Fix region setup and simplify the bilingual documentation

[n30nex]

Summary

Implements the September English/French site audit on top of merged PR #89.

• Preserve local radio settings unless the operator explicitly chooses a profile; include the documented BC Mesh profile.
• Preserve unrelated MQTT connections, reuse existing Canada slots, and fail without changes when both brokers cannot fit.
• Keep location and setup choices through map/configuration and language handoffs; make ambiguous searches actionable.
• Improve light/dark focus contrast and mobile map layout, shorten setup wording, and remove unrelated next-role navigation.
• Load offscreen maps and collapsed tables only when needed; show map tiles before the boundary overlay, and embed the 17 icons used by the tools instead of downloading the full icon library.
• Simplify feedback, preserve its source page, and render the broker reference without JavaScript.
• Clarify conditional erase/DFU instructions, experimental hardware limits, listing review dates, and registry adoption status.
• Add regression tests, contributor instructions, and map/editor performance coverage without lowering budgets.

The audit follow-up maps A01–A17 to changes and verification. Existing community listings, restored guides, build downloads, admin contacts, and subscriber inventory are retained. No region boundaries, broker accounts, or credentials are changed. No new dependencies are added.

Important: gateway before site

The shorter anonymous feedback form needs the gateway update in this PR deployed first. Verify /config returns communityIdeaOptionalDetails: true, then publish the site. Existing complete v1 submissions and boundary proposals remain compatible. Against an older gateway, the new form offers copy/GitHub fallbacks rather than sending an incompatible request.

See the rollout instructions. This PR does not deploy or merge itself.

Verification

All three quality jobs passed for af65bc6: final workflow run.

• 530 browser tests passed, with no failures or flaky results; 34 project-specific skips are intentional. Coverage includes all 128 rendered non-404 routes on desktop/mobile and English/French critical journeys across six browser configurations.
• 160 JavaScript tests, 16 Python content tests, 47 gateway tests, and 17 automation tests passed. Both region-geometry checks passed.
• The strict build and link audit passed for 129 pages and 12,990 local references. The separate noindex subpath build passed 12,988 link checks and 10 local desktop/mobile setup and language-switch journeys.
• All 12 Lighthouse audits passed, including the map, bilingual mobile map, and editor. The final map scores were 96 on desktop and 100 on both mobile language routes; accessibility was 99–100 across the audited routes.

The final local map regression pass also completed 24 tests across all six browser configurations, including deferred loading, slow boundaries, retry after tile failure, marker images, and keyboard access. English desktop and French mobile layouts were visually checked.

Tests use temporary broker fixtures and simulated submissions, not live broker changes or public test issues.

The performance server now sends HTTP gzip for text assets, matching the verified live GitHub Pages responses. Performance thresholds and the original simulated mobile/desktop throttling profiles are unchanged. The server is bound to localhost and has compression, identity-response, and HEAD regression tests.

Still requires human evidence

Formal registry adoption and proposal #63; missing community contact/profile facts; physical reproduction and safe limits for the unfinished solar builds; infrastructure authentication/retention policy. Those are documented explicitly rather than presented as verified by software tests.

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-05T01:23:34.496246Z	af65bc6	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: af65bc60c0

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve the wizard step while traversing the map

When an operator reaches step 4, follows Explore regions, and then returns through Configure this region (or switches languages from the map), the incoming step=4 is dropped because this map state never initializes wizardStep, while mapHrefForState() only serializes state.wizardStep. The returned configurator therefore defaults to step 3 instead of restoring the review step; initialize a validated step value here alongside the other handoff settings.

Useful? React with 👍 / 👎.

[n30nex]

Added the requested Canadaverse entry to the read-only MQTT inventory in English and French in dfa4923, with n30nex@gmail.com as the contact and GTA Regional Tools & CartoLive map as the purpose (translated on the French page). The 14 Analyzer content tests, full content validation, strict build, and checks of both rendered entries passed. This is a documentation update only; no broker accounts or services were changed. GitHub checks are rerunning for this commit.