One of my favorite ways to learn something new is to just dive right in and try it out. And while this is my favorite way, it’s not often what I end up actually doing. Most of the time what I prefer to do is look at a few tutorials, and follow along, while adding my own spin once I get the hang of it. That’s how I learned Photoshop in high school, and how I first learned basic web development. That said, I’ve been really trying to embrace the “Dive Right In” method for this capstone project.
Because I approached this project in that way, I’ve found myself getting stuck a lot more frequently than when I’ve followed tutorials for other new skills. Luckily, failure only stresses me out a little bit! I know that figuring out how to fix mistakes is a really amazing way to learn, and boy have I learned a lot about Unity and AR this term.
One of the things I’ve been doing to avoid getting overwhelmed when I hit a wall is to try to solve my problem one piece at a time. If you look through the commit history on some of the branches I’ve worked on, you can see exactly this.
The problem I was solving in the above branch was a bug with my iOS device, where the AR Camera Background was not rendering properly on my device, despite working for my teammates. At first, I thought it was an issue with my build settings, so I changed those. When that didn’t work, I did a lot of research and eventually found some resources that suggested upgrading the project to the Universal Render Pipeline might fix the issue, which it did! But it also caused another problem, which was that the cube we’ve been testing with was no longer rendering. A little bit more research revealed that the shaders on the cube needed to be changed to URP-compatible ones. After that fix, I’ve had no trouble with the camera.
Currently, I am working on adding an anchor to the scene and attaching it to our tracked image so that we can later attach our game to the anchor. I’ve been running into issues on both ends of the spectrum. Either the anchor does not appear at all, or it appears way more times than I want it to (once!). I was having a lot of trouble with this, so I ended up turning to ChatGPT to try and come up with alternative ways to achieve my goal.
Working with ChatGPT is interesting. I’ve seen first hand some extremely silly mistakes from it, but overall I feel like I have learned some valuable lessons in the process, both about ChatGPT and about my project!
I just built my latest attempt at getting this anchor working, using some ideas from ChatGPT. Standby while I see if it worked…
… (For the full effect, wait about 90 seconds before continuing on) …
… The anchor is not being added, so that is a little bit anticlimactic. I guess I’m going back to the drawing board!
Leave a Reply