Skip to content

Add weight-chambers prototype of the A1 score triangle - #142

Merged
minoreality merged 3 commits into
mainfrom
weight-chambers-prototype
Sep 25, 2026
Merged

minoreality merged 3 commits into
mainfrom
weight-chambers-prototype

Conversation

@minoreality

@minoreality minoreality commented Sep 25, 2026 •

Copy link
Copy Markdown
Owner

Summary

  • Add prototypes/weight-chambers/, a standalone page that moves the primary weights over the integer lattice w_G + w_R + w_B = 168 and shows the order the additive score σ(S) = Σ_{c∈S} w_c gives the eight states. Six walls cut the weight triangle into twelve chambers; the three A1 conditions (w_B > 0, w_R > w_B, w_G > w_R + w_B) hold exactly in the chamber ordered K<B<R<M<G<C<Y<W. A1 and Theorem 1 follow the numbering Derive GRB ranks from the three score conditions #144 brought to docs/algebraic-color-model.md.
  • model.test.ts enumerates all 14,365 lattice points. It checks the funnel of 48 inclusion orders → 12 reversed by complement (the twelve chambers, which are also the de Finetti orders) → 6 with additive rank → 1, and that each wall swaps exactly the pairs that tie on it. Expectations are rebuilt from subsets of {G,R,B}, independently of the model's bit arithmetic. Like hue-euler's, the file runs under npm test, so the CI checks job covers it.
  • Register main.ts, playwright.config.ts and browser.check.ts as knip entries. The page is not a build input and nothing in the app links to it; dist/ contains no prototypes/.
  • The page and its README describe the order of an additive score, not perceived or photometric brightness, and use no photometric coefficients (§Evidence Boundary).
  • Second commit: the reduced ratio next to the weights was hidden on the whole w_G = 0 edge (121 of the 5,077 reducible lattice points), because ratio[0] === weights[0] stood in for "no common factor". It is now hidden only when the reduction leaves every weight unchanged, and a browser check covers all three edges.

main is merged in (#139, #141, #143, #144). The only conflict was the knip entry list, where #144 added the De Morgan prototypes at the same place; both sets are kept.

Checks

On 5cfa60c. The first five ran inside npm run verify (exit 0, 96 s).

  • Ran npm run format:check (it does not cover prototypes/; npx prettier --check prototypes/weight-chambers is clean as well)
  • Ran npm run lint (it does not cover prototypes/; npx eslint prototypes/weight-chambers is clean as well)
  • Ran npm run deadcode
  • Ran npm run typecheck:all, plus npx tsc --noEmit -p prototypes/weight-chambers/tsconfig.json
  • Ran npm run build (as typecheck:app and build:bundle)
  • Ran targeted tests if code behavior changed (npm run verify: 113 files, 1,311 tests; npx playwright test -c prototypes/weight-chambers/playwright.config.ts --workers=1: 8 passed — CI does not run these browser checks)

Visual Review

  • Checked affected UI in a browser, or this change has no UI impact — the app is untouched; ?w=0,84,84 now reads 0 : 84 : 84 = 0 : 1 : 1 in the dev server, and the prototype's own checks drive the page in Chromium at 320, 390, 576, 834 and 1186 px, asserting that no readout moves across six weight states and that every SVG label stays inside its figure
  • Updated screenshots or icons if visual assets changed — none changed

Documentation And Licensing

  • Updated relevant docs, or no docs update is needed — the prototype carries its own README; docs/ is unchanged
  • Checked whether changes affect MIT / CC BY 4.0 license boundaries — the prototype is outside the CC BY 4.0 set that docs/LICENSE.md enumerates (the listed research documents and Theory tab content), so no boundary moves
  • Checked whether security reporting or deployment notes are affected — not a build input, so nothing new is deployed

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
minoreality and others added 2 commits September 25, 2026 13:38
On the w_G = 0 edge the readout never showed the reduced ratio: 0,84,84
showed nothing while 84,84,0 showed = 1 : 1 : 0, 121 of the 5,077
reducible lattice points in all. It hid the ratio when ratio[0] ===
weights[0], meant as "no common factor", which holds for any factor once
w_G = 0. Hide it only when the reduction leaves every weight unchanged.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@minoreality
minoreality merged commit fcad51b into main Sep 25, 2026
6 checks passed
@minoreality
minoreality deleted the weight-chambers-prototype branch September 25, 2026 04:47
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant