News for hackers. Stuff that matters.

A comment on Zig, Skia, Clojure, Geometry and the Japanese TV Show: ICFP Contest 2021. Parent comment

Re:Zig, Skia, Clojure, Geometry and the Japanese TV Show: IC... by Arnavion on Thursday July 15, 2021 @07:49PM (#27851489)
They certainly have their problems. They're functions that get invoked without any code to explicitly invoke them. You have to consider whether a destructor can be invoked on a partially constructed object. You have to consider whether a destructor can be invoked more than once. You have to consider providing the ability to not call the destructor for performance. etc etc

It's understandable why a language trying to be close to C in spirit does not have them.

[ Parent | on HN ]
Re:Zig, Skia, Clojure, Geometry and the Japanese TV Show: IC... by ghoward on Thursday July 15, 2021 @09:33PM (#27852181)
Yes, those are all problems with destructors, but they are well understood. However, in a language I am currently creating, calling a destructor on a partial object is impossible because the object is created in "one" step. In other words, constructors themselves don't have access to the object in a partially created state.

It does make things much easier.

[ Parent | on HN ]
Re:Zig, Skia, Clojure, Geometry and the Japanese TV Show: IC... by hnakamur on Thursday July 15, 2021 @08:13PM (#27851696)
I agree. For example, most of Rust libraries uses the drop handler (something like desctructors) to close a file descriptor. However it has problems since in async model closing is also async and it may return an error. So in tokio-uring explicit async close will be provided. https://github.com/tokio-rs/tokio-uring/pull/1/files#diff-3d...

And you can call close like: file.close().await.unwrap(); https://github.com/tokio-rs/tokio-uring/pull/17/files#diff-b...

[ Parent | on HN ]
Re:Zig, Skia, Clojure, Geometry and the Japanese TV Show: IC... by pcwalton on Friday July 16, 2021 @12:08AM (#27852967)
Note that there are ideas to fix this issue in the async foundations WG. It's arguably more of a problem with async as implemented in Rust than with the concept of RAII.
[ Parent | on HN ]