Jump to content

Talk:Total functional programming

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Misconception Regarding "Turing Complete"

[edit]

This article repeats a common misconception regarding total programming languages: that they are not Turing complete. The question of Turing completeness with regards to total languages like Agda is subtle (it involves codata as mentioned in the article), but it is certainly not true to say that Agda is not Turing complete (yes, even with all of the checks on termination and safety turned on).

In particular, it is possible to write a Turing machine or lambda calculus interpreter in Agda, using methods like corecursion (although corecursion is not strictly necessary).

More details are available in this paper. — Preceding unsigned comment added by Doisin (talkcontribs) 02:03, 29 July 2020 (UTC)[reply]

Missing material

[edit]

I'm sorry, but without descriptions of codata and dependent type systems the article is not even worth looking at. --141.228.106.148 (talk) 18:42, 20 January 2015 (UTC)[reply]

Bad example

[edit]

There are various manners in which tooling can prove termination of quicksort and there are provers where quicksort can be shown to be trivially terminating. Some writer seems to have lifted some tool-specific behaviour to a general quality. I suggest to either clarify the problem of automated termination checking or remove the example completely. — Preceding unsigned comment added by Marco devillers (talkcontribs) 09:34, 15 August 2022 (UTC)[reply]