Fork me on GitHub
#clojure-europe
<
2022-04-12
>
genRaiy07:04:45

Good morning

pez07:04:34

Good morning. Trying to figure out if I can visit :clojureD in June. Wish me luck!

❤️ 7
🤞 6
1
1
thomas08:04:38

So I just got information about a potential competitor... and they claim that the technology they use is not Turing complete. Now that doesn't sound good... but what does that actually mean?

Eugen08:04:15

to my knowledge - they lack some of the things to make it a general purpose language. Presumebly they are refering to this: https://en.wikipedia.org/wiki/Turing_completeness . Whether that is a hindrance or not, I don't know, Traditional SQL is not touring complete https://stackoverflow.com/questions/900055/is-sql-or-even-tsql-turing-complete

thomas08:04:57

they link to that page as well. And as you say... in what respect is that a hindrance. does that mean that there are certain problems you can't solve?

Eugen08:04:18

yes, some problems will not be solvable. but that might not be an issue for the domain in question

Eugen08:04:46

touring complete means you could write an OS with the language (in theory).

thomas08:04:57

it claims to be a 100% declarative system, I assume that there are limitation to that at a certain point.

Eugen08:04:05

but not all languages are meant to be used that way

thomas08:04:22

true, and this one certainly isn't.

Eugen08:04:35

yes, depending on what they implement - but they could always extend it

thomas08:04:37

but what kind of problems can't be solved with a non-Turing complete system? Or does that depend on what specific things makes you non-Turing complete?

Eugen08:04:23

it depends on the specifics

Eugen08:04:39

marketing descriptions are not enough

slipset08:04:08

A feature of non-turing completeness is that there is a bunch of stuff it cannot do.

slipset08:04:53

ie you can make some guarantees about bad stuff not happening.

genRaiy09:04:37

The problem with Turing complete is that there are no limits to what can be computed.

1
genRaiy09:04:09

@U0BKWMG5B had a great talk about the various degrees of power back in 2017 at Dutch Clojure Days

borkdude09:04:55

One problem with Turing completeness is the halting problem: if you are Turing complete you cannot proof that your program will ever terminate.

4
borkdude09:04:19

Therefore things like Dhall exist, which are not Turing complete: > Dhall is not Turing-complete. Evaluation always terminates, no exceptions https://github.com/dhall-lang/dhall-lang

Ben Sless10:04:02

This is why Turing incompleteness is a feature ^

borkdude10:04:39

If that wasn't clear by now, yes :)

Ben Sless10:04:27

Isn't the halting problem solved for simply typed LC?

ordnungswidrig13:04:02

If it's turing complete you'd better use clojure anyways 😛

☝️ 1
lread13:04:26

Good morning

Michaël Salihi19:04:55

Sunny Morning from Annecy lake! 🌅 First time but not last...very nice place!

😍 3
otfrom07:04:52

nice trailer

🙂 1