This past week the Carrier team completed our object interaction sprint with some success. While our engine received some much needed architectural additions and we arrived at a point where object collision is detected, our game does not resolve collisions appropriately. Buildings track when objects hit them but neither accept damage on collision or are destroyed. This sprint is somewhat shortened, so I decided that the game needed some polish and that object collision resolution, given that it is so crucial to gameplay, should be the primary focus of this sprint. I divided our goals into three objectives:
- Collision resolution between objects and buildings
- Collision resolution between objects and other objects
- Collision resolution between objects in the ground
While this sprint seems fairly straightforward, in meeting our team became aware of several potential design issues. For example, given that our environment is a city and thus should be densely packed with objects and buildings, keeping in line with our general policy of physics and mechanics leaning in the player’s benefit, how can we detect what the player is aiming for when they release an object from the winch? In example, if a car is directly in front of a building, how can we tell if the player is aiming for the car or the building if they toss an object in that direction? We decided to play with a couple different techniques and use the playtest on Tuesday to determine which makes the most sense. We may use tight collision boxes for increased precision, or use the angle of the thrown object to determine target.
Further, I decided as Product Owner that my team needs to focus solely on development this week, so I told them that I will be handling the playtest questionnaire and I’ll bring my proposed questions to them on Monday. Ideally I’d like to be testing complete object collisions, but the team may not be able to complete it in time, so I may have to adjust my questions appropriately.
Thanks much for reading.