Interesting that OCaml was flexible and expressive enough to be used as a prototype testbed but not chosen as the implementation language, especially given the maturity of both. I would be surprised if Zigs incremental builds could be meaningfully faster than dune's.
Cross compilation is great, but not mentioned in the "why Zig" section. Is memory control that crucial for a compiler?
Rust itself was originally written in OCaml, same with WASM. I'm curious about what milestone gets reached where the maintainers collectively decide to transition away.
Rust moved away from OCaml when it decided to be re-written in Rust. The post alludes to this as being a usual time for a wholesale re-write, and I'd agree.
I appreciate the insight, and on closer reading the post clearly states that realistically only Zig and Rust were ever considered anyway.
Since you're here, could you comment on the approach Rust took in their rewrite? Was it more of a straight translation like Go did when they self hosted -- similar to the recent Bun transliteration? Or were there architectural changes made along the way like this article describes with Roc?
The Rust re-write happened before I got involved. If pcwalton is around and sees this comment, maybe he can provide a more first-class account.
> Was it more of a straight translation like Go did when they self hosted -- similar to the recent Bun transliteration? Or were there architectural changes made along the way like this article describes with Roc?
From what I remember, it was a whole-sale re-write from scratch, not a transliteration. While Rust took a lot of inspiration from OCaml, especially in those days, it was different enough that I'm not sure that a more direct transliteration would have been particularly possible, though again, see above, I wasn't there, so I don't know for sure.
OCaml compiler is incredibly fast. I wonder how it'd fare with Jane Street's extensions for the borrow checker etc in OxCaml, if it's good enough for their HFT I'm sure it's good enough for a new language.
I wrote a toy Scheme implementation in OCaml by using the Camplp4 preprocessor. In benchmarks, it was faster than Gambit Scheme, which compiles through C.
No. I wrote it about 10 years before GitHub existed, and it’s just a toy. All it does is transform core Scheme syntax to Ocaml syntax, converting untyped values in the usual way to a ‘Value’ sum type.
I originally thought it would be slower than the faster Scheme compilers, like Gambit, because of how naive it was, but I was surprised to find that on benchmarks, including compute-heavy ones like fractals and I/O heavy ones like web serving, it outperformed Gambit. That’s really a reflection on Ocaml though, I didn’t do anything special.
If you asked an LLM to do that today, it could probably produce something better for you pretty quickly.
I suspect this "not a systems language" alludes only to OCaml's rather steeper learning curve and until-recently difficulty with multiple threads. I am sure it could roll just fine as a single-threaded compiler language written by a small team, which indeed, it was.
OCaml has often historically been considered a language that's been appropriate to write systems tooling like compilers, runtimes, and unikernels in, even though GC'd languages were/are not often considered for such projects.
I don’t think there’s too many of us on the ‘GC did nothing wrong’ hill.
Reading the average HN opinion, it seems everybody is writing high-performance latency-sensitive systems that would implode if a response would take 1 ms longer than normal.
Sampling bias. Most of the people responding are probably those with a strong opinion because of what they work on. Everyone else is likely relatively indifferent to it.
It is a misconception that GCs only affect latency-sensitive systems. High-performance throughput-optimized systems are also sensitive at ~1µs granularity for different reasons, so GCs are not used there either.
That a GC is adverse to the performance both latency-oriented and throughput-oriented workloads doesn't leave many use cases in "high-performance" systems. Maybe systems that are severely I/O bound but is barely a thing these days.
> Maybe systems that are severely I/O bound but is barely a thing these days.
Any kind of web service is barely a thing today? Which is what 99% of HN posters are working on, hence my comment.
> High-performance throughput-optimized systems are also sensitive at ~1µs granularity for different reasons, so GCs are not used there either
Games are high-performance throughput-optimized systems that have adopted GC languages for 15+ years now, and again a type of application which is much more latency sensitive than most people deal in their day to day.
Nobody is claiming GC is a panacea, but it’s good enough for a lot more use cases people give it credit for.
If you are severely I/O bound it isn't intrinsic, it means your server is badly under-provisioned in the I/O department. Linux on a modern server can push 200 GB/s of I/O. Even if web services were engineered to a standard that could consume that much I/O, which they are not, you would have to be astonishingly wasteful to burn it all.
