v0.54 — Code audit bugfixes

• Checked JavaScript syntax, JSON files, HTML asset references, duplicate IDs, and Cloudflare Pages Function imports.
• Fixed the starter-search button so it no longer auto-checks the Privacy/Terms agreement.
• Fixed print packet information capture so field-box searches are included in the printed packet, not only the optional notes box.
• Added a concise “Information searched” results section so families can review the exact names, numbers, places, and record clues before contacting agencies.
• Updated cache-busting to v=054.

v0.48 Universal Search Readability Cleanup

• Renamed public-facing “clue” language to “Information you have” / “Information I have.”
• Moved the user guidance boxes above Universal Search so families read the instructions before searching.
• Kept Universal Search as the single public homepage tool instead of exposing several separate search forms.
• Tightened site/index matching so person or roll-number searches do not show every township/section neighbor as a direct match.
• Improved print CSS so agency request packet text prints fully instead of appearing as scroll boxes.

v0.46 — Unified horizontal family-land search

• Replaced the separate Guided Family-Land Finder homepage section with one Universal Search form.
• Expanded the form to include the questions from the older What do you know?, Official-source lookup clues, Dawes search, and current-index search workflows.
• Moved Unified Results directly below the Universal Search form.
• Reordered unified output so the Built Research Path appears first, followed by summary, local/index matches, official-source leads, prepared links, and record-request packet.
• Added an “I don’t know where to start” button that builds a starter research path when the user has no clue yet.
• Improved built research path cards to explain what to look for and where to search for each record type.
• Added guards so older advanced-tool JavaScript does not fail when the old wizard section is no longer present on the homepage.

v0.45 Unified Search

The homepage now has a one-search engine. Users can enter a name, Dawes roll number, census card number, address, latitude/longitude, township/range/section, county, family story, or legal-notice clue. The site will:

• classify the clue type;
• build a research path;
• search local map, Dawes, allotment, and approved-record data;
• query official source connectors where useful;
• prepare official source links;
• build record-request text for NARA/OHS, county clerk, and BIA/LTRO; and
• print/save the full packet as a branded PDF.

No new D1 migration is required. Deploy the full v0.45 folder with Wrangler, not as a changed-files-only upload.

v0.44 address-to-coordinate lookup

Adds an address resolver for Census/TIGERweb research leads:

• New endpoint: /api/address-resolve?address=...
• Uses the U.S. Census Geocoder first.
• If Census cannot match the address, uses OpenStreetMap Nominatim as a low-volume fallback to produce latitude/longitude.
• The Census geography lookup can now use resolved coordinates to request Census geoLookup and TIGERweb AIANNH/OTSA/off-reservation-trust geography leads.
• Results remain research leads only; users must verify coordinates and official records.

OpenStreetMap/Nominatim usage must stay end-user-triggered and low-volume. Do not use this endpoint for bulk geocoding.

v0.36 — Dawes quick-search layer and branded PDF packet

This release adds a fast homepage search layer for Dawes / Five Tribes record leads. It loads data/dawes_index.json and lets users search by name, tribe, enrollment category, roll/enrollment number, census card number, and keyword before the full allotment-map transcription is complete.

The release also adds a branded research packet print feature. Users can add Dawes lead results to a packet, click Print / save PDF, and use the browser print dialog to save a branded PDF. The packet includes AllottedLand.com branding, result fields, next research steps, source links, and a research-only disclaimer.

New files:

• assets/dawes-search.js — static Dawes search and branded print-packet logic.
• data/dawes_index.json — starter JSON index with one NARA guide example record for testing.
• data/dawes_index_template.csv — bulk-entry template for reviewed public Dawes index rows.
• tools/build_dawes_index_from_csv.py — converts the reviewed CSV into data/dawes_index.json.

Important: this version does not scrape Ancestry, FamilySearch, or other restricted databases. Add only data that has been lawfully obtained, reviewed, and allowed for public indexing. Treat all matches as research leads until checked against original records.

v0.35 — Land Loss Project pattern-evidence intake

This release expands evidence.html from a short evidence form into a structured pattern-evidence intake. It asks for allottee identity, land description, land-loss mechanism, actors, protection/vulnerability flags, document checklist, federal approval status, and pattern-research consent.

The form still submits the existing compact backend keys (family_name, tribe, county, decade, loss_method, source_type, summary, source_note, contact, and consent) so the current pending-evidence queue can keep working. Extended details are compiled into the review summary/source fields and included as draft_details for future backend support.

