After a fourth failed attempt at completely rewriting Psyia, I am officially calling it quits. It sucks, but I’ve failed to turn the initial creative experiment of GPU Cubes, built three years ago into a viable VR product. I don’t think I’ve ever really failed at anything that I’ve tried my best at before, so it’s a strange and upsetting thing – but in some ways quite liberating. If I can try my best, fail, and come out of it wiser, then there’s no reason not put my all into everything I do!
The reasons for my failure are complex and interwoven. I did my best to explain it in this twitter thread. Psyia has been holding me back through guilt, shame and anxiety for a long time. I’ve always had to do battle with occasional long bouts of depression, and that was definitely part of the emotional clusterfuck that has become Psyia.
But I’m actually feeling pretty good! Over the past two years I’ve created a huge list of achievable, interesting experiments I’d like to work on but haven’t because I felt that it was irresponsible to work on anything but Psyia. I feel like now I have the freedom to get stuck in and start creating again.
If you are one of the few people who remembers Psyia and was looking forward to it coming out, I’m sorry for letting you down. I really, really tried. But there’s some good news!
Despite never reaching a point where I considered it ready for a public release, I do have a string of working builds that I’ll upload for free on this site. I’ll upload all the builds I have over the next few days (my upload speed is terrible). The builds I recommend trying if you enjoyed the original GPU Cubes demo are:
- Psyia 2.2.0a4 – this is the most up-to-date build of Psyia I have, featuring effectively all the features and updates that came from the 2018 / 2019 rewrites. It’s obviously unfinished and buggy, but the main gameplay loop works and it contains only a few bugs that realistically prevent anyone from enjoying Psyia as I had hoped.
- Psyia 1.4.0 – This is the rewrite of Psyia that I originally was working on for release after GPU Cubes became popular on Reddit. If you ever tried Psyia at a festival or exhibition, you tried this version. It’s stable and reliable – the only reason I didn’t release it was because I got depressed and decided it wasn’t good enough.
I’ve also open-sourced both Psyia repositories (1.x and 2.x) in case anyone can learn from them. I imagine full project repositories for closed-source VR apps are pretty hard to come by, so hopefully that’s a useful resource for somebody. The repository links are here:
- Psyia – the repository for all Psyia versions 2.x. It contains lots of branches of failed experiments. Was originally created as a fork of PsyiaLegacy.
- PsyiaLegacy – the repository for all Psyia versions 1.x. It also has a branch that became the new Psyia respository. A little messy because I didn’t really know how to use Git when I started working on Psyia.
Note that I’ve had to remove some of the music in the repositories as I had permission to use it in the app, but I imagine the artist would be less keen on me publishing .mp3s on GitHub for free.
So that’s it! I’m not going to work on Psyia anymore. You can expect to see a lot more interesting creative work being posted here in the future – I have a lot to do!
Off the top of my head, I’ve got:
- Unity ECS to play with
- Creating custom render pipelines in 2019.x
- The combination of knuckles and compute shaders
- Fluid simulations
- RealSense depth cameras
- Augmented reality
- Generative music / audio
- Arduino / electronics + VR
- A secret art project I’ve been planning for 6 months…
- Tesselation shaders
- The VFX graph
- Ecosystem simulations
I’m sure if I look through the notes I’ve been writing for the past 18 months, I’d remind myself of all the other projects I’ve had on the backburner.
So, apologies for my absence, everybody. Sorry for not finishing Psyia. I’ve learned my lesson in regards to being indecisive about whether a project is a creative exploration of a product. I won’t make that mistake ever again.