Rewriting Bun in Rust
bun.com
bun.com
As an aside, I don't know why anyone would not want to use a memory-safe (and possibly race-safe) language in 2026. Rust gives you that in a performant package, so if you are turned off by GCs and immutability for performance reasons, you still have the option to use Rust.
I can understand when you need the absolute best performance and you decide to drop to down to C++, and I also relate with just personal preference, but beyond those it seems a no brainer to me.
The rust compiler is very slow. The best way to speed it up appears to be organizing a codebase in many crates. This is not preferable ergonomics to many. Beside that, for many problems, a garbage collector eliminates a large amount of defects (including the ones stated in the article) without any added friction, whereas Rust asks that you think in terms of ownership. This is not preferable ergonomics to many.
I realize what I'm saying above, while true, doesn't give a clear example. Many gamedevs would rather iterate with a language that is lower friction, not only because game code is finnicky (like frontend UI code) but because the build process can be unique. Many gamedevs prefer to iterate with hot-reloading, and asking them to use a slower compiler is asking them to accept greater latency in that cycle.
I do not claim that these reasons apply to everyone.
A "crate" in Rust is the unit of compilation. In C, a file is the unit of compilation. Rust just lets you have a compilation unit that's composed of more than one file (without having to resort to C-style textual inclusion). But if you want, you can certainly have one-file-per-crate, just like you would in C. And what's nice about having many crates is that crates forbid circular dependencies, which trivially enables coarse-grained parallelism in the build system. So yes, organizing a large codebase into crates is the best way to achieve parallelism, but that isn't something to be deplored (and strictly controlling circular dependencies is useful for comprehending large codebases in general).
In C, usually yes.
In C++, sometimes yes. It depends on how template-heavy the code is, but if you have some discipline you can keep most logic out of headers and thus easily splittable.
In Rust, almost always no, because of circular dependencies. You can try to work around it by adding `dyn Trait` everywhere, but that requires a lot of code changes and comes at big ergonomic costs (and a small runtime cost).
Which is why in practice, Rust compilation units are almost always larger than C++ or C compilation units. Rust can sometimes be competitive with C++ on compilation speed anyway, thanks to a smarter build system and not having to re-parse headers a billion times, but usually it's slower.
This feels like a strange, overly-specific complaint. It reads a bit like “When I write entangled code, it’s hard to untangle”. Like, yeah, the only thing that’ll save you from that is…not writing entangled code? I’m not of the opinion that the argument of “yeah but C lets me do whacky stuff” is a particularly strong line.
FWIW, letting a module grow, and then splitting modules up by cut-and-pasting stuff out along natural domain lines generally _is_ how I write Rust. Largely due to how easy it makes it to construct modules and submodules.
https://news.ycombinator.com/item?id=48856158
I may have been a little overly specific, since there are other issues besides cycles that also block splitting crates apart, but from what I’ve seen it’s very common for modules within a crate to have cyclic dependencies and therefore not be easily factorable into separate crates.
That said, you also run into related parallelism blockers without cycles. For example, the orphan rules force most trait impls for a type to be in the same crate as the type definition. Also, a module which has no source-level cycles will often have cycles after monomorphization. In this case, Rust doesn’t prevent you from splitting the code into crates like it does with source-level cycles, but you do lose most of the actual codegen parallelization unless you can switch from generics to trait objects.
That's not to say that you couldn't write a commercial game engine with something like C# that stands shoulder-to-shoulder with unity and unreal, but it doesn't seem like anyone has attempted to do so. Maybe it's the decompilation fear.
Also, it would continue to make sense to use a scripting language alongside Rust.
Or you could do it [as I recall the project being called] the scientist way. You still have the old code, so you could replay inputs against each and compare. Probably more realistic because uncompressed video would be a ridiculously huge dataset. This would be more resilient in the face of testing hardware and driver drift.
