Day 16 — Crafting System Foundations

Context

Today marked the beginning of one of the largest and most important systems in Island Crossing: Crafting. This feature ties together nearly every other gameplay loop—gathering, tools, resources, progression—and will become a core part of the long‑term player experience.

Rather than trying to implement everything at once, I split the work between the AI Agent and myself so progress could happen in parallel.

System Architecture & Requirements

I began by outlining the full set of requirements for crafting, including:

• Crafting recipes
• Required ingredients and quantities
• Output items
• Crafting stations
• Crafting UI flow
• Validation logic
• Inventory integration
• Placement and pickup behavior

The list is large, but defining it early will reduce rework later.

Parallel Development Workflow

To keep things moving efficiently:

• The AI Agent began implementing the backend logic and data structures.
• I focused on building the UWs (User Widgets) that will form the Crafting UI.

This separation made it easier to work iteratively without blocking on UI or logic dependencies.

UI Work

I created the early versions of the UWs that will handle:

• Displaying available recipes
• Highlighting required materials
• Showing the player’s current inventory
• Crafting confirmation flow

Most of these are placeholder visuals, but they establish the overall layout and structure the system will build on.

---

Summary

What I accomplished:

• Defined the full requirements for the Crafting system.
• Began implementing Crafting backend logic via the AI Agent.
• Created the initial set of Crafting UI widgets.
• Established a parallel workflow to speed up development.

What I learned:

• Large systems benefit from early requirement gathering to avoid rework.
• Parallel work between UI and backend is effective when the system is well‑scoped.
• Crafting will likely be one of the most interconnected systems in the project.

Day 15 — Resource Gathering & Tile Coexistence

Context

Today’s work expanded gathering mechanics and improved tile flexibility.

Resource Additions

• Branches from shaking trees.
• Wood Piles from chopping trees.
• Stone from pickaxing rocks.

Tile Coexistence

• Grass can now coexist with other tile occupants, increasing interaction flexibility.

Interaction Prompt

• Added UI prompts for interactions to improve clarity.

Asset Expansion

• Used the automation pipeline to generate icons and meshes; also purchased assets for future crops.

---

Summary

What I accomplished:

• Added Branches, Wood Piles, and Stone.
• Implemented tile coexistence.
• Added interaction prompts.
• Expanded asset library.

What I learned:

• Coexistence greatly simplifies world design.
• Interaction prompts are essential as interactions increase.

Day 14 — Trees, Rocks & Asset Automation

Context

Today focused on expanding world variety and building tooling to accelerate asset creation.

Natural Resources

• Added random trees and rocks during initial world generation.

Item Catalog Refactor

• Centralized item definitions to remove scattered Item IDs.

Asset Pipeline

• Set up Python automation to:
  • Generate concept images via OpenAI.
  • Send images to Meshy.AI to create meshes.
  • Download FBX files.
  • Compress textures.

---

Summary

What I accomplished:

• Added trees and rocks to world generation.
• Refactored item catalog.
• Built an automated asset generation pipeline.

What I learned:

• AI pipelines are powerful but inconsistent for stylized assets.
• Automated tooling pays off even if early outputs aren’t perfect.

Day 13 — Generative Grid & Grass-Cutting Loop

Context

Today focused on making the island generate itself on a player’s first session.

Procedural Grid

• Replaced the static 2,500-tile prefab with procedural generation. Initial load now creates the island, which is then saved for future sessions.

Island Size Adjustment

• Reduced the grid size by 50% to bring generation time to ~20 seconds.

Grass Cutting

• Implemented the Grass resource loop:
  • Sickle harvests grass.
  • Grass is added to inventory.
  • Dropping Grass replants it.

---

Summary

What I accomplished:

• Implemented procedural grid generation.
• Added the grass harvesting and replanting loop.
• Optimized initial load times.

What I learned:

• Procedural systems must balance performance and player expectations.
• Resource loops quickly add depth to the world.

