. Archives, search engines and third parties copy it into their own records.

That is why uniqueness matters. The number is not the source of the Board's authority; the Board's valid act is. But the number is the public pointer to that act. A pointer that resolves to two targets cannot demonstrate which text was approved, when it became effective or which later decision superseded it. The administrative record has not seized authority from the Board. It has failed to project the Board's authority unambiguously.

The two affected policies show the range of reliance. Trustee-selection procedures shape who may stand, who may vote and how an election is certified. The travel policy controls reimbursement, advance approval and receipts. One concerns the route into the governing body; the other concerns how members of that body spend organizational funds. Both deserve a provenance chain more durable than a page header.

The approval routes must stay distinct

The June and July records also illustrate why a canonical ledger must preserve approval method.

The trustee-procedure amendments appear on the page for a Board meeting held on 10 June. The travel-policy amendments appear in the separate written-consent register and were approved on 6 July. Internet Society's bylaws permit action without a meeting by unanimous written consent of all current trustees. Its electronic-vote procedure says every trustee must consent, the effective date must be set out in the proposal, the result must be announced to the Board list and the resolution must be noted in the minutes of the next Board meeting.

Neither method is a weaker imitation of the other. They are different authorized routes with different evidence. A meeting decision is joined to notice, quorum, agenda, vote and approved minutes. A written consent is joined to the motion, every trustee's consent, the announced result, effective date and later notation in meeting minutes. Compressing both into an unqualified number throws away information needed to audit the act.

The public index already recognizes the distinction by listing meeting resolutions separately from e-votes. The missing piece is one identity layer above both lists. That layer should make the identifier unique across approval routes and should show, rather than obscure, which route produced the decision.

The strongest defence is ordinary publishing error

The most plausible benign explanation is simple: a number was copied to the trustee-procedure page that belongs to the travel policy, while the meeting page retained its pre-publication placeholder. Complex sites contain hand-edited headers, structured registers, cached indexes and policy pages owned by different teams. Drift happens.

That defence should be taken seriously. It also makes the correction straightforward. Internet Society can name the final resolution identifier for the trustee procedures, correct the affected header, publish or link the Meeting No. 195 resolution record, and add a dated correction note. If the meeting action never received a separate final number, it can say so and explain the authoritative citation. If 2026-13 was reassigned, it can preserve that history.

What would not solve the governance problem is a silent text edit. A future visitor would see one clean number, but anyone who saved the earlier page would not know whether the Board record changed or only the web projection did. Search indexes and third-party citations would preserve the former value without an authoritative bridge. A correction should add state, not erase the fact that the public record once forked.

A canonical Board-decision identity receipt

The remedy can be compact. Each decision should have a public, machine-readable receipt generated from the authoritative corporate record.

At minimum, the receipt should contain a globally unique resolution identifier; the exact action title and decision text; whether approval occurred at a meeting or by unanimous written consent; approval, effective and publication dates; the meeting, minutes or e-vote source; and the version or cryptographic hash of every approved policy artifact. It should also list the public pages that project the identifier, the decisions it amends or supersedes, and the decisions that later amend it.

Corrections need their own fields: previous public value, reason class, approving role, correction time, affected projections and whether readers relying on an earlier version were notified. The earlier value should remain in an append-only history. A human-readable page can explain the record, while a structured endpoint lets policy indexes and website headers derive from the same source rather than being typed again.

This is not a proposal for a website database to decide what the Board did. The authority remains with the Board and its valid corporate records. The receipt is a narrow public mirror of that authority. It lets a reader move from policy to decision and back again without asking which of two identical keys to trust.

What a correction would prove

A complete correction would answer four questions.

First, what final identifier belongs to the 10 June approval of the Procedures for Selecting Trustees? Second, what authoritative artifact contains that decision—approved minutes, a resolution page or another corporate record? Third, does Resolution 2026-13 continue to identify only the 6 July travel-policy consent? Fourth, when and why did the conflicting public header appear, and which surfaces were corrected?

Those answers would not require public disclosure of trustee signatures, privileged advice or internal drafting exchanges. They would not reopen either policy on the merits. They would simply restore the one property an identifier must have: one stable route to one identifiable act.

Until then, the safest reading is also the least dramatic. The travel-policy assignment has the stronger public chain because the policy header and formal e-vote register agree on title, number and date. The trustee-selection amendments have public decision language but an unresolved placeholder at the meeting source. That is evidence for triage, not authority to rewrite the record from outside.

Sources

  1. Internet Society — Procedures for Selecting Trustees
  2. Internet Society — Resolutions Approved via Unanimous Written Consent
  3. Internet Society — Travel Policy for Trustees
  4. Internet Society — Board Meeting No. 195
  5. Internet Society — List of All Board Resolutions
  6. Internet Society — Procedure for Conducting Electronic Votes
  7. Internet Society — Amended and Restated By-Laws
  8. Lu Heng — The Policy Mirror
  9. Lu Heng — On When the Bookkeeper Auditions for Olympus
  10. Internet Society — Resolution index by type