Historically game engines are the worst offenders when it comes to unit testing. I'm not sure if that's still the case - but that's why I erred on the side of integration tests.
I suppose rendering is a component of a game engine too though, not just physics. I don't know how to do that reliably. Even if you captured pixels, it'd be annoying. If you've ever tried doing screenshot based diffing on web you will know that slight changes in aliasing in Chrome bugger everything up. Things that should be equivalent randomly aren't but not in a way that any human would care.
I have no idea how game devs handle big refactors other than lots of manual testing.
What? I do understand that a CRUD app with little to no logic do not focus on test.
But any kind of engine.. i can not understand that it's not a priority? In general I'm not a big fan of unit test on simple websites, they do not give any real value compared to effort
But I remember working on a complex codename. It was extremely important to have close to 100% unit test in the core part. Saved me a lot.
Its primary output is pixels and sound. Those are hard to test in a reasonable way. Screenshot testing is useless in a codebase where most of the changes are about making the pixels prettier.
And then your site gets more traction, features are added, things are refactored.
And now you're looking at what used to be a simple site, still with no tests.
It's a habit, that pays off long term. You don't do it for yourself, you do it for who changes the code next. Might be you, or not. Even when you do it yourself, you can't hold all the requirements in you brain, you write those up in tests
The development practices are not exactly up to date, and game development is in no hurry to change. It doesn't help that software development wages there are not at all competitive - game development selects for passion, not skill. People who want to build robust modern codebases and people who want to build AAA games are different people. So there aren't many game devs who want to push for better test coverage.
But it's also because game engines are dealing with many, many things that are hard to test for.
You know how messy it is to test a website for "does this layout look right" or "can you navigate from A to B"? Now multiply that by complex 3D geometry. A lot of what game engines do is dealing with complex 3D geometry, where the primary verification is "does it look right" and "does this interaction feel right". Which is why game development traditionally has wide human QA, and slim unit testing.
Only now do we have software that can sort of, semi-reliably, automate testing for "does it look right".
Stride should mostly fit the bill: https://github.com/stride3d/stride
Their homepage: https://www.stride3d.net/
Even the physics engine they use is in C#: https://github.com/bepu/bepuphysics2
Not a very popular game engine (never got the attention that the likes of Godot did), but it's nice to work with! It’s up there with Stride in regards to the “indie” game engines I like, maybe alongside jMonkeyEngine (since you typically don’t see that many Java game engines either).
Edit: I mean that it's up there with Flax: https://flaxengine.com/ whoops
Stride, Flax and jMonkeyEngine are all pretty cool, not just Godot and O3DE!
[1] https://docs.rs/hot-lib-reloader/latest/hot_lib_reloader/
Such as? You can't be referring to hot reload alone because you can already do that in both C++ and C#.
C# AFAIK holds a minor place in gamedev and also has slow compile times maybe? I assume it has better support for hot reload times. But generally the performance profile isn’t there for the most demanding games even if the DX is.
C# compile times are fast. Performance is a lot better than you probably think, especially with modern .NET (not what Unity uses).
Rust’s hot reload is cross platform if I recall correctly and is always available until you disable it from the build (eg debug on, release off). It also has a better story for working around the limitations that Visual Studio just throws up on.
C++ hot reload is not a common experience in the ecosystem.
Really shouldn't be a dealbreaker when 95% of your users will be using Windows (to play games), so your most tested platform should be Windows.
> when you started it under a debugger
Hot reload isn't a replacement for scripting IMO. It will never be as good. Even hot reload in Rust would have more limitations than an actual scripting language.
In total Microsoft only holds ~12% of the total video game market share across PC and console in terms of owning the platform (i.e. Xbox + Windows). I think you're overestimating how many game developers get to use reverse debugging for gaming. In practice given the limitations for Visual Studio's implementation, I highly suspect that number is very rarely used anyway.
Likewise CERN's ROOT is quite famous, nowadays superseded by CINT, and integrated in Jupiter netbooks via Xeus.