Fast data that can be collected before full map transcription: Dawes roll/card basics, allotment-packet status, county book/page, probate case number, tax-sale year, sheriff deed, mortgage/release, BIA/LTRO file clue, and oil/gas lease clue.

v0.34 — Site-wide map background

This release makes the angled source-map pattern consistent across all included pages by using the shared assets/v032-visual.css override and adding that stylesheet to About, Contact, and Source Records. The green header remains solid for readability.

v0.33 — Green header + Land Loss Project

This release keeps the original green header color while leaving the angled source-map pattern in the page body only. It also renames the public evidence initiative to Land Loss Project while preserving the existing evidence.html URL.

v0.32 — Home mission, consent placement, and visual background

• Moved search/research-path consent checkboxes next to the buttons they unlock.
• Removed the separate “Before using the research tools” box from the home page.
• Expanded the Mission panel with more detail about source-first, family-guided research.
• Added a light angled map-pattern background outside the green header area.

AllottedLand.com

Project contact: allottedland@gmail.com

AllottedLand.com is a free public research-tool prototype for helping Native families find allotted land from partial clues such as name, tribe, roll/enrollment number, township/range/section, county, town, cemetery, or family story.

Current status

Beta / Phase 1. Current public version: v0.9. See CHANGELOG.md and changelog.html for project updates.

Beta / Phase 1. The current live search starts with the Library of Congress 1909 Cherokee Nation atlas map index. Name, roll-number, allotment-number, county-routing, testimonial, and land-loss datasets will expand only as verified records are added.

v0.5 changes

• Added public project email: allottedland@gmail.com.
• Updated contact.html with general question, correction/removal, and volunteer mailto links.
• Added contact email to footers and policy pages.
• Added a reminder not to email sensitive private family documents unless redacted and permitted.
• Added basic Organization structured data on the homepage with project URL and contact email.

Site pages

• index.html — homepage, guided finder, search, process guide, request builder, stories/data preview
• about.html — mission and beta scope
• sources.html — source-record explanation and official source links
• transcribe.html — indexing/transcription plan
• contact.html — contact/correction guidance and privacy warning
• privacy.html — starter privacy policy
• terms.html — starter terms of use
• submission-policy.html — starter submission consent policy
• sitemap.xml — sitemap for search engines
• robots.txt — crawl rules and sitemap location

Safety rule

Do not accept uploads, testimonials, corrections, or family documents until Privacy, Terms, Submission Consent, and a review/removal process are finalized.

Deployment

Static site for GitHub + Cloudflare Pages. Suggested Cloudflare Pages settings: Framework preset None, build command blank, build output directory / or ..

v0.6 update

Replaced primary contact buttons with Gmail web compose links, kept a default-email-app fallback, and added a copy-email button so contact works even when mailto links are not configured in the visitor's browser or operating system.

v0.7 update

Added public project-update tracking with changelog.html and CHANGELOG.md, updated navigation/footer links, and added the updates page to the sitemap.

v0.9 update

Added a local Map Indexing Agent starter kit:

• tools/map_indexing_agent.py — downloads one public LOC map image, tiles it, runs Tesseract OCR, and writes candidate rows.
• tools/review_candidates.html — local browser review helper for OCR candidate rows.
• tools/requirements.txt — Python package requirements.
• docs/ocr-workflow.md — installation, running, review, and safety workflow.
• data/allotment_records_candidates.json — unverified OCR candidate holding file.

OCR candidates are not public verified records. Move rows into data/allotment_records.json only after human review against the original source image.

v0.10 Windows Tesseract path helper

If Windows says tesseract is not recognized even though Tesseract is installed, run the agent with:

python tools\map_indexing_agent.py --page 29 --max-tiles 12 --psm 11 --min-conf 60 --preprocess threshold --tesseract-cmd "C:\Program Files\Tesseract-OCR\tesseract.exe"

The agent also now checks common Windows install locations automatically.

v0.11 — Tile Review Upgrade

The local OCR review helper now shows saved tile images beside candidate OCR rows, includes a bounding-box overlay where available, and adds filters/reject buttons so noisy map-label text can be screened before any row is drafted for data/allotment_records.json.

v0.28 homepage behavior

The public home page now reads approved records from /api/records and database counts from /api/stats. Pending records remain hidden from public search until approved through admin.html. The top navigation now links only to full pages instead of internal home-page anchors.

v0.29 admin/homepage cleanup

admin.html is now the place for internal counts, latest approved records, pending records, and review workflow. The public homepage no longer displays latest approved records or Phase 1 data-status cards.