Day 12 — Persistence Hardening & Farming Completion

Context

With the farming loop in place, today was dedicated to polishing persistence and fixing edge cases.

Inventory Improvements

• Added Stack Drop support.
• Refined smart pickups to prevent overfilling.

Crop Persistence

• Implemented a “ratchet system” to ensure crops never regress when rejoining the world (i.e., When testing, the date always resets when restarting a session…but don’t regress the crops).

Critical Bug Fix

• Fixed a major issue where updating a tile incorrectly erased its saved growth stage.

---

Summary

What I accomplished:

• Completed farming persistence.
• Added advanced inventory behaviors.
• Fixed a major tile save corruption bug.

What I learned:

• Persistence must be validated at every step.
• Multi-day systems require safeguard logic.

Day 11 — Farming Loop Foundations

Context

Today marked the beginning of the farming gameplay loop.

Tile Lifecycle

• Tiles now support the full farming lifecycle:
  • Empty → Hole → Planted → Growing

Tool Behaviors

• Shovel digs holes.
• Seeds plant crops.
• Watering Can waters planted crops.

Growth Simulation

• Crops progress using timestamps stored at planting. The world_time_component now drives multi-day growth.

---

Summary

What I accomplished:

• Implemented the basic farming loop.
• Added tool behaviors for planting and watering.
• Built time-based crop stage progression.

What I learned:

• Time-based systems need careful persistence handling.
• Tile state transitions must remain deterministic.

Day 10 — Drop/Pickup Completion & Inventory Model Redesign

Context

With the refactor complete, today focused on finishing the Drop/Pickup milestone and stabilizing the inventory architecture.

Drop & Pickup Completion

• I implemented the full priority-based tile selection for dropping items. If no valid adjacent tile is available, the system now displays a toast message to the player. Items can also be picked up and placed into the correct inventory slot.

Inventory Model Redesign

• I refactored the inventory data model to use explicit slot-based storage rather than a virtual stack count. Each slot now stores both an Item ID and Quantity. This ensures the UI, logic, and persistence layers stay tightly aligned.

---

Summary

What I accomplished:

• Completed the Drop/Pickup milestone.
• Rebuilt the inventory model using explicit slot-based storage.
• Added toast notifications for invalid drop attempts.

What I learned:

• Clear data ownership prevents subtle bugs.
• User feedback (like toasts) matters even in early builds.

Day 9 — Grid Math Refinement & Prefab Stability

Context

Refactoring continued with a focus on stabilizing grid behaviors.

Grid Alignment Fixes

• Inconsistent grid math caused tool and item misalignment. I unified these calculations.

Prefab Cleanup

• Cleaned up several prefabs to reduce Scene Graph crashes.

---

Summary

What I accomplished:

• Unified grid calculation logic.
• Improved prefab clarity and stability.

What I learned:

• Grid calculations must be centralized.
• Scene Graph can be fragile with frequent prefab edits.

Day 8 — Major Refactor & Early Asset Tests

Context

Today focused on reorganizing the project into cleaner, more modular components while exploring options for generative tiles/grid.

System Decomposition

• Large, overloaded systems were broken down into focused components.

Tile Generation Tests

• Attempts to generate 2,500 tiles at once caused editor instability (5-10 minutes to save and Push Changes).

Asset Tests

• I tested Fab assets with the Curved World material to explore stylistic compatibility.

---

Summary

What I accomplished:

• Completed a major structural refactor.
• Tested large prefab generation.
• Evaluated external assets with curvature.

What I learned:

• Prefabs can overwhelm the editor at scale.
• Strong architecture accelerates future features.

Day 7 — Drop/Pickup Foundations & Architectural Limits

Context

Adding Drop and Pickup actions revealed structural limitations.

Drop & Pickup

• I implemented basic dropping and picking up of items.

Architectural Issues

• Several systems were doing too much, signaling the need for a larger refactor.

---

Summary

What I accomplished:

• Added early Drop and Pickup behavior.
• Identified structural bottlenecks.

What I learned:

• Refactoring is essential when scaling core systems.