add more territories

@@ -5,4 +5,12 @@ A 0–100 score of personal cannabis freedom for every country — what you may
 **Live site:** https://cannabisfreedom.fyi/
 
+## What's on the site
+
+- **The index:** 214 countries, territories and dependencies scored, with state-by-state scores for the United States, Canada and Australia, a world map that can be coloured by any segment or context section (seeds and clones, clubs, driving, visitors, cannabis use), rankings and side-by-side comparison.
+- **History:** the laws, rulings and treaties that moved each score since 1900, with a yearly score for every place rebuilt from them. More than 500 events have a full article.
+- **Culture:** a cultural atlas of words for the plant, traditions and the people who changed the law.
+- **State of Cannabis:** a yearly, pseudonymous community survey of how people use, grow and think about cannabis.
+- **Sources:** every source the index cites, including the books it draws on.
+
 ## How the score works
 
@@ -40,8 +48,27 @@ Built with [Astro](https://astro.build). Pushing to `main` deploys to GitHub Pag
 | `segments.yaml` | The six segments, their weights and level descriptions |
 | `changes.yaml` | The History page: laws and events that moved a country's score, with the scores before, plus treaties; the home page shows the latest four |
+| `articles/` | Longer articles on single history entries, one Markdown file each, named after the entry |
+| `culture/` | The cultural atlas: `language/`, `traditions/` and `people/`, one Markdown file per entry |
 | `prices.yaml` | Sourced flower prices by country, date and market (street or medical) |
+| `survey/` | The State of Cannabis questionnaire (`2026.yaml`) and its published, counted results |
 | `src/` | The website: pages, components, styles |
 | `docs/DATA-MAINTENANCE.md` | How the country profiles and the history are checked and extended, and what is still open |
-| `scripts/` | Helpers for that work: apply a batch of history edits, list what still needs a source |
+| `docs/CULTURE.md` | What belongs in each part of the cultural atlas and how to write an entry |
+| `docs/BOOKS.md`, `docs/books.yaml` | How the research books are read and cited, and the catalogue of what has been read; the books themselves sit in `books/`, which is not in Git |
+| `docs/research/` | Dated research notes behind new entries and changes |
+| `source-archive.json` | Internet Archive copies of the cited sources |
+| `scripts/` | Helpers for that work: see below |
+
+Helper scripts, run from the project root:
+
+| Command | What |
+|---|---|
+| `node scripts/data-status.cjs` | Where the data stands: unchecked history entries, estimates, thin profiles, sections not yet researched |
+| `node scripts/history-batch.cjs` | Apply a batch of history edits from a JSON file |
+| `npm run articles:backlog` | Which history entries most deserve an article next |
+| `npm run books:status` / `books:check` | Reading status of the research books; `books:report` rewrites `docs/BOOK-RESEARCH.md` |
+| `node scripts/archive-sources.cjs` | Save new sources to the Internet Archive and update `source-archive.json` |
+| `node scripts/book-covers.cjs` | Fetch covers for the books listed on the Sources page |
+| `node scripts/survey-collect.cjs` | Decrypt and count submitted survey answers into `survey/results-<year>.json` (needs the project's secret key) |
 
 The build checks every data file against a schema (`src/content.config.ts`) and fails with a clear message if, say, a score is out of range or a segment cites a source that isn't defined.
@@ -69,4 +96,6 @@ Please include a source link or a complete bibliographic citation. Or just [open
 ### Contributing flower prices
 
+Prices are still incomplete, so the site hides them unless it is built with `PUBLIC_SHOW_PRICES=true` (see `.env.example`).
+
 Add sourced per-gram observations to `prices.yaml`. Only two markets are used: `street` for
 illicit or non-prescribed flower, and `medical` for prescribed flower. Keep the original currency,
@@ -80,8 +109,14 @@ example, source-quality rules, currency conversion and validation checklist.
 Scores are editorial estimates, not legal advice. Check current local law before travelling with, buying or growing cannabis.
 
