Let's Play Catch

Friday 9/18: A little work on the beginning of an AI scripting system.
Monday 9/21: Tweaked the camera and throwing mechanics. Added some space at the bottom of the screen when holding the frisbee that you can use to turn.
Tuesday 9/22: Tweaked the AI system. Started on the frisbee reading/catching logic.
Wednesday 9/23: Finished the frisbee reading/catching. More camera tweaks. Added a ground "shadow" to make it easier to read the frisbee.

I think I'm almost to the point where I can implement a simple drill.

Plodding along

The last week has been fairly terrible as far as visible progress is concerned, but I think I got a lot of necessary work done that will pave the way for future improvements.

Thursday 9/10: Reworked the frisbee flight calculations using some actual physics equations. Added transition actions to show a smooth transition between running and standing, for instance.

Friday 9/11: Started to expand the action system to allow for more complex actions.

Monday 9/14: Day off.

Tuesday 9/15: Used the new action system to create a more complex throwing motion, including pivoting.

Wednesday 9/16: Continued working on the throwing mechanics.

Thursday 9/17: Created a more advance UI display system. Finished (for now) the throwing system. Basically, a target arrow pops up after you pivot out that helps you judge the direction and distance of your throw.

Next up...AI?

I can run! I can throw!

I ran into some frustrating problems getting my animations to work. I ended up sticking with my own modeling format (instead of .md2). I might change later, but for now it serves my needs.

By the way, I think I work in cycles. First, I carefully plan and execute a feature. Then, when the end is in sight, I rush to get it finished so that I can see it on the screen. Then, hopefully, I clean up my work so that it will work in the future.

Tuesday 9/8: Got the run animation loaded. This took way longer than it should have.

Wednesday 9/9: Made a forehand animation. Refactored the animation code to handle more than one animation in the same model. Made an action system to handle the timing of animations (as well as the actions they represent). All of this also involved animating the frisbee to act with the player. Added animation frame interpolation, so that animations are smoother and require less disk space.

Virtual Axis

Making steady progress with the graphics...

Saturday 9/5: Made a skeleton for the model and bent it into a quick pose. Also made a basic texture with skin color. Got the model loaded and displaying. Added simple lighting. Made a better frisbee model and texture.

Sunday 9/6: Improvements to texture loading system. Gave the field a grass texture. Added hair and eyes to the player texture.

Monday 9/7: Filmed myself running around and throwing frisbees. Completed a 'run' animation, but haven't loaded it yet. I think I'm going to use .md2 models (the Quake II format), because they are pretty compact.

More Lies

So, it seems I keep lying. Every time I write here I say the graphics are good enough and that I'll spend the next few days working on game mechanics. And then, invariably, inevitably, I go back to work on the graphics. Maybe it has something to do with my insecurity over the look of the game, along with my confidence in hashing out the gameplay.

Wednesday 9/2: Added some structure for player AI. Currently, once the other player catches the frisbee, it will chuck it back at you as soon as it can see you, but it doesn't move at all or throw accurately. I also did some work on loading models with multiple frames (i.e., animations).

Thursday 9/3: I spent the day following modeling tutorials and making something terrible from scratch. I was happy with it at the time though. I won't show the result.

Friday 9/4: I spent the day modeling myself. It looks great. I also created the texture mapping.

Modelin'

Monday 8/31: Two-finger swipe camera panning. Model loading. Basic frisbee model.
Tuesday 9/1:  Basic player model. Added another player to throw to. UI drawing layer, including a nice button class. Different camera modes and a change mode button (the white rectangle). Added rotation to the player based on direction.

So, I lied about how long it would take to get real models in the game. The transition from Blender to OpenGL is actually pretty painless, although I don't have any animation yet. Now that it looks a bit better, I can move on to throwing / catching mechanics and basic player AI.

Zoomin'

Thursday 8/27: Day off (moving)
Friday 8/28: New field class (rendered as a big green rectangle). Two-finger pinch zoom.

So, after one week, there is basic code in place for graphics, input, camera control, movement, and throwing. It's not pretty, but I'll probably wait to do a whole round of graphics improvements at once. I bet that will come in week 3 at the earliest.