Monthly Archives: September 2012

Bleeps Run – First Screenshot

A couple weeks ago I was working on improving my character making and animating skills.  The little blue guy below is the result.  My goal was to make a cute simple character I could use in run and jump type games (2.5D/3D platform or runner type games, etc).  Everything else is the result of this weekend.  This is just a test level with a bunch of extra junk so I could test rendering performance on mobile.  After some tweaking of the shader I was able to keep the FPS at or above 32 FPS on a Motorola Droid 2.  Thats about a 2 year old phone, so I declare the performance testing as “pass”.  Actual levels will probably have less junk in the back to slow down performance.

GeomeTree LD48 Post Mortem

It was great to get so much feedback on a game. Some good, some bad, but overall very valuable. Overall I haven’t disagreed with any of it. The good parts validate that the concept I was going for was worth pursuing.

The Good:

The simple graphics worked out well. At first it was too big and bulky but after some tweaking I got it to look alright, and without spending any time on textures or models. Not that textured models are bad or that I didn’t want to do them at all, they were just out of the scope I laid down. They weren’t important to the overall concept

The single 1 2 3 encoding for the trees worked out well and was easy to mutate. If I end up doing something more with this, the coding would be different, or at least the walking of tree nodes. Along those lines, tree branching just takes an angle in one direction and goes with it. It’s very ‘good enough’. I had no idea if the algorithm/technique in my head was going to work at all, and it did with just a small amount of tweaking.

The bad:

I spent most of my time just getting the tech to work. I neglected objectives. In part I didn’t want there to be an objective. I wanted it to be about random tree stuff, exploration, that sorta thing. I really wanted to throw in a piece of server code and share trees and have the game show other trees, let people vote on them. In retrospect, I should have, I think it could have made the deadline still.

Personally I think the genetics could have been better, as I mentioned above I would do something bigger if I follow up on this concept. I won’t say it was totally bad because I knew going into it that I needed to limit my time on the tech/algo. (related, early on I decided against having an algo just make symmetric trees, they look to uniform, and it’s been done in my opinion).

I really wanted more trees to pick from. 3 fit on the screen easily, so I went with that. Really though I think there needed to be about 20 per generation. If I expand on this maybe it’ll be as a tree farm.

The end:

My original vision was a lot different than the end result. I planned to have a root system, and have the leafs on the trees gather light, even have the tree get larger around as it grew. So a tree without enough leafs or roots wouldn’t survive, that sorta thing. I haven’t thrown out the idea of doing more with this concept, at moments I am really excited about the idea, but I think it needs to simmer for now.

Overall I had a ton of fun and look forward to next time.