It's not “very slow”, that's a tired meme. It's slower than it could/should, but complaining about rustc being “very slow” is a clear misrepresentation, especially when everybody seems to have been fine with tsc's historical performance for instance. It could be nice if it was faster indeed, but people claiming it's “very slow” are just showing they never worked with it.
> The best way to speed it up appears to be organizing a codebase in many crates. This is not preferable ergonomics to many.
In this context (where you don't plan on publishing you stuff on crates.io) a “crate” are just a directory at the root of your repo, the ergonomic impact is literally zero.
Rust is not compiling any slower than a comparable C++ codebase for me.
It is compiling much slower than a managed language like Kotlin, C# etc. though. Which is an unfair comparison anyway.
The problem with e.g. gamedev is the iteration cycles. It's very creative work and not so much your average engineering job (apart form game engine dev).
But again, that's an unfair comparison, because even in Unity and many if not most other game engines, there is a scripting language that is used for most of the game logic, that doesn't need hyper performance...
Still, when compared to such languages, Rust indeed does compile slowly AND one has to say: Rust is selling itself not only as a safe, blazingly fast systems programming language these days, but it also is used for its type system etc. and then, when you come from a managed language, compile times really suck.
(Although debug builds are better, they are not an option for game dev for example, because the binary is just too slow then... 200fps vs. 20 :X)
> In this context (where you don't plan on publishing you stuff on crates.io) a “crate” are just a directory at the root of your repo, the ergonomic impact is literally zero.
Is not true, you can't have circular out of crate dependencies. This often means you now need a third crate that's a trait crate, but then you can't implement external traits on external types, so you need bridge crates, and so on.
Rust's limitation of performance requiring lots of crates indeed has real impacts on projects beyond simple hello worlds or trivial cli apps.
Considering it to be a zero impact issue is rather reductive, even in the context of the language's design principles itself.
Rust for all it's good sides has had a lack of interest from core team and energy to drive real valuable changes beyond the nightly blockers into stable, or maybe they are working real hard and the boulders are so hard to move that we can't see any change looking outside in.
Is it justified after the gargantuan effort that was merging Async and GATs? Yes.
But acknowledging the problem doesn't help us solve it.
This is to say, Rust is an amazing labour of love project that seems rather stuck in time due to lack of investment/time/effort or all of the above, I am not sure, but it's moving slower than I would like, at solving the problems Rust developers face everyday.
And yes Rust compiler is slow (very slow is arguable, compared to modern C++ it isn't that bad, but compared to say Go without cgo, its horrid), Cargo is just bad, without proper hermetic builds and stuff, even when I setup sccache for our team and our cache hit rate remained below 20% and most of it was just C++ deps hitting the cache.
Just to be clear Zig builds are quite slow too, especially on windows where debug builds also use llvm.
TBH Zig debug builds on Linux also don't really feel that fast, C still compiles faster for me by a considerable margin.
Either way as someone doing Rust everyday for last 8+ years, 5+ in small/large teams, I have lots of complaints and I am sad, it has been over years of me complaining without nearly enough progress, they have a survey declare ambitions, and then well... things just don't move much.. not nearly as much as I would have expected.
Honestly given I have been a rust dev for over half a decade now, I should instead of commenting here probably be figuring out if I can contribute to Rust to help things along (faster?).
But most meetings and discussions happen at very EU/US centric times, and number of non US/European core contributors in Rust is also rather small(I don't know of one but I hope there are a few) so as someone not in those circles, I don't have the energy to figure out my way in, with my day job.
Tldr; Is Rust the language for the job here, likely. But the question should be why couldn't have been the language Bun was written from the very start. Why does Zig or C++ or C seem so much more productive.
Sorry for ranting about this but this felt a little relevant since you claimed people complaining, are likely people who have never worked with Rust.
The average cargo check for the projects I've worked on, usually finish in less than 1 second, with `cargo build` completing in a single digit second (often below 2s), it's not slow by any means.
