yeah, to me right now when I say "FP" I'm more referring to the way of thinking and the general idea of how you would structure programs vs things like Rust, Java, C, etc. I think I'm maybe too dead-set on Haskell as well for now, maybe trying things like OCaml/Scala would be better for myself

had a pepsi a little too late so now I'm nyoooom even though it's so late

I tried to read through some haskell book that I forget the name of but I honestly think it did more harm than good, because it felt like it assumed a lot about the background of the reader, and I wasn't that reader. I'm just not an academic person so it confused me :p

its too early for dog brain to be thinking about anything, let alone jvm. im go nap

*shutdown_compilation >.<

yeah, i already even have the (terribly written) code written to do this in a minecraft p2c i made some amount of time ago. im also remembering now that theres a stop_compilation value for _should_compile_new_jobs so that would "work" i guess

this wouldnt even really work in practice i dont think. too rusty with my jvm internals

*_should_compile_new_jobs = CompileBroker::stop_compilation;
good fix i think

yea, iirc removing force_inline and adding dont_inline is sufficient but its been a few years since ive actually done anything with it so im domt rember

oh for sure. i never said the idea was practical >:3. unfortunately it's also like #3 on my list of "most interesting things to implement" once i get this rust ir library written

something something overwriting the code attribute with my own vm's bytecode something something, would also probably need the verifier to not touch it and i dont know much about that internally

yea, ive done virtualization projects for .NET & native bin2bin but doing it in the jvm seems more interesting in a way. could probably get away with forcing c2 to not JIT the methods & do the mc cheat dev:tm: thing of stubbing _from_interpreted_entry. idk, have to write the actual libs first to try

i live in ohio (/neg) so i cant see dms without age verification but now i have a permanently stuck (1) notification there lmao

i feel compelled to mention this is completely useless in practice but definitely a "this would be neat :point_right::point_left:" deal

im ranting again ><. i will head myself to sleep before i turn this into a 10 part reply