Man, I loved Unity. I did serious engineering work in it, full 3D realtime industrial simulator running actual lidar packet protocols and spoofing remote PLC I/O. Our project hardware had NO IDEA...
Man, I loved Unity. I did serious engineering work in it, full 3D realtime industrial simulator running actual lidar packet protocols and spoofing remote PLC I/O. Our project hardware had NO IDEA it wasn't actually in the system, so we could test the code before the site even started production. Classically the integration test fell to the field engineering team.
This was back in 2018. I used it all the way up to spring of last year, when we were in serious discussions with their new industrial automation team. The problem I had is that I changed careers and was (am) now doing automotive automation, and they didn't have a good core radar model. Also the batched raycast command was a CPU operation with no way to leverage GPU.
There was something something DOTS, which I kept wanting to try, but man they really half-assed that release. The feature announcement at the conference looked so awesome, but then there was never any solid documentation, no good examples, and then after a few YEARS of it lingering they decided to can the feature officially to "polish" and I thought, "well there's no sense trying to learn it now if they're going to change everything."
C# is an amazing language (the best, I'll happily die on that hill) and I've been dreaming of some way to get back to Unity to do... anything. But now? If they're going to unilaterally yank the terms out from under everyone for games that get installed, what about industrial ToS? The simulators I've made only get deployed on a few machines, but they're absolute beasts and provide a ton of value.
The thing is, there's this value decision of "should I build this myself or pay someone to do it for me," and part of that is cost. I couldn't imagine sinking a few years into a project just to then have Unity decide that they're going to make it prohibitively expensive.
I don't see how anyone goes back to Unity. Ever. I wouldn't.
Man, I loved Unity. I did serious engineering work in it, full 3D realtime industrial simulator running actual lidar packet protocols and spoofing remote PLC I/O. Our project hardware had NO IDEA it wasn't actually in the system, so we could test the code before the site even started production. Classically the integration test fell to the field engineering team.
This was back in 2018. I used it all the way up to spring of last year, when we were in serious discussions with their new industrial automation team. The problem I had is that I changed careers and was (am) now doing automotive automation, and they didn't have a good core radar model. Also the batched raycast command was a CPU operation with no way to leverage GPU.
There was something something DOTS, which I kept wanting to try, but man they really half-assed that release. The feature announcement at the conference looked so awesome, but then there was never any solid documentation, no good examples, and then after a few YEARS of it lingering they decided to can the feature officially to "polish" and I thought, "well there's no sense trying to learn it now if they're going to change everything."
C# is an amazing language (the best, I'll happily die on that hill) and I've been dreaming of some way to get back to Unity to do... anything. But now? If they're going to unilaterally yank the terms out from under everyone for games that get installed, what about industrial ToS? The simulators I've made only get deployed on a few machines, but they're absolute beasts and provide a ton of value.
The thing is, there's this value decision of "should I build this myself or pay someone to do it for me," and part of that is cost. I couldn't imagine sinking a few years into a project just to then have Unity decide that they're going to make it prohibitively expensive.
I don't see how anyone goes back to Unity. Ever. I wouldn't.