This week was our final week of development. We focused on finishing all unfinished features, adding in the remaining animations, bug fixes, and polish.
For animations, we had a few remaining animations to implement, in particular eating, pooping, and sleeping. These were made with Maya, and were implemented in a similar way to the idle animation. To handle animation states, we made a simple animation state machine with a priority system - each animation had a priority, and animations would only play if they had a higher priority than the current animation (For example, the idle animation would not play if a sleeping animation was currently playing).
The pet now fetches both the ball and frisbee when they’re thrown at the dog park. They’ll run after the toy and return it to the player in their mouth afterwards!
We also finally implemented a stat icon that appears above a pet to show how their stats are changing without the player having to check the menu. This will help players more easily understand what the effects of different interactions have with the pet. These icons center above the dog and automatically stack above each other to allow for easy reading. They last for 3 seconds before disappearing, and will automatically rotate to face the player.
Additionally, we implemented some sounds for the dog that are stat-based; for example, the dog whines if it is hungry below a threshold, and pants if its energy level is too low. The dog barks at the park as well. We obtained these sounds from freesound.com as we typically do.
As noted by Prof. Yarger and some playtesters during the alpha phase, the area behind the house was awkwardly empty. We fixed this by adding some houses, fences and grass to simulate a neighborhood environment.
Some bugs we tackled this week included adjusting the way we handle states when teleporting to the park; we had a frustrating bug where the dog would try to run to the bed or the food bowl while at the park. This made it so that we couldn’t play fetch with the dog until returning to the living room. We fixed this by resetting the ‘state’ of the dog each time we returned to the park to ‘follow’, rather than ‘sleep’ or ‘eating’. Additionally, we fixed some other miscellaneous bugs as pointed out in the most recent playtest. For example, users noted that the ball didn’t soar as far as expected when thrown; we simply made the ball smaller, lowered its mass, and lowered its linear damping to mitigate this. There was also a bug concerning the dog not positioned correctly while eating, which was an easy fix made by moving a target follow position slightly.
Comments