> I believe c, Fortran, zig, C#, Java and golang are all faster compiling languages.
Sure, but the difference between type checking is 10ms and type checking in 500ms is barely noticeable for a human being anyway, despite the x50 difference.
> That makes rust pretty slow in my book.
“Slow” is a perceptual thing. It doesn't matter if it's slower in absolute benchmark performance. If it doesn't slow you down in your work it's not “slow”.
> As far as I know the only language that compiles slower is C++
Typescript's compiler is much slower than Rust's, but it's plenty fast enough for most people and you almost never see complains about it because it mostly doesn't matter outside of pissing contests.
But you have to compile Rust code to run it. You can run TypeScript code without type-checking it. That’s a massive difference in the development workflow.
The new TSC, supposedly 10x faster, will be very pleasant to have but not as much of a game-changer as you might expect. A 10x faster Rust compiler would be incredible.
And yet I'm waiting for TSC every day while almost never thinking about rustc…
> The new TSC, supposedly 10x faster, will be very pleasant to have but not as much of a game-changer as you might expect.
It will be very nice, but I don't expect it to be a game changer, tsc isn't fast but it's fast enough to get the work done, the annoyance is there but it's objectively minimal. Anything else is pointless internet language war.
> A 10x faster Rust compiler would be incredible.
For development? Not really, not for me at least. Against the endless rants about rustc's performance on HN, absolutely.
For our small codebases, maybe. The tsc-in-Go announcement had VS Code compilation go from >120 seconds to ten.
If you're talking about bundling the entire app, 120 seconds down to 10 seconds is a great saving, but other bundlers can do it much faster already. Most people don't bother using TSC to bundle their code. I can understand why Microsoft might insist on using it, but it's not typical.
For type-checking the entire app, I would see that as akin to running all the unit tests. 120 -> 10 seconds is excellent, but not something that should be a constant bottleneck in your development workflow.
I can definitely see that 10x improvement being crucial on large teams (and/or LLM swarms) where the speed of the CI queue becomes a bottleneck.
But it's not too slow for most people.
Likewise, Rust may be slow enough to be painful for some big projects that need to often rebuild from-scratch in release mode, but that niche is definitely much smaller than the size of the “rust compiles too slowly” crowd on HN.
I gave Rust classes in university a few years back, and literally none of my students complained about slow build times. At this point it's much more of an internet meme than an actual pain point for most devs.
By the way, for most apps, rhe difference from an app we'd write in class is not as nearly as big as you'd expect: most of the code lies within dependencies anyway, and you need roughly the same amount of them in both cases.
Of course there are exceptions, apps that are tens of millions of LoCs exist, but it's not the majority.
(And they tend to be naturally split in multiple crates, which is the default workaround if a rust project takes too much time to compile).
In fact tsc was slow enough that it was rewritten into a different language for version 7, with the performance of the compiled output as the primary motivation.
Additionally there are ways to have interactive code reloading, e.g. Visual Studio and Live++.
Or even a proper REPL, ROOT, CINT, Xeus.
Naturally all things that Rust could also have, only it hasn't been the focus and there are several decades to catch up.
Sure, maybe rustc's performance compares favorably to how tsc used to be, but that's not the benchmark most Rust developers (such as myself, for more than 10 years now) care about.
> a “crate” are just a directory at the root of your repo, the ergonomic impact is literally zero.
Nonsense. That's another Cargo.toml to maintain, and another place you might need to add/remove dependencies, and you have to manage the dependency tree among your sub-crates. The ergonomic impact is absolutely not literally zero, and I'd even say it's enough to be annoying.
It was very slow. It's gotten a lot faster over time (over 2x faster). It's still not exactly fast, but it's definitely faster than C++. Although C++'s slow compile times are often complained about they were never really enough to stop most people using it, including for games.