+## Privacy
+
+The site uses [Plausible](https://plausible.io/) for visitor statistics: no cookies and no personal data. Survey answers are signed with a key made in the respondent's browser and encrypted so that only the respondent and the project can read them; only counts are published.
+
 ## Data for research
 
 - `/data/panel.csv`: a score for every place in every year since 1900, with the six segment scores, rebuilt from the history. The `basis` column separates recorded years from assumptions.
 - `/data/history.csv`, `/data/scores.csv`: every history entry with its sources, and today's scores with ISO numeric codes for joining.
+- `/data/prices.csv` and `/data/survey.json`: the sourced flower prices, and the counted survey results.
+- Every file is also available as JSON (`/data/panel.json`, `/data/scores.json`, …).
 - `source-archive.json`: copies of the sources in the Internet Archive, kept up to date with `node scripts/archive-sources.cjs`.
 

@@ -368,4 +368,19 @@ still require a URL.
 - **`country`** may also be a disputed territory (`xa`, `xn`, `xs`, `xt`). Its entries show on its
   page and in the history list, but its scored changes never move the world average.
+- **Dependencies with laws of their own** have their own profiles where they matter enough:
+  Jersey (`je`), Guernsey (`gg`), the Isle of Man (`im`), Bermuda (`bm`), the Cayman Islands
+  (`ky`), Gibraltar (`gi`) and the British Virgin Islands (`vg`) under the UK, and Greenland (`gl`)
+  and the Faroe Islands (`fo`) under Denmark, French Polynesia (`pf`) under France, and Puerto Rico
+  (`pr`), Guam (`gu`), the Northern Mariana Islands (`mp`) and the US Virgin Islands (`vi`) under
+  the United States. A profile marks this with
+  `dependency: {of: gb, kind: "Crown Dependency", note: …}`;
+  like a disputed territory it is scored and drawn on the map but takes no rank and stays out of
+  the averages. Their history entries use their own code.
+- **A place without a profile of its own** goes under the country that answers for it
+  internationally: the Falklands and the other British Overseas Territories under `gb`; French
+  overseas territories such as New Caledonia under `fr`; American Samoa under `us`. Leaving `country` out files the entry as
+  International, which is only for treaties, UN decisions and other events that belong to no one
+  country. Such entries are normally context entries: they must not move the profile's scores.
+  When a place has gathered enough to deserve a profile, give it one as above and move its entries.
 - **`checked: true`** means: the date and the event were compared with a source that was read.
   A search-result summary is not a read source (see "What went wrong").
@@ -924,4 +939,17 @@ drugs. No articles were changed; see the research note for source limits.
 ## Recent law-news scan (12 October 2026)
 
+A [17 October follow-up](research/law-news-2026-10-17.md) adds four official-source
+context entries for Virginia's hemp cap and retail-rule draft, Jersey's planned
+prescriber oversight, and Bermuda's medical/hemp timetable. Virginia's profile
+now gives the scheduled retail start date and hemp restriction. Announcements
+and drafts remain distinguished from operating rules; scores were preserved.
+The build passed with 3,102 pages.
+
+The Canada follow-up in the same note adds Nova Scotia's June fine increases,
+the April veterans' reimbursement reduction, and the earlier commercial-edible
+classification guidance raised by a cooking-policy lead. Canada and Nova Scotia
+profiles were updated together. The guidance is dated to its verified September
+2024 announcement, not presented as a new 2026 cooking ban. Scores were preserved.
+
 Added five primary-sourced context events: Ireland's October medical-access
 review timetable, Canada's October regulatory commencement and September

