> The end-game is just dissolving any distinction between compile-time and run-time This was done in the 60s/70s with FORTH and LISP to some degree, with the f

> The end-game is just dissolving any distinction between compile-time and run-time This was done in the 60s/70s with FORTH and LISP to some degree, with the former being closer to what you're referring to. FORTH programs are typically images of partial applications state that can be thought of as a pile of expanded macros and defined values/constants (though there's virtually no guardrails). That being said, I largely agree with you on several of these and think would like to take it one step further: I would like a language with 99% bounded execution time and memory usage. The last 1% is to allow for daemon-like processes that handle external events in an "endless" loop and that's it. I don't really care how restricted the language is to achieve that, I'm confident the ergonomics can be made to be pleasant to work with.