Website overhaul: accurate content, creative inventory, roadmap page

[MichaelFisher1997]

Brings the marketing site in line with how far the game has come, and turns it into a multi-page property with a full roadmap.

What changed

Content corrections (homepage)
The site was stuck at the early-prototype state. Updated all copy to reflect the real game:

• Stack fixed: Three.js → Babylon.js (the game runs on @babylonjs/core)
• Dropped the stale "creative-mode slice / 4 biomes / 23 blocks" framing
• Now reflects: 11 biomes (8 climate-selected + ocean/mountain/snowy mountain; beach is a reserved palette entry selectBiome() never returns), survival + creative modes, day/night cycle, two-channel voxel lighting, flowing-water simulation, caves/ores, 6 tree species, per-seed saves, scalable graphics
• Fixed the cd Minebench repo name in the CTA

New /roadmap page (src/pages/roadmap.astro)
A vertical phase timeline with a status legend:

• Phase 1 — The foundation (Live now)
• Phase 2 — Crafting & survival depth (In progress)
• Phase 3 — Life, danger & the wider world (Planned) — incl. world-generation modes (Superflat, Amplified, Large Biomes, sky islands, cave world)
• Phase 4 — Multiplayer & beyond (Planned) — incl. multiple dimensions (Underworld, Void, Aether)

Creative inventory (Features.astro + site.ts)
Was showing a stale 9-block hotbar labelled "the creative palette". Now renders the full 39-item creative inventory (CREATIVE_PALETTE from Items.ts) in the game's curated order, with real representative colours from Blocks.ts.

Showcase

• Wired in two real screenshots (main-menu.png, in-game.png) — placeholders auto-replaced via build-time discovery
• Removed dev-facing "How to capture these screenshots" block + refreshed the lede

Plumbing

• Layout.astro forwards per-page title/description to BaseHead
• Roadmap link added to Header + Footer nav
• Narrowed root screenshots/ ignore (/screenshots/) so site screenshots track while the root benchmark screenshots/ dir stays ignored

Verification

• astro check — 0 errors, 0 warnings, 0 hints
• astro build — builds /, /roadmap, /play cleanly with the game bundle synced

Note

The root game dist/assets/ is owned by nobody (a sandboxed build), which blocks the full bun run build sync:game step with EACCES. Worked around by running astro build against the already-synced bundle. Worth clearing with sudo rm -rf dist before the next game-code change.

[github-actions]

📋 Summary

PR #6 does not reference any linked issues (no Fixes #N / Closes #N in the title or body).

This PR refreshes the Astro marketing site to match the current game: it corrects the stack from Three.js to Babylon.js, updates biome/block/mode claims, adds a new /roadmap page, replaces the stale 9-block palette with the full 39-item creative inventory, wires in real screenshots, and adds roadmap links to the header/footer. The code is clean and both astro check and astro build pass. The main remaining problem is a stale bundle-size claim that is now factually wrong by nearly an order of magnitude.

📌 Review Metadata

• Reviewed Commit SHA: b19f151147cca34aa6023ff7518d643c9106d925
• Reviewed PR: Website overhaul: accurate content, creative inventory, roadmap page #6
• Linked issues: None found.

---

🔴 Critical Issues (Must Fix - Blocks Merge)

None identified.

---

⚠️ High Priority Issues (Should Fix)

None identified.

---

💡 Medium Priority Issues (Nice to Fix)

[MEDIUM] website/src/data/site.ts:12 — bundleGzip: "~135 kB" is factually inaccurate
Confidence: High
Description: The marketing site still advertises a "~135 kB gzipped" production bundle, but the current Vite build emits a single index--*.js chunk that is 1,170.26 kB gzipped (5,229.64 kB raw). This claim appears in both Hero.astro and Technical.astro via site.bundleGzip, so it is user-facing and materially misleading.
Impact: Prominent marketing copy contradicts the actual shipped artifact.
Suggested Fix: Measure the real gzipped payload and update the constant. For the current build that is roughly the main JS chunk alone:

bundleGzip: "~1.2 MB",

Or remove the claim entirely if an exact bundled size is not being tracked in CI.

---

ℹ️ Low Priority Suggestions (Optional)

[LOW] website/src/components/Features.astro:28 — "placeable blocks, plants and foods" overstates placeability
Confidence: Medium
Description: The creative palette contains 34 blocks/plants and 5 food items. Food items are edible, not placeable, but the label says "placeable blocks, plants and foods", implying all 39 items can be placed.
Impact: Minor wording imprecision on a marketing page.
Suggested Fix: Change the line to something like:

<p class="palette-sub">{blockPalette.length} blocks, plants and foods — open the full grid with <kbd>E</kbd> in creative mode and place anything instantly.</p>

[LOW] website/src/data/site.ts:16-17 — treeSpecies and dayLength are unused
Confidence: High
Description: These new site fields are added but not referenced anywhere in the site components. They are accurate (6 tree types, 600-second day) but currently dead weight.
Impact: Negligible; just adds unused constants.
Suggested Fix: Either surface them in copy (e.g., "6 tree species" chip, "10-minute day" fact) or remove them to keep site.ts lean.

