rootrd , the use of AI gives primarily the legal risk since it might inject code stolen from licence-incompatible projects since these LLM models are learned on thousands of unknown origins. And additionally, without understanding the code use of AI may introduce problems that you will never understand and they will work like a timed bomb. And additionally if speak about myself, I don't like using of AI in any way, I want to control my code fully, and I don't want robots inject various junk. The only good use of AI is gathering of hints, but implement the final thing just by yourself. Additionally, there are LOTS of "drowned stones" might appear that needs an attention.
And ye, just my recent example how I built the iOS port. I did it fully by myself without any robots. Mostly I seeked for random code samples or documentations that tells about how each function works, etc. Most of the work I spent on implementing the proper vibration support on modern iPhones which have the Taptic Engine component. It was a pain, and the most annoying bug that it might start to buzz infinitely, and this bug still appears sometimes, and I going to find a better way to fix it. However, after overall, the done build became awesome for myself, designed properly, every icon properly formatted, almost every known behaviour adjusted to fit into the specifics of the system, etc. Even the tvOS is implemented, but it's another pain since it doesn't offers the persistent storage for user data, only freaking cache. I can say it was a very interesting and cool quest.
Try to perform the porting by yourself, and you'll be much more coller than you now! 🦊 👍️ It's no need to hurry, it's not a business with vampirish managers with burning arses from deadlines. It's a hobby project. An art! Even you will spend a week or month for a work, the result would be much more polished and stable, and you'll navigate the code better than when the case where everything a sly robot 🤖 will implement instead of you.
