Would really appreciate any feedback about the syntax, docs, features that are glaringly missing, etc. Before anybody mentions it: I know the performance is pretty lousy when dealing with a lot of data. If you can believe it, the runtime is about 100x faster than it used to be! Long term I'd like to switch to a proper bytecode interpreter, but so far performance has been Good Enough for my use cases.

Thanks for taking a look!