Talk:Substructural type system

Latest comment: 5 days ago by 193.0.82.141

There is a note that says the article appears to contrict itself regarding whether aliases are allowed. In what way does it contradict itself?

My mistake. I didn't read the article carefully.—greenrd (talk) 09:48, 11 July 2008 (UTC)Reply

It might be good to mention C++11's unique_ptr and move semantics/rvalue references here as an example; it is better than auto_ptr because it allows more static checking. Jewillco (talk) 22:07, 20 September 2012 (UTC)Reply

Someone did it, but now it is incorrect - the C++ code in the article is not correct C++ code, as after std::move'ing from a unique_ptr, the pointer is nulled :/ 193.0.82.141 (talk) 19:15, 26 November 2024 (UTC)Reply

Links linking back the same article

edit

In the section "Different substructural type systems" there are links for each of the four substructural type systems, but they all link back to the same article (namely this one). Is that desirable? KristofferHaugsbakk (talk) 19:52, 4 January 2014 (UTC)Reply

Ordered, Stack, "Same order"?

edit

"Ordered type systems (discard exchange, contraction and weakening): Every variable is used exactly once in the order it was introduced."

"This can be used to model stack-based memory allocation"

Wouldn't that require *reverse* order?

50.204.20.34 (talk) 21:51, 20 February 2014 (UTC)Reply

C# Static Constructors

edit

Are static constructors similar to affine types? 169.244.51.180 (talk) 12:49, 18 May 2022 (UTC)Reply

Clarify when "use" only concerns transfer of ownership

edit

This nomenclature is used in discussions of lifetime semantics, where limiting how many times you can transfer ownership out of a variable binding is a meaningful distinction (say between Rust and C++), whereas how many times you can "use" a variable sounds rather esoteric. Is that a thing? The definition seems to rule out the content.Anordal (talk) 00:14, 22 November 2023 (UTC)Reply

Done: Added section about the resource interpretation (inspired by the linear logic one). As far as I know, this narrower definition is the only practically relevant of the two. I don't know if the main definition is secretly describing the same while just being vague, but I'm erring on the side of not changing the main definition.—Anordal (talk) 21:43, 27 November 2023 (UTC)Reply