> ML is not suitable for writing systems programs. Type inference causes no end of confusion, and most systems tasks lend themselves more easily to iteration and sequential execution rather than functional definitions. ML is just too much of a mind warp. I’m targeting a different audience than ML does.
Well, there is OCaml though!
Active Oberon aka BlueBottle (I think?) aka aos is Oberon with concurrency you can run natively or hosted and either way you can build “native” code.
It’s a good time, probably doesn’t meet all of Russ’s goals though!
Good post. Enjoyed the read. Saved the PDF for re-reading.