> a garbage collector eliminates a large amount of defects (including the ones stated in the article) without any added friction
I'd be careful about that "without any added friction". Rust's lifetime/borrowing system tends to lead to less buggy code because it encourages structuring code in a less spaghetti way. GC does eliminate memory errors but you also lose that non-spaghetti code structure.
You are seldom compiling the world from scratch.
Especially in the platforms dear to game devs.
True in Rust too though - you're normally doing incremental compiles. And even though you're compiling the world, it's still on par with C++. For example I just tried compiling a hello world Bevy project, which has 462 crate dependencies - pretty big. It took 2m20. Totally reasonable to build an entire game engine and all it's dependencies.
I wish they had reported the compile time of Bun before/after the Rust port - that would have been very interesting.
Same applies to most commercial libraries, where code is usually provided for debugging purposes, not for building everything from scratch.
Also incremental linking on Rust, or hot code reloading is still not something that works out of the box.
How beefy is that machine to achieve 2m20? I started using C++ on MS-DOS, on a 20 MHz 386SX PC with 2MB RAM and 20 MB HDD.
Especially relevant in current times.
Ryzen 9 3950X (16 core). I think it was top of the line 5 or 6 years ago but it only cost me about £300 second hand for the whole PC (with 128 GB RAM - before the RAMpocalypse!) Definitely not a crazy machine.
> I started using C++ on MS-DOS, on a 20 MHz 386SX PC with 2MB RAM and 20 MB HDD
Yeah, I mean early C++ was quite a lot simpler than modern C++23! Of course it's going to compile way faster. You're not compiling UE 5 on that machine! Compilers didn't have so many slow optimisations back then either.
While I could comfortably still develop C++ in a Asus 1215B netbook with 8 GB and SSD, thanks to all the native libraries I could install, and make use via pkgconfig, the same could not be said to play around with Rust on the same netbook.
This matters, because not everyone can afford top of the line desktops, especially in 2nd and 3rd level countries, which then plays a role in language adoption.
Some even make do with what they can run on tablets.
I think the biggest issue is actually disk space. C++ isn't exactly efficient with disk space but Rust is on another level. You're going to run out of disk space on that netbook it you have any more than like 2 Rust projects built.
https://github.com/pjmlp/gwc-rs
The original C++ versions was done in 2 minutes, because it spent zero minutes compiling all dependencies.
Needlees to say that first compilation had an impact on battery charge as well.
Naturally on such a beefy machine it will fly given how fast Bevy gets compiled.
Languages with garbage collection are generally considered "memory safe". GP was talking about choosing a language that requires manual memory management, but doesn't have something like rust's lifetimes to catch things like use-after-free.
Rust is pushed by many as the replacement to C, because of the memory safety guarantees. I'm sympathetic. I worked with Haskell for a time, so I get it. But Rust seems quite complex. There are so many language features that there's memes about it. There's also the friction and learning curve.
So, for fun, I choose zig because, like C, I can hold most of the language in my head and "just write." I choose zig because it does a great deal to help me write correct and highly performant code. I can use arena allocators and defer and cure my code of many memory issues. Then there's the various language rules around pointers (optionals, slices, etc) that help me write correct code. There's the built in testing and the test allocator. I love that comptime and the build system are not special cases, but rather are just garden variety zig. I love the simplicity and elegance of it all.
I also choose zig because I prefer the liberty it affords me. I am responsible for each and every allocation. It appeals to my libertarian sensiblities.
Like many memes, these are misleading. Rust is a solidly medium-sized language; smaller than Python, certainly, though with a perilously steeper learning curve than Python.
The Rust stdlib has a lot of essential low-level types needed for adding a 'semantic layer' on top of the language so that the language user can exactly 'express intent' (types that arguably should be language features instead). Just look at all those detailed methods needed to make RefCell work, and what does 'into_inner' or 'undo_leak' even mean?
https://doc.rust-lang.org/std/cell/struct.RefCell.html
E.g. what's a single concept in C (e.g. "the pointer") easily has a dozen specialized equivalents in Rust, just because Rust needs the additional information to do its memory safety magic correctly.
