What did everybody make of the benchmarks cited in the article? I couldn't really understand what they actually represented, or if I should put much stock in them. I was surprised that Rust didn't come out better in those.
A bunch of changes landed right before the beta release some weeks back, and a bunch of examples broke. We also haven't invested much time in making sure that the examples have kept up lately. We used to beat C++ in several of them, but who knows what's happened since.