They way collisions work in Identity is interesting. There are four ideas at work here. First, there is a bounding sphere at the base of the player. This is the one that stops you from walking through walls.

Next, there is a larger sphere around the bounding sphere. I call this the recognition sphere. All things colliding with this sphere will be considered when you press the action button. This is how you open doors, read books, talk to characters, etc.

For gravity, there is a ray trace from the player to the floor. Whenever the ray is longer than zero, then player is in the air - so we pull the player down until the distance is zero. The

Finally, there is a ray traced from the players head to the camera. The camera always moves to the first point closest to the player when there is a collision. This prevents things from blocking your view of the player.

The environment is scaled to realistic proportions (and yes, I did use measuring tape to make sure it's just right). Everything was modeled in Blender. Collision surfaces were added here, too.

I wanted the menu to pop out at you, so I made it 3D. You can zoom it in and out. It appears at a random angle every time you press the menu button (the angle is limited to that you always have a clear view of the menu options).

If you've played Ragnadox games before, you'll recognize these characters. This is the first time they've appeared in 3D.

I use an unwrapped texture to skin the characters. Packed textures allow multiple characters to load from a single resource.

All planning starts with a pen and paper. From there, I make basic models, then add textures, then update the models, then add effects (lighting, shaping details, etc).

Identity was developed in Python using a very simple text editor with syntax highlighting.