@@ -0,0 +1,87 @@
+# Cannabis policy scan — 17 October 2026
+
+Four gaps were added to `changes.yaml` after reading official sources. The scan
+looked for recent developments and checked existing coverage before adding records;
+it is not an exhaustive worldwide audit. The repository's database is its YAML
+data, rather than a separate SQL store.
+
+| Event date | Jurisdiction | Addition and legal stage |
+| --- | --- | --- |
+| 12 October 2026 | Jersey | Ministerial answer plans legislation overseeing medical cannabis prescribers, expected in early 2027; not enacted rules. |
+| September 2026 | Virginia, US | Regulator timeline records initial retail-rule draft; sales scheduled for 1 July 2027. |
+| 15 August 2026 | Virginia, US | Regulator confirms removal of the CBD-ratio exception to its 2 mg total-THC hemp-package cap. |
+| 23 July 2026 | Bermuda | Minister sets Q4 target for initial medical/hemp reforms; wider adult-use work takes longer. |
+
+## Sources read and limits
+
+- [Jersey WQ.201/2026, original one-page PDF](https://www.statesassembly.je/getattachment/0ca87632-b6c1-4e9c-9d99-80e90eaef827/WQ-201-2026.pdf?ext=.pdf&lang=en-GB).
+  The answer date, rather than the question's 5 October submission, is used.
+  It does not establish passage, final prescribing requirements or commencement.
+- [Virginia regulator timeline](https://cca.virginia.gov/retailmarijuanamarket).
+  September draft release is recorded only to month precision. Later milestones
+  remain scheduled. The linked SharePoint draft required authentication and was
+  not read; no individual draft provisions are asserted.
+- [Virginia regulator implementation notice](https://cca.virginia.gov/news/virginias-new-marijuana-and-hemp-laws-what-you-need-to-know).
+  Uses the stated August effective date, not the July publication date. The HB30
+  legislative link returned a JavaScript shell; this addition relies on the
+  regulator's express implementation guidance, not a claimed full statute review.
+- [Bermuda Senate statement](https://www.gov.bm/news/advancing-public-safety-across-the-ministry-of-national-security).
+  The Q4 target is an announcement, not evidence that implementation has occurred.
+  No subsequent commencement was established in this scan.
+
+Jersey and Bermuda have no standalone country profile here. Their entries have
+no `country`, following the existing Cayman convention; neither is assigned to
+the UK's domestic score. Virginia's entries belong to the US context timeline,
+with its state profile updated together. All four entries omit `before`.
+Scores remain unchanged: the new material establishes a hemp-market restriction,
+draft rules and planned reforms, not a change to the scored adult flower rights.
+
+Already-covered Canada, Ireland, NSW and Cayman developments were not duplicated.
+A New Zealand controlled-drug administration amendment was screened but not added:
+the material retrieved did not establish a sufficiently specific cannabis change.
+Virginia's September medical-dispensary opening was read as an access lead but
+not added as a separate policy change.
+
+## Validation
+
+`npm run build` passed (3,102 pages). History IDs are unique, all four additions
+are checked context entries without score baselines, and Virginia's six scores
+match their pre-edit values. The US score-chain check reports zero flagged steps.
+Generated HTML contains all four new history pages and their source links;
+Virginia's page contains the new sales date, hemp cap and regulator citation.
+`git diff --check` passed. Pre-existing README and source-archive edits were
+preserved. No articles were changed.
+
+## Canada follow-up: commercial edibles and two 2026 changes
+
+The user recalled a cooking-related story and narrowed it to commercial edibles.
+The closest verified match is Health Canada's ingestible-product classification
+guidance. This remains a possible match, not a confirmed identification of the
+story. No new home-cooking ban was established.
+
+Three additional records were saved:
+
+| Event date | Record | Source and dating |
+| --- | --- | --- |
+| 3 September 2024 | Expanded public classification guidance | [Health Canada industry announcement](https://www.canada.ca/en/health-canada/services/drugs-medication/cannabis/industry-licensees-applicants/updates-cannabis-industrial-hemp.html) dates inclusion of the policy; it was shared with licence holders in December 2023. The [current guidance](https://www.canada.ca/en/health-canada/services/cannabis-regulations-licensed-producers/classification-edible-cannabis.html) carries 18 November 2025 page metadata, which is not treated as first commencement. |
+| 25 June 2026 | Nova Scotia fine increases and new offences | [Provincial announcement](https://news.novascotia.ca/en/2026/06/25/higher-fines-new-offences-support-safe-legal-cannabis-market) explicitly confirms proclamation that day. Its linked statute PDF returned a rejection page; penalty details rely on the government's implementation notice. |
+| 1 April 2026 | Veterans Affairs reimbursement reduction | [Official FAQ](https://www.veterans.gc.ca/en/contact-us/help/faqs/cannabis-medical-purposes-revised-reimbursement-policy), especially questions 8 and 13–15, separates reimbursement from retail prices, authorised quantities and eligible product forms. |
+
+Canada's product context now explains the food-mixing classification, and its
+medical context acknowledges the separate veterans' benefit. Nova Scotia's
+enforcement section now records the provincial fines. All three additions omit
+`before`: classification guidance clarifies existing product rules, the benefit
+does not alter adult-use rights, and the provincial entry follows the federation
+context convention. Existing national and provincial scores were preserved.
+
+The commercial guidance is recorded under its verified earlier announcement,
+not presented as a newly adopted 2026 rule. The October industry notice about
+administrative monetary penalties links to guidance carrying July metadata;
+no fresh penalty law was inferred from that notice. The previously recorded
+March 2025 streamlining reform was not duplicated.
+
+Follow-up validation: `npm run build` passed (3,105 pages). History IDs are unique,
+all three additions are checked context entries, and Canada/Nova Scotia scores
+match their pre-edit values. Canada's score-chain check reports zero flagged
+steps. Generated HTML contains the three new history pages and the updated
+profiles with citations. `git diff --check` passed. No articles were changed.