The admin page does not load database content until the ADMIN_KEY is entered and verified through /api/admin-dashboard. The D1 APIs still enforce the key server-side. For stronger page-level protection, place /admin.html behind Cloudflare Access / Zero Trust.

v0.29b navigation note

The Workbench header now includes a direct Admin link for trusted reviewers. The admin page still requires the ADMIN_KEY before database data loads.

v0.30 notes

This version adds public Testimonials and Land Loss Project pages, a /login.html key helper for reviewers/admins, and D1-backed pending review queues for testimonials and evidence submissions.

After uploading v0.30, run this migration in Cloudflare D1 Console:

-- open migrations/0002_testimonials_evidence.sql and paste the whole file into D1 Console

The public source image still is not stored by the website. Workbench users load the LOC image locally in their browser, enter human-reviewed rows, and submit record data only.

County lookup is now supported through data/township_county_lookup.json. Do not rely on a county suggestion until the township/range-to-county crosswalk has been independently verified.

v0.31 consent gate update

Public search and research-path tools now require a Privacy Policy / Terms of Use checkbox before the buttons activate. Testimonial, evidence-project, and Workbench submission buttons remain disabled until required consent boxes are checked. Server-side Functions also reject submissions missing required consent confirmations.

This is not a substitute for legal review. The project should continue to review privacy, accessibility, child-privacy, data-retention, and state privacy requirements before wider public launch.

v0.37 hotfix deployment notes

Upload/overwrite these changed files after v0.36:

• index.html
• assets/dawes-search.js
• assets/v032-visual.css
• CHANGELOG.md
• changelog.html if using the public changelog page

This release fixes the Dawes search/PDF display bug. The print packet is hidden on screen and appears only in the browser print/save-PDF output. Dawes results no longer display before a user enters a search clue and clicks Search.

v0.38 deployment note

Upload these changed files over the current site:

• index.html
• assets/print-packet.js
• assets/v032-visual.css
• assets/app.js
• README.md
• CHANGELOG.md
• changelog.html

After upload, hard refresh the site with Ctrl+F5. The print packet must remain outside <main> in index.html; otherwise print CSS can hide the packet's parent container.

v0.38 data-source roadmap

The next data layer should prefer official/public source paths:

1. NARA Catalog API for archival descriptions, digital-object metadata, OCR/extracted text, and JSON search results.
2. Library of Congress JSON API and IIIF image services for map metadata, map pages, citations, and images.
3. Oklahoma Historical Society Dawes database fields as a search/path reference, subject to permissions and practical access limits.
4. BIA/LTRO as the official record path for trust/restricted title documents, including patents, deeds, probate orders, leases, rights-of-way, cadastral surveys, plats, and title status records.

Do not scrape paywalled/restricted sites such as Ancestry, Fold3, or FamilySearch. Link out or use user-provided/downloaded records only where permitted.

v0.39 official source connector hub

Upload these changed files over v0.38:

• index.html
• sources.html
• assets/source-connectors.js
• assets/v032-visual.css
• functions/api/nara-search.js
• functions/api/loc-search.js
• data/source_catalog.json
• README.md
• CHANGELOG.md
• changelog.html

Cloudflare setup

Set this Pages environment variable if you want live NARA API results:

NARA_API_KEY=your_read_only_nara_catalog_api_key

Do not place the NARA key in browser JavaScript or in the public GitHub repository. The site calls /api/nara-search, and the Cloudflare Function adds the x-api-key header server-side.

The LOC connector does not require a key. It uses /api/loc-search to query the Library of Congress JSON/YAML map endpoint and return lightweight source leads.

Source behavior

• NARA Catalog API: live source leads when NARA_API_KEY is configured; official link fallback if not configured.
• LOC JSON/YAML API: live map/source leads through /api/loc-search.
• OHS Dawes Rolls: official state linkout/search path.
• BIA/LTRO: official title-record request path.
• BLM GLO Records: official federal land-record linkout.
• NARA AWS dataset: bulk metadata path for future offline import work.

