4 ms·
YT talk from Ron Pressler: "Why user-mode threads are (often) the right answer" [1] and slides [2]. 1: https://www.youtube.com/watch?v=KmMU5Y_r0Uk https://www.
by emmanueloga_ 5y ago
YT talk from Ron Pressler: "Why user-mode threads are (often) the right answer" [1] and slides [2].
1: https://www.youtube.com/watch?v=KmMU5Y_r0Uk https://www.youtube.com/watch?v=KmMU5Y_r0Uk
2: https://assets.ctfassets.net/oxjq45e8ilak/5QM86VAnN9XJ9HUIs2er5v/8240f681da10461f7d63a83429c90e4c/JPoint2021Loom.pdf https://assets.ctfassets.net/oxjq45e8ilak/5QM86VAnN9XJ9HUIs2...
- zvrba 5y agoSomeone would disagree (though in context of C++): http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2018/p1364r0.pdf http://www.open-std.org/JTC1/SC22/WG21/docs/papers/2018/p136... Perhaps it's easier to address the problems in a managed environment and I really do hope they pull it off. Also it's unclear whether virtual threads will support async file I/O out of the box, or ever. (C# does have Async methods on files.)
- pjmlp 5y agoC++ still needs to get the executors mess sorted out. Java supports executors since Java 5, and async IO exists since ages with NIO. Nowadays C++ gets lost discussing language minutiae that it isn't as much fun as it used to be.
- zvrba 5y ago> Nowadays C++ gets lost discussing language minutiae that it isn't as much fun as it used to be. Yeah, I agree. C++ is no longer fun at all.