The entirety of Rust and its stdlib has a huge 'semantic surface' compared to most other languages (even C++), and I think this difference in the semantic surface size to other languages is exactly the one thing that either attracts or repels people to/from Rust ;)
That "the pointer" is a single concept in C is the root cause of absurd numbers of bugs over the history of C.
The questions of how long the memory a pointer points to lives, what threads it is safe to access from, and how it is allocated and destroyed all still exist in C. Those answers are just implicit rather than explicit like they are in Rust. It's not like Python/etc where you don't have to worry about these things.
I prefer a middle ground between strong and weak typing (but yeah, C is arguably too weakly typed, but the exact sweet spot is different for every programmer).
A reasonably smart C compiler that has visibility on all code in a project could theoretically infer a lot of the same semantic details that must be manually provided in Rust by looking at how the pointer is used in the wider codebase (basically what static analyzers are doing by reconstructing control flows). Of course such an extensive analysis would result in even worse compile times than Rust ;)
I bet $4.20 you didn’t write C. You wrote something C-like which the compiler didn’t reject because the C standard has a gigantic surface of ‘undefined behavior’ which means once your program does one thing out of spec it isn’t C anymore silently.
What I am trying to express with the statement you refer to is that zig appeals to me on an aesthetic level. It suits my tastes.
I mentioned liberty because I appreciate that zig allows me to have complete control of memory allocations, and gives me quite a deal of control over the metal. I like this and I like that I don't have to appease a borrow checker. Zig's compiler and type system is strict enough to guide me to correctness, but the experience, broadly, is not so onerous that programming becomes tedious and frustrating.
These qualities, I suspect, resonate with my individualistic tendencies.
Your explanation lands otherwise. I can't relate to "I don't have to appease a borrow checker" though. That reads like a deliberately uncharitable read of what Rust's compiler is doing. That being said, it's definitely not for everyone.
I tried Zig. Loved it. But then after a few weekends I said to myself: "Wait a bit. I've been down this path before with C and C++. I know exactly where it ends." -- and bailed.
To me these days low-level programming is basically: either go all the way, or don't go at all. Sadly Rust is not like, going really all the way, but it's IMO the closest we got. And before somebody starts enumerating obscure languages that barely have a community: I include multi-axial evaluation here i.e. PL features, stdlib, ecosystem, richness of education (Rust has _a lot_ of good books and courses), and others I am surely forgetting at the moment.
Zig is great. But I'll not take marginal improvements anymore. Not in my personal life and not in my work. I'll get big ones, or not get them.
Since we are explaining philosophies, now without political undertones: that's mine.
Khronos, Open Group, NVidia, Microsoft, Sony, Nintendo... aren't going to change their APIs and SDKs, just because of social media discussions on the merits of C, C++ vs other safer alternatives.
I agree we should minimise their use, however not everyone accepts a dual language approach, nor there are alternatives in such domains, even if they technically exist, you still need to overcome the human and political factors.
Hence why it is so relevant to fix C and C++ security flaws to some extent as well.
Though that’s one nice thing about Swift: it has a very good interop’ with C now, and a “starting to get pretty good” interop’ with C++. So that can help, sometimes. (Obviously, I reiterate, I understand there are situations where the choice is just not possible, and enhancing C and C++ is indeed a good thing.)
if you read the article carefully, jarred is pretty clear about how their specific requirements with Bun cause friction when bridging the manual memory management of Zig with a garbage collected JS runtime. at face value, that makes quite a lot of sense to me, and it's a pretty specific scenario that is not the full on condemnation of memory unsafe languages that your comment is.
All that said, I use rust for everything.
If I'm vibe coding something I'm always just going to do it in Rust.