Not shown: a picture, a generated file, or past the length shown per change.

@@ -91,6 +91,6 @@ const culture = sub ? [] : cultureOf(c);
         {sub
           ? <p class="sub">#{sub.rank} of {group.units.length} in {c.name}<br>would rank #{sub.worldRank} among countries</p>
-          : c.disputed
-            ? <p class="sub">Not ranked: disputed territory<br>would rank #{c.worldRank} among countries</p>
+          : c.territory
+            ? <p class="sub">Not ranked: {c.territory.label}<br>would rank #{c.worldRank} among countries</p>
             : <p class="sub">#{c.rank} of {countries.length} worldwide<br>#{c.regionRank} of {c.regionCount} in {c.region}</p>}
       </div>
@@ -113,6 +113,6 @@ const culture = sub ? [] : cultureOf(c);
       </div>
     )}
-    {c.disputed && !sub && (
-      <p class="c-disputed"><strong>Disputed territory.</strong> {c.disputed.note} See also <a href={`${base}country/${c.disputed.country.slug}/`} data-country={c.disputed.country.id}>{c.disputed.country.flag} {c.disputed.country.name}</a>.</p>
+    {c.territory && !sub && (
+      <p class="c-disputed"><strong>{c.territory.kind}.</strong> {c.territory.note} See also <a href={`${base}country/${c.territory.country.slug}/`} data-country={c.territory.country.id}>{c.territory.country.flag} {c.territory.country.name}</a>.</p>
     )}
     <h2 class="d-h">In short <span><a href={`${base}can-i/`}>all countries →</a></span></h2>

