Gerbil Benchmarks

[vyzo]

I spent some time writing some (optimized) Gerbil code for the now defunct language shootout.

The results are here: https://vyzo.github.io/lisp-benchmarks-game/
TL;DR We are doing really well. I started writing them thinking we are competing with Racket, but it turns out we are in an entirely different ballpark; we are really competing with C and Go.

There are some additional benchmarks from the (nowadays unmaintained) r7rs-benchmark suite, with results rendered here: https://vyzo.github.io/r7rs-benchmarks.
I find those less interesting, as they are vanilla Scheme benchmarks, not idiomatic Gerbil, and I am not really interested in optimizing that code.
Regardless, they are useful for tracking progress in future compiler optimizations, where we try to improve the performance of safe code by surgical type annotations and inference.

As always with benchmarks, take them with a grain of salt.