Turing Complete Version Numbers

[edward]

http://joeyh.name/blog/entry/turing_complete_version_numbers/

[OJFord]

It makes for an interesting-sounding title, but:

> This is simply Brainfuck with operators that are legal in (Debian) version numbers kept as-is, and some numbers replacing the rest.

Is it known to be (or else what is) a minimum number of such operators for Turing completeness?

I ended the post thinking that was the more interesting question, but now that I comment, I wonder if that's even meaningful without restricting what it means to be an 'operator' - 6 and 9 from the OP are quite convoluted for example.

[greyfade → OJFord]

Yes. https://en.wikipedia.org/wiki/One_instruction_set_computer

[tromp → OJFord]

Combinatory logic [1] is universal and needs only 2 primitive combinators, S and K, as well as an application operator. So 3 operators suffice for completeness without any convoluting.

[1] https://en.wikipedia.org/wiki/Combinatory_logic

[thethirdone]

I'm not sure exactly what I was expecting, but I was expecting something more than brainfuck encoded into "(Debian) version numbers".

I think I was expecting it to be about the semantics of the sequence of numbers and turing complete metric for deciding if a sequence is valid or not.

[vortico → thethirdone]

Yeah, at the end of the article I thought "Wow, he found an example of a set containing at least 8 elements!"

[andrewshadura]

Package: node-debbundle-acorn (6.2.1+ds+~0.4.0+~4.0.0+really4.0.0+~1.0.0+~5.0.1+ds+~1.7.0+ds+~0.1.1+~0.3.1+~0.2.0+~0.1.0+~0.3.0+~0.3.0-5)

[teddyh → andrewshadura]

…in Debian experimental¹. In Debian testing it will probably have a more sane version designator, since all the versions of the package in testing, stable and unstable all have normal versions.

1. https://tracker.debian.org/pkg/acorn