[corresation]

Rust seems like the answer here.

How so? To me the callout to Rust diminished the entire article because it made it almost anti-Go for no particular reason (ala "Go is poopy anyways, but maybe [some other unproven option] will be the savior]". This same sort of nonsense occurred with the statement "on my not-very-fast-at-all Core i3-2100T" regarding the C++ performance, which is just narrative nonsense given that the same circa-2012, AVX-equipped processor was what yielded his Go numbers).

The author demonstrated C code that was 2x faster than Go code run through the standard Gc chain, when the C code is run through a very mature, hyper-optimized C compiler, which is actually really impressively good for Go. They then try to pound home the point by inlining AVX which is absolutely ridiculous for such a language comparison. It borders on pure troll, and I'm really surprised that so many people are falling for this.

[theatrus2]

Both Rust and C/C++ sit squarely in the "no GC" or "optional GC" camp, which has very important implications for certain applications which strongly favor that compared to a generalized system-wide GC.

[Dewie]

> How so? To me the callout to Rust diminished the entire article because it made it almost anti-Go for no particular reason

It might seem that way if Go and Rust were competitors, but given that they seem to have pretty different philosophies and goals, that doesn't seem to be the case.