CAR TYCOON

WHAT WENT RIGHT:

This thing runs at 60 fps on a $50 android phone… great success! I was banging my head up against a wall to shave it down, and I finally figured it out

I got the libraries I wanted to test working on a prototype level. This is usually the biggest breakthrough, since it runs on a production device and it works. The weeks of polish will bring it to a real product that consumers can use.

WHAT WENT WRONG:

Man, getting Unity to run on a POS mobile device is a challenge. The truth is, there's nothing tougher than getting your game to run at 30-60FPS on a bad cell phone that is stable and looks good. I don't think a crappy looking unity game that runs on a decked out gaming rig has market value. It won't run on a Switch or legacy hardware, plus it can't compete with the AAA guys who actually know what to do with the hardware. If you can achieve 60 fps on mobile, you are already in the top 5% of of the market. Especially if that product is 3D. Minimalism is king.

Almost everything broke on my mobile builds and I spent 3 days stripping everything down and going into the weeds with every parameter/checkbox in Unity's rendering toolbox. Unity's documentation sucks. Period. It's better to ask GPT. Also Unity's MUSE is a long way from being useful, the interface sucks and it's too slow. Right now, I've got the choice between asking a robot that's two years behind, or using another robot that doesn't have the UX figured out, or wading through poorly indexed API documentation on Unity's website.  Software dev is going to get much easier with AI, and right now we're in this weird limbo of "kinda useful, kinda awkward and kinda reliable." This will probably get ironed out over the next year.

I was reminded of my experience shipping 2 other 3D products on mobile and how paranoid I was about resources. For Game-A-Week, I've mostly been ignoring those performance optimizations and have been creating with reckless abandon. Just getting things down was the priority. Going forward, I'm going to clamp down on my resources and start the discipline of making my builds run lean and mean.  Since the last time I shipped a mobile game,  I have a better understanding of what/why I need to do things instead of just guessing and hoping for the best. 

My rendering stack is pretty bare bones and I didn't get a chance to work on any art design. I can tune that next week

WHAT I LEARNED:

• Make simple, achievable goals that can be accomplished in a day without killing yourself or working 12 hour days. I like success and smaller, achievable goals are better for morale. If I hit the target, I can ride that momentum into other tasks outside of the minimum daily goal
• If you are making a cross-platform game, build it on mobile first- getting your Unity game to spec is difficult on that platform and once you do it, the other platforms come for free
• Don't be afraid to delete things and rip out large chunks of software that you are iffy about. Just keep a copy of the source on a harddrive for reference if you ever need it again. What matters is the production build and getting things working at the target framerate.
• Minimalism pays dividends in game development. However pre-optmization is the root of all evil. It's all about walking that tightrope.

-CREATEURS SAN FRONTIERES 

"Stay low. Stay humble."

JOIN GAME-A-WEEK:  https://discord.gg/j8QvpxzZZK