I should get on with the rest of the puzzle book. I think I've got 12 puzzles left, but haven't really done anything for a week or so because I was busy with other stuff.
The compile-to-TT thread is interesting. But I feel like the GoL approach is overkill. It should be easier to find a minimalist esolang and compile that to TT. And then either you're lucky and someone already wrote up a reduction to that esolang, or you work your way through other esolangs to Brainfuck (I think there's a C to BF compiler somewhere?).
Hmm, you have a good point. I guess the question then is "What do you want to do?" because if the answer is "Tetris", then the GoL approach is the way to go given the complexity of Tetris. If you just want a primality checker, then BF will be good enough. You could say so yourself in that thread.
My point was that once you get to Brainfuck, you get the leap to "real" languages for free, so you could still implement Tetris, you'd just go through a different path of languages.
Yeah, I was planning to write something up once I've found a language that I think is suitable for translation into TT.
@MartinEnder Thinking about it more, I think that even if you just try to implement BF in TT, then you still have to meet most, if not all, of the criteria I listed. You need an instruction tape (ROM), a data tape (RAM), a way to advance to the next instruction (PC), the ability to add/subtract (small ALU), and conditional jumps.
I'm actually wondering how Paul knows that Turing Tumble is Turing Complete. I wonder if he or a friend proved it.
@El'endiaStarman I didn't realize you got married - the last time I remember seeing you mention your SO, you said she was your girlfriend. Congratulations!