Plan for John Vanderbeck

My disastrous attempt at GID #11

My disastrous attempt at GID #11. Whoah what a mess. You can read a play by play journal that I kept during the event with screenshots and a link to the game itself (what there is of it) here.

It started out what I thought was simple enough. I wanted a simple game that would help me learn Torque2D, but that would be fun and addicting so that after GID was over I could polish the game up as a commercial release. My choice for a game was what I call TileIt. Essentially you have a vertical game board consisting of a fixed hard border around the edges and a series of moveable tiles inside the border. You remove tiles by clicking on adjoining tiles that match each other. To remove tiles they must match and be touching either horizontally, vertically, or diagonally. The twist is that as you remove tiles, you can also rotate the game board. When you do so, tiles will fall down through the spaces left by removing tiles. In this manner you can continue to remove tiles and rotate the board to give you openings to remove more tiles It sounds like loads of fun to me. I also thought it would be fairly simple because I could just let the Torque2D physics handle the falling tiles.

Boy was I in for an eye opener. For one thing, I had to work both Saturday and Sunday at my "real" job, so in the end during the whole GID event I only had about 12 total hours to apply to my entry. A game in 24 hours is hard enough, but 12 for my first GID was hair-raising!

For another this was intended as alearning excercise. I was still a relative noob with T2S having only purchased it a week or two ago. I understood the raw basics and nothing else. This ended up causing problems for me time and again as I wasted valuable time struggling to figure out how to do something, or fix something. Despite that however, T2D was a virtual gem to me and it was only because of T2D that I did in fact get so much done in so little time.

In less than 2 hours I had managed to get a large majority of the game coded up, giving me a partial game board (I only coded up enough for prototyping at this stage), and the basic gameplay elements. It was very nice.

My first major problem though was when I realized the built in T2D physics just weren't giving me the ultra simplified tile dropping that I required. The physics were TOO good :) Tiles that I wanted to drop STRAIGHT down and stop were in fact spinning due to friction, bouncing all over the place, and being nudged out of thier slots. After a long few hours I managed to tweak the physics properties to fix most of those problems but a few I never could fix including the fact that even with my friction set at 0 the tiles acted like they had surface friction applied, causing them to drop down very very slowly when between other tiles. Sometimes not drop down all the way becuase they'd get stuck. They also kept getting "nudged" over a few pixels every now and then screwing up the whole game board. So eventually afgter probably half my GID time was over I ripped out all the physics stuff and tried to put in my own collision response. My real simple collision response was if two tiles collide stop them :p I then went to bed as I had to work on Sunday.

Sunday morning before work I grabbed a few minutes to apply some AWESOME graphics provided by a fellow GIDer Joe Lesko. Thanks again Joe! Then I had to head off to work for the majority of the day, ug.

When I got home and started to work on fnishing things up I knew my super simplified collision response just wouldn't cut it. I tried various things like checking positions on the tiles to see if they were next to each other or above, below, whatever, but that never panned out. I didin't understand the collision normals enough to do the same, but I suscpect even if I had it wouldn't have worked in some cases. After about 2 hours of the approximately 5 I had left till midnight I threw out the colllision all together and completely wrote a function to scan the whole game board and manually move the tiles around. I had come a long way from what I had intended. Originally planning to simply rotate the board and let gravity and T2D Physics do the rest, here I was writing and algorithm to scan each and every tile, look underneath it, and move it down if there was nothing there. Using a grid system.

Now I FINALLY had my game board workign in that I could remove tiles and other would fall properly. Yay!

Only one nagging problem left, and that was that the board rotated funny. I was having weird issues with the T2D rotation that seemed to be causing the boards' tile transforms to get messed up when rotating. I set down to fix it and 4 hours later was still unable to do so :( At this point I gave up in fristration, declared a bug in T2D (Which in fact I don't know for sure I was just mad at the world by this point) and said "Heres my GID its unfinished but i'm out of time".