[LOW] website/src/data/site.ts:170 — Roadmap copy says "launch twelve" biomes while the site claims 11
Confidence: Medium
Description: Phase 3's "Advanced biomes" item says "expanding the climate map beyond the launch twelve". The PR and the rest of the site state 11 selectable biomes; beach is a reserved BiomeId that selectBiome() never returns.
Impact: Minor inconsistency in a forward-looking roadmap blurb.
Suggested Fix: Change "twelve" to "eleven" to match the selectable biome count:

body: "Mangrove swamps, cherry groves, mushroom fields and more — each with unique flora, surface rules and generation, expanding the climate map beyond the launch eleven.",

---

📊 SOLID Principles Score

Principle	Score	Notes
Single Responsibility	8	Components and site.ts data are cleanly separated; each file has a clear purpose.
Open/Closed	7	Roadmap is data-driven and easy to extend; minor tweaks would make unused site fields useful.
Liskov Substitution	9	No inheritance hierarchies to violate.
Interface Segregation	8	Small, focused Astro Props and data interfaces.
Dependency Inversion	8	Components depend on imported data/config, not concrete internals.
Average	8.0

---

🎯 Final Assessment

Overall Confidence Score: 75%

Confidence Breakdown:

• Code Quality: 85% (well-structured Astro/TS, clean CSS, no type errors)
• Completeness: 80% (content refresh is thorough, but bundle-size metric was missed)
• Risk Level: 20% (marketing-site only; no game logic touched)
• Test Coverage: 50% (no new tests, but astro check and astro build both pass)

Merge Readiness:

• All critical issues resolved
• SOLID average score >= 6.0
• Overall confidence >= 60%
• No security concerns
• Tests present and passing (typecheck + build verified)

Verdict:

MERGE WITH FIXES

Update the stale bundleGzip figure before merging; the other low-priority items are optional but would polish the PR.

opencode session  |  github run

[github-actions]

📋 Summary

PR #6 does not reference any linked issues (no Fixes #N / Closes #N in the title or body).

This PR refreshes the Astro marketing site to match the current game: it corrects the stack from Three.js to Babylon.js, updates biome/block/mode claims, adds a new /roadmap page, replaces the stale 9-block palette with the full 39-item creative inventory, wires in real screenshots, and adds roadmap links to the header/footer. The second commit addresses all issues raised in the previous automated review. Both astro check and astro build pass cleanly.

📌 Review Metadata

• Reviewed Commit SHA: 85b87634e037ad9fd99795c695cb2206fd4f617a
• Reviewed PR: Website overhaul: accurate content, creative inventory, roadmap page #6
• Linked issues: None found.

---

🔴 Critical Issues (Must Fix - Blocks Merge)

None identified.

---

⚠️ High Priority Issues (Should Fix)

None identified.

---

💡 Medium Priority Issues (Nice to Fix)

✅ [FIXED] Previous issue: website/src/data/site.ts:12 — bundleGzip: "~135 kB" was factually inaccurate
The value is now "~1.1 MB", which aligns with the actual gzipped production chunk (~1,170 kB). No new unresolved medium-priority issues identified.

---

ℹ️ Low Priority Suggestions (Optional)

✅ [FIXED] Previous issue: website/src/components/Features.astro:28 — "placeable blocks, plants and foods" overstates placeability
The subtitle now reads "blocks, plants and foods — open the full grid with E in creative mode", removing the implication that foods are placeable.

✅ [FIXED] Previous issue: website/src/data/site.ts:16-17 — treeSpecies and dayLength were unused
Both fields are now surfaced in website/src/components/Hero.astro:38,40.

✅ [FIXED] Previous issue: website/src/data/site.ts:170 (now website/src/data/site.ts:294) — Roadmap copy said "launch twelve" biomes
Now correctly reads "launch eleven" to match the selectable biome count.

No new unresolved low-priority issues identified.

---

📊 SOLID Principles Score

Principle	Score	Notes
Single Responsibility	8	Components and site.ts data are cleanly separated; each file has a clear purpose.
Open/Closed	8	Roadmap is data-driven and easy to extend; new site fields are now consumed.
Liskov Substitution	9	No inheritance hierarchies to violate.
Interface Segregation	8	Small, focused Astro Props and data interfaces.
Dependency Inversion	8	Components depend on imported data/config, not concrete internals.
Average	8.2

---

🎯 Final Assessment

Overall Confidence Score: 92%

Confidence Breakdown:

• Code Quality: 90% (well-structured Astro/TS, clean CSS, no type errors)
• Completeness: 95% (content refresh is thorough; previous factual issues resolved)
• Risk Level: 5% (marketing-site only; no game logic touched)
• Test Coverage: 60% (no new tests, but astro check, astro build, and game tsc --noEmit all pass)

Merge Readiness:

• All critical issues resolved
• SOLID average score >= 6.0
• Overall confidence >= 60%
• No security concerns
• Tests present and passing (typecheck + build verified)

Verdict:

MERGE

The PR is ready: it is a low-risk marketing-site-only change, all previous factual issues have been fixed, and both typechecking and production builds pass cleanly.

opencode session  |  github run