I skipped the progress report in April because there was very, very little progress to report. My sole "accomplishment" for that month was to analyze the physics and AI code from MHFramework 3 and refresh my memory on its features, interfaces, strengths, and weaknesses. That was all.
May was a bit more productive. Not nearly enough to catch up with the project plan, but it began a long and complex path toward the implementation of some core functional requirements for Operation SCARE.
Accomplishments
- Through a lot of trial and error, I ported my old Java code to C#, and then refactored it into a more modular and dynamically customizable set of behaviors. Testing is currently in progress.
Known Issues
- The tile map subsystem is spaghetti. Before I go any further with the ghost AI, I need to clean it up.
- Though it has been improved, there is still some poor cohesion that should eventually be addressed.
- Potential (likely) performance issues in tile map actor collision checking
- Need to improve player movement and wall collisions through steering forces.
Next Steps
- Refactor engine's tile map system
- Implement dialogs and text areas
- Choose and create a game font for normal text
- Create container objects
- Cabinets, chests, etc.
- Continue to improve the game's runtime architecture
- Consolidate game's input functionality
- Decompose game screen code to MVC pattern
- Consider cohesion issues vs. class explosion risks
- Experiment with better wall geometry
- Implement ghost controller
