Alex Nedelcu on Nostr: npub1yy8jn…vyfzq I don't necessarily agree. Blocking I/O, in the kernel, is still ...
npub1yy8jnmvkegndrlygssm979wf04jema23uc3384fu8nklacxfqyas5vyfzq (npub1yy8…yfzq) I don't necessarily agree.
Blocking I/O, in the kernel, is still blocking platform threads. However, I'd argue that these instances are both a temporary implementation detail and uninteresting.
Blocking I/O is now an implementation detail in the sense that future JVM versions may be able to lift such restrictions with zero changes to existing code. It's also uninteresting as we're basically talking about local file I/O, which is super fast these days.
Blocking I/O, in the kernel, is still blocking platform threads. However, I'd argue that these instances are both a temporary implementation detail and uninteresting.
Blocking I/O is now an implementation detail in the sense that future JVM versions may be able to lift such restrictions with zero changes to existing code. It's also uninteresting as we're basically talking about local file I/O, which is super fast these days.