#COBOL (*suspiro*), #Prolog (*ains*), #Simula67 (oh!), #Modula2 (ah!), #SML (+ #MoscowML y #PolyML), #APL (jijiji), #BASIC modernizao, #LISP, #C, #JULIA, ¡¡y #FORTRAN!!
#COBOL (*suspiro*), #Prolog (*ains*), #Simula67 (oh!), #Modula2 (ah!), #SML (+ #MoscowML y #PolyML), #APL (jijiji), #BASIC modernizao, #LISP, #C, #JULIA.

Io che cerco di curare la depressione bevendo un moscow mule ottimo ❣️

#habitat #leiria #portogallo #cocktails #moscowml #moscowmule

Does *anyone* actually like Moscow ML's insistence that structures live in files with identical names (e.g., `structure Foo_Fabricator` MUST live in the file `Foo_Fabricator.sml` and MUST have a signature with identical name `signature Foo_Fabricator` which itself lives in the file `Foo_Fabricator.sig`)?

I always thought it idiomatic Standard ML to use SCREAMING_SNAKE_CASE for signature names, which breaks Moscow ML's conventions (unless you put it in SCREAMING_SNAKE_CASE-sig.sml and compiled it as a toplevel unit...which kinda breaks the whole module-as-compilation-unit idea?).

And Moscow ML *hates* it when you include the signature and a structure within the *same* file, because that violates its hardcoded assumptions about the world.

And another thing...

#MoscowML #StandardML #functional_programming #naming #convention #programming

If you are trying to get HOL4 working on a Raspberry Pi, you need to use MoscowML.

Poly/ML native compilation is not really supported for Raspberry Pi, and its bytecode interpreter performs worse than MoscowML's bytecode interpreter anyways.

I learned this after spending a few frustrating days trying to get PolyML to build HOL4, fruitlessly, then gave Mosml a shot (and it worked like a charm).

The things you learn...

#HOL #MoscowML #PolyML #StandardML #RaspberryPi