User talk:Rainwave/Drafts/Active Replay

Apparently I have a similar language's specs saved on my Samsung Notes from 2022 but with "flip instruction" and with additional "if 0 jump to start" --ChuckEsoteric08 (talk) 14:21, 14 October 2026 (UTC)

I presume that you're talking about Flippocalypse. Yeah, actually I came up with this idea quite recently after seeing someone's (I think it was Umnik?) proposal of a maximally annoying BF derivative on Esolangs Discord. This language I can't easily prove/disprove TC and working in it looks pretty daunting (monotonicity makes it highly nontrivial). It's definitely interesting to hear that you've developed a similar idea a few years prior. --Rainwave (talk) 16:40, 20 October 2026 (UTC)

Decidability

I think this language is decidable, and more specifically a program always loops if ? ever jumps back to the beginning even once.

• The important property to note is that if a sequence consisting of ><! ends by pointing to an active cell starting from a tape of all-zeroes, it'll point to an active cell on any tape.
• Now imagine a program a?b?c? where a, b, and c all only consist of ><!, and upon starting from a tape of all-zeroes, the program jumps back on, say, the third ?.
• This means that the sequence abc, starting from a tape of all-zeroes, points to an active cell, and by the property given above, abc will always end pointing to an active cell.
• Therefore, no matter the state of the tape at the beginning of the program, the program will always jump back at or before the third ?. Therefore, the program loops.

I hope this makes sense, I'm not sure if I've written it correctly. –PkmnQ (talk) 17:25, 20 October 2026 (UTC)

I see your point. I think there's still a possibility where computation is possible despite the fact that the IP always resets after a particular ?. Imagine what happens in the second pass if the first ? now stumbles on an active cell and resets the IP. That active cell wasn't there before, so decisions can still happen this way. The downside is that most useful programs just wont halt. --Rainwave (talk) 17:59, 20 October 2026 (UTC)

Yes, and I do actually still think there is a chance the language is Turing-complete in the sense of "when the behavior goes into a certain predictable pattern, the program is considered to halt". I don't have a proof or disproof for that, though. –PkmnQ (talk) 11:47, 21 October 2026 (UTC)

That also is probably a reason why I did not upload Flippocalypse earlier. --ChuckEsoteric08 (talk) 11:16, 21 October 2026 (UTC)

This decidability argument actually doesn't apply to Flippocalypse, since its tape-modifying command flips the bit instead of setting it to active, which is the only reason why the first proposition is true. I have no clue how to approach it, though I haven't looked at it too much. –PkmnQ (talk) 11:47, 21 October 2026 (UTC)

I know this proof doesn't apply, just that it seems it can be decidable too. --ChuckEsoteric08 (talk) 13:16, 21 October 2026 (UTC)