@@ -367,5 +367,5 @@ const columnGroups = [["Shown by default", BASE_COLUMNS], ...["Score segments",
                   {...Object.fromEntries(EXTRA_COLUMNS.map((col) => [`data-${col.key}`, col.sort(c)]))}>
                 <td class="num">{c.rank ?? "–"}</td>
-                <td><a class="cname" href={`${base}country/${c.slug}/`}><span class="flag">{c.flag}</span>{c.name}{trendOf(c) && ((t) => <span class={`trend ${t.delta > 0 ? "up" : "down"}`} title={`${t.delta > 0 ? "+" : "−"}${Math.abs(t.delta)} ${Math.abs(t.delta) === 1 ? "point" : "points"} since ${t.since}. Latest change: ${t.last.title} (${changeDate(t.last)})`}>{t.delta > 0 ? "▲" : "▼"}{Math.abs(t.delta)}</span>)(trendOf(c))}</a>{c.disputed && <span class="tag-disputed" title={c.disputed.note}>disputed</span>}</td>
+                <td><a class="cname" href={`${base}country/${c.slug}/`}><span class="flag">{c.flag}</span>{c.name}{trendOf(c) && ((t) => <span class={`trend ${t.delta > 0 ? "up" : "down"}`} title={`${t.delta > 0 ? "+" : "−"}${Math.abs(t.delta)} ${Math.abs(t.delta) === 1 ? "point" : "points"} since ${t.since}. Latest change: ${t.last.title} (${changeDate(t.last)})`}>{t.delta > 0 ? "▲" : "▼"}{Math.abs(t.delta)}</span>)(trendOf(c))}</a>{c.territory && <span class="tag-disputed" title={`${c.territory.kind}. ${c.territory.note}`}>{c.territory.tag}</span>}</td>
                 <td class="hide-sm" data-col="region">{c.region}</td>
                 <td class="num-col hide-md" data-col="users" title={useNote(c)}>{c.use.estimate && <span class="approx">~</span>}{compact(c.use.users)}</td>
@@ -503,5 +503,5 @@ const columnGroups = [["Shown by default", BASE_COLUMNS], ...["Score segments",
     const byNum = new Map(COUNTRIES.filter((c) => c.num != null).map((c) => [c.num, c]));
     // Map features that inherit another entry's score (atlas name or numeric id → country id).
-    const ALIASES = { Kosovo: "XK", Somaliland: "XS", "N. Cyprus": "XN", 304: "DK" };
+    const ALIASES = { Kosovo: "XK", Somaliland: "XS", "N. Cyprus": "XN" };
 
     // ── Map ───────────────────────────────────────────────

@@ -59,4 +59,9 @@ const countries = defineCollection({
       // name of the country it is internationally regarded as part of.
       disputed: z.object({ part_of: z.string().regex(/^[a-z]{2}$/), note: z.string().min(1) }).strict().optional(),
+      // A dependency with its own laws, outside the country that answers for it internationally
+      // (a Crown Dependency or an overseas territory). Shown and scored like a disputed territory:
+      // a page, no rank, no part in the averages. `of` is that country's file name; `kind` is how
+      // the place is described ("Crown Dependency", "British Overseas Territory").
+      dependency: z.object({ of: z.string().regex(/^[a-z]{2}$/), kind: z.string().min(1), note: z.string().min(1) }).strict().optional(),
       flag: z.string().optional(), // an emoji to show when the two-letter code has no flag of its own
       // Population by year: a snapshot every ten years from 1900 (Our World in Data) and the

@@ -65,4 +65,9 @@ const everyPlace = (await getCollection("countries"))
       flag: data.flag ?? flagOf(id),
       disputed: data.disputed ?? null, // { part_of, note } for a territory that is not a recognised state
+      dependency: data.dependency ?? null, // { of, kind, note } for a Crown Dependency or overseas territory
+      // what an unranked place is, for both kinds: { kind (to start a sentence), label (inside one),
+      // tag, note, of (file name), country (set below) }
+      territory: data.disputed ? { kind: "Disputed territory", label: "disputed territory", tag: "disputed", note: data.disputed.note, of: data.disputed.part_of }
+        : data.dependency ? { kind: data.dependency.kind, label: data.dependency.kind, tag: "dependency", note: data.dependency.note, of: data.dependency.of } : null,
       num: data.iso_numeric ?? null,
       populationByYear: Object.entries(data.population).map(([y, v]) => [+y, v]).sort((a, b) => a[0] - b[0]),
@@ -88,8 +93,9 @@ const everyPlace = (await getCollection("countries"))
   .sort((a, b) => b.raw - a.raw || a.name.localeCompare(b.name));
 