All output remains labeled as research leads. Users should verify every lead against the official record page or original source record.

v0.40 connector hotfix

This release keeps the same v0.39 Official Source Lookup interface but makes it safer for real public use. If the Library of Congress JSON endpoint returns HTTP 403, rate limiting, or another proxy failure, /api/loc-search now returns official LOC search-link cards with printable source leads instead of an error. /api/nara-search now tries broader searches and returns official NARA search/guide cards when the live API returns zero rows or the NARA_API_KEY is missing/placeholder.

Deploy with Wrangler from this folder:

npx.cmd wrangler pages deploy . --project-name allottedland --branch main

Then hard refresh https://allottedland.com with Ctrl+F5.

v0.41 deployment warning

This package is a full Cloudflare Pages deployment snapshot. Deploy this entire folder with Wrangler:

npx.cmd wrangler pages deploy . --project-name allottedland --branch main

Do not deploy a changed-files-only folder as the whole site. Cloudflare Pages deployments replace the deployed asset snapshot with the folder you deploy. A partial folder can leave the live site missing shared CSS, JavaScript, data, and supporting pages.

v0.43 public source-lead connectors

This build adds public source-lead tools that make the site useful while map indexing continues:

• /api/chronicling-search — Chronicling America / LOC historic newspaper source leads for tax-sale, sheriff sale, guardian, probate, oil/gas, and family-name searches.
• /api/fr-search — FederalRegister.gov public API source leads for BIA, Indian Affairs, tribal ordinances, land acquisition, and related notices.
• /api/census-lookup — Census Geocoder + TIGERweb AIANNH/OTSA source leads for address or coordinate lookups.
• Record request packet builder — printable request text for NARA, OHS, county clerk, BIA/LTRO, and BLM/GLO follow-up.

No new D1 migration is required. No new API keys are required for Federal Register, Census Geocoder/TIGERweb, or Chronicling America/LOC. NARA still requires NARA_API_KEY for live NARA rows.

Deploy the full folder with:

npx.cmd wrangler pages deploy . --project-name allottedland --branch main

Do not deploy a changed-files-only folder as the full website snapshot.

v0.43 Census lookup note

If Census returns no address match, use latitude/longitude. Rural, historic, or non-standard addresses may not geocode from Census address-range data.

v0.47 deployment note

Deploy the full v0.47 folder with Wrangler. This version makes Universal Search the single homepage workflow. The older advanced search tools still exist as code/API paths, but the homepage no longer asks families to choose among multiple search forms.

npx.cmd wrangler pages deploy . --project-name allottedland --branch main

After deployment, hard refresh the browser with Ctrl+F5.

Key test cases:

• Click “I don’t know where to start.” Results should begin with Built research path.
• Search Claude Ketcher roll 1637. It should not show unrelated public index rows unless they actually match.
• Search T24N R14E Section 1. Matching map/index rows should appear.
• Search an address. Census/geography leads should appear as research leads only.
• Click “Print all / save PDF” after results appear. The PDF should include the built research path and result sections.
• Click each Copy request button in the Agency record request packets section.

v0.49 — Homepage simplification and field-first Universal Search

• Combined the repeated homepage instruction panels into one Start Here panel.
• Moved Testimonials / Land Loss Project above the Universal Search form.
• Removed the redundant homepage “Submissions require consent” panel because consent is handled at the submit forms.
• Made Universal Search field-first instead of starting with a large text box.
• Removed the visible Search Summary section from results.
• Added plain explanations for Official source leads and Prepared official links.
• Removed “In plain English” wording from agency request packet bodies while keeping request text understandable.

v0.52 public-page simplification

All public-facing pages except admin.html and workbench.html were simplified around a single user goal: help families search first, understand the record path, and avoid confusing tool sprawl. Reviewer/admin tools remain available but are not promoted in the public navigation.

v0.52 - Free-first FamilySearch resource layer

• Added FamilySearch as the first free outside genealogy starting point in Universal Search results.
• Added data/partner_links.json for future free-first partner/resource cards.
• Updated Privacy Policy and Terms to explain outside resource links, future partner/affiliate disclosures, and the no-sale/no-paywall approach for families.
• Kept paid partner links inactive in this version.

v0.53 - FamilySearch start-here link

• Added a FamilySearch link directly inside the Start Here panel so families can gather names, dates, relatives, and places before searching AllottedLand.com.
• Removed the public-facing “Paid partner links are not active in this version” sentence from the homepage Start Here box.
• Updated homepage cache-busting to v053.