Slide's Adventure Development Info
The production of Slide's Adventure began during the summer of 2005. Andy and I set out to accomplish an entirely original 3D game with eight different worlds. After the first world took a solid month to complete, we had to rethink our goals. If I was to continue to create one world a month, it would take me about eight months alone to complete just the levels, and that was too much time. There were also many other factors to be taken into account: animation, enemies, scripting, and playtesting.
After working with the editor for a few months, I was able to create just about every simple object I could think of. I wasn't too worried about how the levels were coming along, but I had no idea how we were going to animate our main character, let alone model him in the first place. I did some research and ran across a program called Blender, which is a 3D modeling and animation program. After completing a few tutorials, I set out to try and model our main character.
My first attempt
As you can probably tell, my first attempt didn't turn out so well. The main character ended up way too rigid and had thousands of tiny, individual surfaces. Basically, this model was inefficient and would bog down our engine, so I made a second attempt keeping these issues in mind.
My second attempt
Now this model was starting to look good. He has a decent sword and head covering this time, and I was even able to put in a nice bone structure to allow for maximum flexibility. In the end, I didnít end up using this exact model, but I used him as a strong guideline for later.
(Click to Enlarge)
Here we have a picture of the playerís animations. We have (from left to right) - running, walking, attacking, idling, jumping, climbing, and victory. Even though I had created all of these animations, I still had no way of putting them into the game, but thatís where Andy came in. With Blender, we were able to export all of the vertices into a raw triangle form, and Andy worked his PHP magic to create a simple program to format the data into a readable text file we could import into the game.
Textures are one of the biggest parts of any game, and they sure as heck better look good if you want anybody to even glance at your title. I started working on the textures for Slideís Adventure once I had completed building most of the objects for a level. My basic cycle was design > model > texture > playtest / edit. When I had initially started placing textures on the fundamentals a level (ground, buildings, sky, etc) I noticed they werenít cheerful enough. To me, the textures changed the atmosphere to an almost serious feel which was the complete opposite type of style I was going for.
Not exactly the style I was going for.
(Click to Enlarge)
Any red rectangular objects you see are roughly the size of the player. I liked using them as a reference for proportion, but I didn't always remember to keep that in mind. The first level of Slide's Adventure was a bit cramped, and the opposite was true of the city level. The second level, Snow Land, was my personal favorite. I think the textures really helped give it a cold feeling, and the level itself was the prefect size.
Level 2 with no textures
Level 2 with textures
Panoramic view of Snow Land
Various items used in Slide's Adventure
After completing four levels, there was still plenty of work to be done. We still didnít have simple things like death, 1ups, or a title screen, but one impressive development tool we did have was a camera track editor. I was able to open up any one of my levels, place and set a camera in a series of positions, and watch the camera playback with the push of a button. I could then save these tracks and code them into the levels (via our simple scripting code) for events like goal completion or level introductions. For example, every time the player grabs a goal, the camera pans away from him.
But in the end, there were still a few things we weren't ever able to finish.
As you have probably noticed, the player doesn't have a texture. He is purly white. We weren't able able to texture each one of his many triangles. When Blender exported our player, he was exported in 354 individual trianges, whereas each enemy was limited to less than 100. It would have taken too much time to texture each frame of his animation, so we left him textureless.
We were never able to perfect a major part of our game: hit detection. Horizontal and vertical platforms are still buggy, and rotating objects don't have working hit detection whatsoever. This was a major frustration for us, and we were never able to overcome these problems, so until we have a better understanding of how to fix this, these platforms will continue to stay broken. Eventually, I image that Slide's Adventures will get an update with the problems removed, but until then, there is nothing we can do.
Some platforms still work
(Click to Enlarge)
I think that Slide's Adventure turned out pretty well for two high school kids. From the beginning until it was finished, it probably took us about 14 months to complete. There were months where we wouldn't touch the game at all, and there were months where we worked quite diligently. If nothing else, I had fun working on it, and I think we can both say we accomplished something nice over the past two years.