Hacker News Reader

Hate to be that person, but Go developers hit it first and then implemented (a really nice in my opinion) concurrent hash map that does stuff similar to what is described in the article, but without copying the entire hash table each time: https://pkg.go.dev/sync#Map

View on HN

Concurrent maps are often easier to implement in GC languages because you can use the garbage collector for cleaning old values. Whereas without a GC you will need an alternative like RCU.

I do agree with the statement that in GC languages concurrent structures are easier to implement (and I actually quite like that about Go too), however, fundamentally, GC languages also implement GC itself somehow :), so the only real difference I guess is that for some (complex) data structures you _need_ GC to implement them cleanly, and that's what RCU does

Reply on news.ycombinator.com