After dialing back the projection of work for the week it's been easier to focus on a few of the smaller things. In that sense, the Androidbot has been removed for now; it was too difficult to deal with a new AI Core and re-scripting it, while simultaneously figuring out how to match the Animation controller to the new AI Core. Unity, it turns out, wont allow me to animate any of the IK rigs for the model inside of Mechanim without buying Unity Pro, so I'll have to reanimate the model inside Blender again and then re-import it. Legacy Animations also would not work from the way the model was rigged/animated/exported, it would all have to be redone anyways. There would be no sense in simply "making it work", I want to do it right; I see that as what will set me apart from other Independent Developers.
Since the original Robot model was still there, I added the new Rain AI Core scripts that I wrote to it and am now testing some of the commands I've written; as well as the new Navigation Mesh, Waypoint Route, and Waypoint Network. So far, the Navigation Mesh is established and working perfectly. I set a robot out to just wander around the map and it only got stuck once, allowing me to learn that two of my buildings had colliders which were set to "Trigger" mode; the player could have simply walked through them. Bug evaded on that one. The Waypoint Route is working fantastic, and after some quick help from the extremely helpful Rival{Theory} support the robot no longer clips through the ground, or falls over due to gravity. Those guys are seriously #1 in my book.
The Waypoint Network is also working well, as long as the robot isn't programmed to wander randomly from the path. Although knocking the robot off the path will lead him to find his way back to the system automatically. He wanders just fine about 75% of the time, then every once in a while he looks for a spot just out of reach and has no way to get there. I'm fairly certain this is due to my Navigation Mesh allowing the tops of buildings as potential navigation points; it's definitely something that can be fixed without too much work from me.
For now, I think the only game-mode which will be updated for the Alpha 3 will be the Robot Overlord level; although as with the rest of the Alpha Builds, all of the old content/game modes are going to be left in-tact. This will mean there will be an Alpha 3 Robot Overlord featuring the new AI Core, and the old one from Alpha 2. The options on the main menu will be changed to reflect this.
My sole goal for Alpha 4 will be to have Androidbot re-animated and implement him with some new attacks into the Robot Overlord game mode. If this proves to be a simpler undertaking then I assume, I will use the rest of the time to spruce some of the game up, and maybe animate/script some of the other monsters into the game world. After this weeks disaster in development though, I don't want to get ahead of myself again. I have 5 more months to achieve my original 6-month to Beta goal, there really is no reason to overdevelop at this point. I think going from not knowing Animation, Rigging, AI scripting, or AI Path Navigation scripting, to implementing all of those into a video game in two weeks will be satisfying to me as a developer; and give me the foundation to easily expand on that knowledge and implement more enemies easier throughout development.
Remember if you are following this blog or project: 3 weeks ago I had roughly ZERO experience in making video games, and I grow more knowledgeable by the hour. That's what 100 hours a week practicing something will do. Do I consider myself a great developer? No, but I hold myself to great developer standards; standards which most companies no longer meet in today's market. I never want someone to be able to call me lazy, regardless of whether they like the finished product or not. That's my only goal in the end, to have a product that people are amazed was made by such a small team (1 developer and a 3D modeler).
No comments:
Post a Comment