schmu VS Pipefish

Since the last time I posted, I finished implementing pattern matching for schmu. To make matching on multiple columns less confusing I also added a tuple syntax to the language (finally), which are treated as anonymous records in codegen. Since then, I'm trying to overhaul my memory management, as my RAII-like solution only worked for linear code. In my first big departure from OCaml semantics, I decided to implement mutable value semantics. The paper linked in the Val language introduction makes a strong case for value semantics and after watching a couple of talks by Dave Abrahams, I wanted to try see how it feels. By making mutability be transitive and explicit, it also fixes one of the (few) gripes I have with OCaml that an array can never be really const as it is a reference type (it's possible to enforce constness with modules, but that's not exactly lightweight, syntax wise). Implementing mutable value semantics was pretty straight forward on the typing side, but I'm still not completely done with the codegen. This is due to 1. Assumptions about immutability I made in a lot of places are now wrong, and I had to completely change the way I pass values to functions. 2. I had to implement reference counted arrays, which was more work than I thought it would be. There are still edge-cases coming up in testing from time to time. Yesterday I finally managed it work for tail recursion, yay! I'm looking forward to getting rid of unneeded reference count updates in the future, by moving them to compile time, at least for linear code, lobster style. That's also an excuse to read that Perceus paper again. For the rest of November, I want to enhance my module system a bit. In particular, I want to add signatures and allow locally abstract types. I hope to have this in place before December to do the Advent of Code in my language.

I'm still working on my toy language schmu, an ML-inspired language which uses LLVM as backend.

I spent the time off over the Easter break to write the first program in my language which is not an explicit test and ended up implementing Ray Tracing In One Weekend. It was very rewarding to see how usable the language is already.

Charm is a language where Functional-Core/Imperative-Shell is the language paradigm and not just something you can choose to do in Python or Ruby or PHP or JS or your favorite lightweight dynamic language. Because of the sort of use-cases that this implies, it didn't seem suitable to write another Lisp or another ML, so I got to do some completely blank-slate design. This gives us Charm, a functional language which has no pattern-matching, no currying, no monads, no macros, no homoiconicity, nor a mathematically interesting type system — but which does have purity, referential transparency, immutability, multiple dispatch, a touch of lazy evaluation, REPL-oriented development, hotcoding, microservices … and SQL interop because everyone's going to want that.

I think it's fair to call this a language announcement because although I've been posting here about this project for a loooong time, I've finally gotten to what I'm going to call a "working prototype" as defined here. Charm has a complete core language, it has libraries and tooling, it has some new and awesome features of its own. So … welcome to Charm 0.4! Installation instructions are here. It has a language tutorial/manual/wiki, besides lots of other documentation; people who just want to dive straight in could look at the tutorial Writing an Adventure Game in Charm.

In my own language there is some syntactic flexibility but the only thing that describe pretty table could mean would be the second of the possibilities above; the first would be expressed by describe prettyTable and the third by describe PRETTY, table. This makes it more readable from the point of view of a coder, and who else is going to want to read it, my mom?

I've been trying to think of a way of doing this which is simple and consistent and which can be extended by other people, so if someone wanted to embed e.g. Prolog in Charm they could do it without any help from me.

Charm does this for declaration of local constants in functions (there are no local variables in functions). So for example if you wanted to write the Collatz function this way (which you wouldn't, it's just a minimal example) then you could do so without worrying about a computational explosion:

I guess it's time for me to put aside my exploration of Charm and set up a collaboration with my son the lyricist.

In fact that's how a lot of Charm programs end up getting written, because you want to pass a whole bundle of stuff to the functions. For example.

No, it's "shell" as in "shell of the code". The idea is that the imperative bits of the language, the bits that do the mutation of state and the IO, can can call lovely pure referentially transparent functions. But functions can't call commands (otherwise by definition they wouldn't be pure). So all your imperative-ness is reduced to about 1% of your code which lives right at the top of your call stack --- the "imperative shell" of your code. See [here](https://github.com/tim-hardcastle/Charm/blob/main/examples/adv.ch) for an example. The "imperative shell" is the main function --- all 13 lines of it --- and everything everywhere else is pure and immutable.

I'm not sure what counts as cool. It's just dogfooding at the moment. I did a bunch of other languages (only the BASIC and the Forth are up to date with the current version of the language I think), and I did a tiny adventure game (and used it as the basis for a tutorial).