The systems language from the '70s that almost beat C

4 alexandratabone 6 8/13/2025, 8:07:04 AM
Back in the mid-70s, there was a systems programming language that, for a while, looked like it might become the standard for writing operating systems. It had the low-level control of C but with better memory safety, a cleaner module system, and syntax that didn’t feel like you were wrestling the compiler.

It even saw some experimental use in kernels and compiler projects. Benchmarks at the time sometimes put it ahead of C on real hardware, and it avoided some of the undefined behavior landmines that C has carried with it for decades.

So why haven’t you heard of it?

It never had the same institutional backing as C, its compiler toolchains were spotty, and by the time it was ready for prime time, universities were already standardizing their teaching around C. AT&T’s influence didn’t hurt either.

I’ve been digging through old manuals, source listings, and mailing list archives to see what made it tick and why it disappeared. I’ll share:

a couple of short code examples next to C equivalents

some notes on how it handled memory

the story of its brief moment in OS development

and what it got right that modern languages like Rust and Zig are rediscovering

It’s strange how much of what we think is “modern” was already there 40+ years ago, quietly ignored.

Comments (6)

cheaprentalyeti · 10h ago
You're not providing us any information on where to read about this further.
jleyank · 9h ago
See https://news.ycombinator.com/item?id=44886930

Being cute with multiple submissions?

jleyank · 9h ago
And the language is … ? Is this BLISS?
FrankWilhoit · 9h ago
If it is, there will be a lot to talk about. But I don't know what he means by "spotty compiler support", unless the absence of a "standard library".
FrankWilhoit · 9h ago
The whole point of BLISS was to reason from object-code optimizations to language features. The goal was to build a compiler that would emit code competitive with handwritten assembler: then ask, what kind of high-level language constructs would facilitate or impede that goal? The language was not allowed to have or do anything that would subvert the optimizer. It is at least an interesting approach and I do not know of any other language that has explicitly adopted it.
LargoLasskhyfv · 10h ago
> I’ll share:

Yah? Where?