DECIDE: licensing-concerns.md says "No License Yet"; the repo root carries LICENSE

devdocs/developer/licensing-concerns.md:10 is headed "## Current Position: No License Yet". The repo root carries LICENSE and LICENSE.md, and the README states a per-component split — zlib / 0BSD / CC BY 4.0.

Found by frankD during an audit-only sweep of the unowned devdocs/developer/ tree. Flagged, not fixed, and routed here rather than to a docs edit — correctly.

Why this is yours and not a documentation task

Every other staleness in that sweep is an engineering cost: an agent wastes ten minutes, a gate is described wrongly, a "not implemented" claim is disproved by compiling. This one is not. A page stating the project has no license, while the project has three, is a claim about what other people may legally do with the code — and it is the kind of claim that is read by exactly the sort of person you do not want reading a wrong one.

A launch is precisely when someone reads it.

The fork

The question is not "should the page be updated" — plainly yes. It is what the page should now say, and that is not derivable from the tree:

1. It is superseded — the position was reached, the LICENSE files are the answer, and the page becomes a historical record with a Status (date): superseded line. (Likeliest, but it is your call to confirm the split is final.)
2. It is live and the concerns are open — the LICENSE files are provisional, the concerns the page raises still stand, and it needs updating rather than closing.
3. The split itself needs revisiting before launch — in which case this is real work, not a docs fix.

I cannot tell which from the repository. The LICENSE files exist; whether they represent a settled position or a placeholder is a fact only you hold, and guessing it into a public- adjacent page is exactly the class of guess Track U exists to prevent.

Method note for whoever picks up the surrounding sweep

frankD's grep for the hook-denied commands tripped the hook itself, because the hook scans command text and does not care that the command is inside a search pattern. Fourth false positive of that shape in one session. The tree cannot be swept with the obvious command, and a REFUSED there is an artefact of the search, not a finding.

RESOLVED 2026-08-30 — option 1, superseded. The split is final.

"we decided about the license long ago. i think that was stale info." — owner, 2026-08-30

And "long ago" is checkable: done/task-license-mpl2-rollout carries a "Decision (final)" section dated 2026-07-02. So the page was eight weeks stale, not the six the ticket estimated from file mtimes — the mtime dates the last edit, and the decision it missed is older than that.

How the page survived the rollout, which is the reusable part

That same ticket's step 3 reads "Update README.md License section (currently says 'no license yet')". The README was updated. This page said the same thing and was not on the list. A rollout checklist enumerates the files its author knew about, so the page that most needed changing was the one nobody remembered — and it then sat for eight weeks, linking to LICENSE.md as "the current public notice", i.e. pointing at its own refutation.

The general form, worth having when the next repo-wide claim changes: a checklist covers what its author recalled, so the search that finds the rest has to be a grep for the CLAIM, not a list of files. Grepping no license yet|remains unlicensed|not open source yet|all rights reserved across docs/ devdocs/ README.md after the fix returns only this page and the rollout ticket's own historical line — so the tree is clean now, and that grep is what should have run in July.

What was done

devdocs/developer/licensing-concerns.md rewritten as a superseded record, not deleted:

Not deleted, deliberately. The reasoning behind the zlib/MPL split is worth more than the file it sits in — the runtime is zlib because it is embedded into every binary the compiler produces, so programs built with pxx carry no obligations from the toolchain. Deleting the deliberation would leave the conclusion with no recorded why, and "why not PolyForm?" is a question a reader will actually ask.

One paragraph inside a live section was struck rather than kept: "This is a statement of intent, not a present license grant. The current repository remains unlicensed until a license or written permission says otherwise." A fork under MIT/BSD/GPL no longer needs permission asked first. The paragraphs around it about a fork's responsibility for its own license and claims are unaffected and still the intent.