-// The index ranks recognised countries. Disputed territories are scored the same way and shown
-// alongside, but they take no rank and do not count towards any average.
-export const countries = everyPlace.filter((c) => !c.disputed);
-export const territories = everyPlace.filter((c) => c.disputed);
+// The index ranks recognised countries. Disputed territories and dependencies with laws of their
+// own are scored the same way and shown alongside, but they take no rank and do not count towards
+// any average.
+export const countries = everyPlace.filter((c) => !c.territory);
+export const territories = everyPlace.filter((c) => c.territory);
 export const places = everyPlace; // both, by score: for pages, the map and the table
 countries.forEach((c, i) => (c.rank = i + 1));
@@ -97,6 +103,7 @@ territories.forEach((t) => {
   t.rank = null;
   t.worldRank = 1 + countries.filter((c) => c.raw > t.raw).length; // where it would sit among countries
-  t.disputed.country = countries.find((c) => c.slug === t.disputed.part_of);
-  if (!t.disputed.country) throw new Error(`countries/${t.slug}.yaml: disputed.part_of "${t.disputed.part_of}" is not a country`);
+  t.territory.country = countries.find((c) => c.slug === t.territory.of);
+  if (!t.territory.country) throw new Error(`countries/${t.slug}.yaml: "${t.territory.of}" (disputed.part_of or dependency.of) is not a country`);
+  if (t.disputed) t.disputed.country = t.territory.country;
 });
 countries.forEach((c) => {
@@ -240,5 +247,5 @@ function averageOverTime(weight) {
     return sum / total;
   };
-  const events = changes.filter((ch) => ch.score && !ch.estimate && !ch.country.disputed).map((ch) => ({ change: ch, date: ch.date, t: dateTime(ch.date) }));
+  const events = changes.filter((ch) => ch.score && !ch.estimate && !ch.country.territory).map((ch) => ({ change: ch, date: ch.date, t: dateTime(ch.date) }));
   for (let year = CHART_FROM; year <= +DATA_SNAPSHOT; year++) events.push({ change: null, date: `${year}-01-01`, t: Date.UTC(year, 0, 1) });
   events.sort((a, b) => a.t - b.t);

@@ -7,5 +7,5 @@ import { priceObservations } from "./prices.js";
 export const SITE = "https://cannabisfreedom.fyi";
 
-// One row per country or disputed territory.
+// One row per country, disputed territory or dependency.
 export const scoreRows = places.map((c) => ({
   id: c.id,
@@ -14,4 +14,5 @@ export const scoreRows = places.map((c) => ({
   region: c.region,
   disputed: Boolean(c.disputed),
+  dependency_of: c.dependency?.of ?? "", // a Crown Dependency or overseas territory: the country that answers for it
   rank: c.rank ?? "",
   score: c.score,
@@ -96,4 +97,5 @@ export const panelRows = places.flatMap((c) => {
       basis,
       disputed: Boolean(c.disputed),
+      dependency_of: c.dependency?.of ?? "",
     });
   }

@@ -11,5 +11,5 @@ const scored = changes.filter((ch) => ch.score && !ch.estimate);
 const intro = (root) => `# Cannabis Freedom Index
 
-> A score from 0 to 100 for every country, for how free a person is to possess, grow, share and use cannabis. ${countries.length} countries and ${territories.length} disputed territories are scored, each with a written explanation, sources and a dated history of how its law changed. ${DATA_SNAPSHOT} edition, last updated ${formatDate(lastUpdated)}.
+> A score from 0 to 100 for every country, for how free a person is to possess, grow, share and use cannabis. ${countries.length} countries and ${territories.length} disputed territories and dependencies are scored, each with a written explanation, sources and a dated history of how its law changed. ${DATA_SNAPSHOT} edition, last updated ${formatDate(lastUpdated)}.
 
