Correctness-first, cross-target functional language
Correctness-first: No numeric overflow, underflow, or precision
loss. No NaN, null, nil, or
undefined. No throw, no catch,
no panic — errors are values. No stack overflow on tail,
non-tail, or even mutual recursion. This trade-off is deliberate:
correctness comes ahead of runtime performance, compile speed, and
syntactic convenience.
Cross-target: The same source compiles to Native (LLVM), JVM, CLR, WASM, or JS, and produces the same result on every one — independent of each target's quirks and built-in types. The compiler accepts only the effects supported by the chosen target and program type.
int32 counter ticking
every 10 ms overflows at 248 days and drops every GCU into failsafe.
Caught in the lab; FAA mandated power-cycles every 120 days until a
patch shipped.
int16 overflow on a 64-bit-to-16-bit
conversion.
Details: Catastrophes Caused by Programming Language Defects.
Short answer: almost nobody is going to solve them.
Maybe with
exhaustiveness enforced by the compiler.
Either — overflow,
underflow, precision loss, and division by zero are typed errors,
handled exhaustively at the call site.
Half a century of programming-language research hasn't produced another language that ships Awsum's guarantees with this breadth of compatibility and first-party tooling — entirely as free and open source software.
Recursion of any shape — tail, non-tail, mutual — compiles to a loop
on every backend, including JVM and JS where the host has no native
cross-method tail calls. Errors are values, not exceptions: no
NaN, no null, no throw, no
silent overflow. The same source produces identical results on every
supported backend — LLVM, JVM, CLR, WASM, JS — each built on open
standards and open implementations, with no proprietary component
anywhere in the chain.
The awsum-lang organization keeps the whole toolchain in
lockstep: compiler, formatter, LSP server, and every major editor
integration. Every release is coordinated across all of them — so
adopters aren't betting on community engagement or third-party
maintainer sentiment for any part of the developer experience,
production tooling, or infrastructure.
Sustaining that work as a digital commons requires funding at senior software-engineer rates.