It is rare to be severely I/O bound because software engineered for I/O performance tends to run out of memory bandwidth first.
Games are not throughput-optimized systems in any conventional sense. They are a canonical example of latency-optimized systems.
I have nothing against GCs, I use them regularly even in performance-sensitive contexts. But too many people understate the adverse impact of GCs on performance contrary to evidence and theory.
I/O bound means waiting on I/O, which isn’t necessarily because it is slow, but because it is simply waiting on data to arrive, like a web service most of its idle time. If your client is dozens of milliseconds away, a GC pause is pretty much invisible, unless you are trying to squeeze every last request/second from a machine (instead of simply scaling horizontally)
That said, from your profile, you seem to work on a very sensitive niche that might colour your opinion, with good reason. What I am claiming is most of us are not building such strict a system.
Even in my toy hobby of OS development a GC isn’t the end of the world unless your goal is to compete with, say, Linux in a some kind of performance challenge, where in that case memory allocation might be the least of your bottlenecks.
It is also a misconception that all GC are born alike, and that don't exit languages with support for value types, stack allocation and GC free memory regions with C like pointer fun if so desired, while being mostly GC enabled.
Could you elaborate on "GCs are not used there [high-performance throughput-optimized systems]"? Are you referring to the cascading effects of tail latency on systems with high fanout?
Sophisticated throughput-optimized systems rely on deep latency-hiding. Schedulers see millions of atomic operations into the future, continuously rewriting the schedule globally to maximize locality and minimize resource contention based on real-time changes to workload, resource availability, and system behaviors.
In short, for each of the millions of in-flight operations (which might only map to a handful of user operations), it is trying to precisely optimize the concurrency, timing, and dependency sequencing such that when operations are executed every resource required is hot, uncontended, and available with high probability. When this works well it dramatically reduces the number of hidden stalls in execution. The schedule is constrained by tail latency requirements; a theoretically throughput-optimal schedule can defer execution indefinitely.
For an analytical database engine, an "atomic operation" is typically a query operation on a database page. A modern server can retire 100M ops/sec. While I am oversimplifying a bit, a 1 millisecond GC pause can blindly wreck the schedule for 100,000 operations in an unpredictable way. In these architectures we try to eliminate all context switches for the same reason which are 100x cheaper.
Practically, 1µs stall is a good heuristic for a noise floor. The schedulers have pretty wide concurrency on big systems, so the implied 100 operations are unlikely to have a dependency. Many stalls that are difficult to precisely control like cache line fills fit in here too.
If there was a GC that had a worst-case stall of 1µs then you could probably use it for these cases. Unfortunately, "low-latency" GCs tend to be more like 1000x that. I don't think there is any way of closing that gap short of putting a GC in hardware.
Real time GC as used by the military in weapons control systems, and on factory automation robots, keeping PTC and Aicas in business, people pay to use them.
TBH, physics limits how latency-sensitive weapons systems need to be and you can largely just disable the GC in these contexts. They use CPUs from the 1990s to do hypersonic terminal guidance. You don’t have to do any performance engineering for many latency-sensitive weapon systems. Could probably write it in Javascript.
For throughput-optimized systems, some of which are real-time, you never see a GC. That loss in performance is simply too large such that the computation becomes intractable. A lot of really poor systems admittedly exist but no one considers them “good”.
I think you may be confused about what "throughput-optimized" means. HFT is not throughput-optimized by definition. The LMAX link literally says it is a latency-optimized system. An optimal throughput-optimized system has unbounded worst-case latency -- the opposite of "latency-optimized".
None of those links contradict anything I wrote, I am already familiar with all of them.
> Most of the people responding are probably those with a strong opinion because of what they work on.
I can assure you that's not the case on here. The people working in truly low latency environments are not commenting on GC threads to begin with because it's a non-starter for them. For whatever reason, there is just a chunk of people that eat a lot of FUD around GCs who are working in the exact domains they thrive in.
Cross compilation is great, but not mentioned in the "why Zig" section. Is memory control that crucial for a compiler?
Rust itself was originally written in OCaml, same with WASM. I'm curious about what milestone gets reached where the maintainers collectively decide to transition away.