 Important when using this data:
@@ -65,5 +65,5 @@ export function llmsFull(root) {
   const profile = (c) => `### ${c.name}
 
-Score ${c.score}/100 (${tierOf(c.raw).name})${c.disputed ? ", disputed territory, not ranked" : `, rank ${c.rank} of ${countries.length}`}. Recreational: ${REC_LABEL[c.rec]}. Medical: ${MED_LABEL[c.med]}. Reviewed ${c.details.reviewed}. Page: ${root}country/${c.slug}/
+Score ${c.score}/100 (${tierOf(c.raw).name})${c.territory ? `, ${c.territory.label}, not ranked` : `, rank ${c.rank} of ${countries.length}`}. Recreational: ${REC_LABEL[c.rec]}. Medical: ${MED_LABEL[c.med]}. Reviewed ${c.details.reviewed}. Page: ${root}country/${c.slug}/
 
 ${one(c.note)}

@@ -79,5 +79,5 @@ export const research = (await getCollection("countries")).map(({ id: slug, data
   const gaps = years.slice(1).map((y, i) => y - years[i]);
   return {
-    slug, name: data.name, flag: data.flag ?? flagOf(slug), region: data.region, disputed: Boolean(data.disputed),
+    slug, name: data.name, flag: data.flag ?? flagOf(slug), region: data.region, disputed: Boolean(data.disputed || data.dependency),
     population: data.population[Object.keys(data.population).sort().at(-1)],
     pct, profile: Math.floor(share("profile") * 100), history: Math.floor(share("history") * 100), signed,

@@ -17,5 +17,5 @@ const byId = new Map(places.map((c) => [c.id, c]));
 const byNum = new Map(places.filter((c) => c.num != null).map((c) => [c.num, c]));
 // atlas shapes that take another entry's score, as on the home page
-const ALIASES = { Kosovo: "XK", Somaliland: "XS", "N. Cyprus": "XN", 304: "DK" };
+const ALIASES = { Kosovo: "XK", Somaliland: "XS", "N. Cyprus": "XN" };
 const ANTARCTICA = "010";
 

@@ -6,6 +6,6 @@ import { SHOW_PRICES } from "../lib/config.js";
 import { PLANTING_STATUS, STATUS_SECTIONS, PLANTING_TRADE } from "../lib/segment-ids.js";
 
-const light = places.map(({ id, slug, name, flag, region, raw, score, rank, s, rec, med }) =>
-  ({ id, slug, name, flag, region, raw, score, rank, s, rec, med, prices: SHOW_PRICES ? flatLatestPricesFor(id) : {} }));
+const light = places.map(({ id, slug, name, flag, region, raw, score, rank, s, rec, med, territory }) =>
+  ({ id, slug, name, flag, region, raw, score, rank, s, rec, med, kind: territory?.label, prices: SHOW_PRICES ? flatLatestPricesFor(id) : {} }));
 const meta = {
   segments: SEGMENTS.map((s) => ({ name: s.name, weight: s.weight })),
@@ -149,5 +149,5 @@ const presets = [
           <div class="bar"><i style="width:${c.raw}%;background:${tierColor(c.raw)}"></i></div>
           <div class="cmp-sub">${tierOf(c.raw).name}</div>`), "cmp-index"),
-        row("World rank", cs.map((c) => (c.rank ? `#${c.rank} <span class="cmp-sub">of ${META.total}</span>` : `<span class="cmp-sub">not ranked: disputed territory</span>`))),
+        row("World rank", cs.map((c) => (c.rank ? `#${c.rank} <span class="cmp-sub">of ${META.total}</span>` : `<span class="cmp-sub">not ranked: ${c.kind}</span>`))),
         row("Recreational", cs.map((c) => `<span class="chip s-${c.rec}">${META.rec[c.rec]}</span>`)),
         row("Medical", cs.map((c) => `<span class="chip s-${c.med}">${META.med[c.med]}</span>`)),

@@ -10,5 +10,5 @@ const base = import.meta.env.BASE_URL;
 const repo = "https://github.com/burningtree/cannabis-freedom-index";
 const files = [
-  { name: "scores", title: "Scores", rows: scoreRows, about: "One row per country or disputed territory: index score, the six segment scores, rank, tier, legal status, seed and clone status where researched, population and estimated cannabis users." },
+  { name: "scores", title: "Scores", rows: scoreRows, about: "One row per country, disputed territory or dependency (Crown Dependencies and overseas territories, with dependency_of naming the country that answers for them): index score, the six segment scores, rank, tier, legal status, seed and clone status where researched, population and estimated cannabis users." },
   { name: "panel", title: "Scores by year", rows: panelRows, about: `One row per place and year from ${PANEL_FROM} to ${DATA_SNAPSHOT}: the index score and the six segment scores as they stood at the end of each year, rebuilt from the history. Made for joining with other country-year data.` },
   { name: "history", title: "History", rows: historyRows, about: "One row per history entry: date, country, category, title, description, the score before and after, how far it has been checked, and its sources." },
@@ -51,5 +51,5 @@ const files = [
         <li><code>baseline</code>: years before the place's first recorded change. The scores are its starting point, usually no cannabis-specific rules at all.</li>
       </ul>
-      <p>Scores before each change are editorial estimates in every case. Places are identified by a two-letter code (<code>id</code>) and by the ISO 3166-1 numeric code (<code>iso_numeric</code>), which is empty for disputed territories. Borders are today's: a country that was part of another state carries the law that applied on its territory.</p>
+      <p>Scores before each change are editorial estimates in every case. Places are identified by a two-letter code (<code>id</code>) and by the ISO 3166-1 numeric code (<code>iso_numeric</code>), which is empty for disputed territories. Disputed territories and dependencies are scored but not ranked, and are left out of every average. Borders are today's: a country that was part of another state carries the law that applied on its territory.</p>
       <h2>How to cite</h2>
       <p>Cannabis Freedom Index, {DATA_SNAPSHOT} edition, cannabisfreedom.fyi, retrieved on the date of download. The data changes as laws do, so keep the file you used; the <a href={`${base}changelog/`}>changelog</a> lists every score change since.</p>

@@ -147,5 +147,5 @@ const related = relatedTo(ch, nearby);
           <p class="d-num"><b>{c.score}</b><span>/100</span></p>
           <div class="d-meter" aria-hidden="true"><i style={`width:${c.raw}%`}></i></div>
-          <p class="ev-tier">{tierOf(c.raw).name} <small>{c.disputed ? "Not ranked: disputed territory" : `#${c.rank} of ${countries.length}`}</small></p>
+          <p class="ev-tier">{tierOf(c.raw).name} <small>{c.territory ? `Not ranked: ${c.territory.label}` : `#${c.rank} of ${countries.length}`}</small></p>
           <div class="d-chips">
             <span class={`chip s-${c.rec}`} title={REC_HELP[c.rec]}>Recreational: {REC_LABEL[c.rec]}</span>

@@ -19,5 +19,5 @@ async function build(kind) {
     return Object.fromEntries(places.map((c) => [`country/${c.slug}/`, {
       f: c.flag, n: c.name, r: c.region, s: c.score, c: tierColor(c.raw), t: tierOf(c.raw).name,
-      k: c.disputed ? null : c.rank, of: places.filter((x) => !x.disputed).length,
+      k: c.territory ? null : c.rank, of: places.filter((x) => !x.territory).length,
       rec: [REC_LABEL[c.rec], c.rec], med: [MED_LABEL[c.med], c.med], x: cut(c.note, 230), h: entries.get(c.slug) ?? 0,
     }]));