I imagine writing plain js or python with it would be much much riskier.
https://github.com/tormol/tiny-rust-executable
This produces a 137 byte binary. Obviously AMD64 isn't used in embedded, but I've seen ARM ones that are in the ~256 range.
It's all in how you use it. Of course, if you don't care about binary sizes, they can get large, but that's very different than actually paying attention to what you're doing.
First of all, I was amazed by how much I could do with Rust (safe Rust, even), and how well it was interfacing with my handwritten RISC-V assembly. I will definitely use Rust again for the next such project.
But, every time my functions would get over a certain size, suddenly some optimizations stopped working, and Rust was trying to put the whole panic/fmt machinery into the thing, going above my linker's flash size limit. It was insanely frustrating, since there was no rhyme or reason to it. Simply adding another branch to a match made it do that. Or another if statement that was exactly the same as the 4 before it.
The 137 binary thing does not scale.
The problem comes from a combination of three things:
1. Rust's standard library (including core) likes to panic a lot, e.g. for code that is genuinely unreachable (and this is fine)
2. Rust/LLVM cannot always optimize unreachable code away (and this is also fine)
3. The presence of at least one panic causes Rust to also include the error string format machinery, which is HUGE (and this is completely unavoidable)
Because of #2, there is no way to prevent this for larger code. I can define my own panic handler, but I can't prevent panic!() calls from inside core from constructing the format message using core::fmt before they call my panic handler.
And so any non-trivial code sometimes just randomly becomes 3x larger, and I can no longer fit it on my MCU, despite never actually printing or caring about any panic messages.
What you're talking about is the ease of which you can not include the panic machinery. I agree that it is not always easy to keep out, and that I would like if it were to be made easier.
But people are doing real, commercial projects on MCUs your size and smaller.
I'm not using panics, and yet I'm getting the machinery. That is where my frustration is coming from.
Mikroe has C, Pascal and BASIC compilers for all kinds of microcontrollers, that is their main market.
How niche are we talking? Rust is deployed on a bunch of popular microcontrollers at this point
Rust is just as fast as C++.
I write in rust and c++, rust isn't as fast. Rust is easier to work with and, compared to the Java crap it's replacing at my work, it's a lot better but it's certainly not zero cost abstractions the way c++ can be, nor is it great for data oriented design because you're hoping the compiler will do the right thing, consistently.
In Rust you are constantly fighting the stdlib and other libraries, and you have to litter your hot code with unsafe blocks to get it to stop adding a branch to nearly every object access, be it for bounds checks or over/underflow checks.
C++ does a much better job at giving you a zero abstraction API, and you can always drop down to raw pointers if you want, without(!!!!) unsafe blocks and weird tricks. Of course it's unsafe in C++ but the friction to writing a branchless hot loop is muuuuch smaller.
When profiling and optimizing Rust code, I very often find myself poring over the generated code, making small changes, reading api docs, and trying again, much more than in C++. Lots of unsafe Rust APIs are not even nearly good enough, even with most checks turned off you will find branches that just branch to panic!(), which is, you guessed it, still more code and a branch than the code would suggest.
I get why people think that most systems languages are the same "speed", but they really are not if you are hitting limits of the hardware in your hot loops.
You couldn't have come up with something more incomprehensible.
If 99% of your code doesn't use unsafe, why contaminate 100% of your code base with footguns?
> How is not having to mark your unsafe code as unsafe a good thing?
The problem with unsafe code in Rust is that IIRC nobody actually figured out yet the "rules" of unsafe i.e. which invariants you can stretch and which can cause UB. My (not super up to date) understanding is that this is an active area of research and progress is being made and also that in practice there are many well understood usages.
In short unsafe rust is somewhat worse than C++ as the boundaries of UB are less well understood/defined
For better or worse they are much more well understood in C as they are front and center of the language semantics
Integer overflows are not checked in release builds by default, since they are not related to memory safety.
On the other hand, rust emits noalias everywhere, which helps in autovectorization.
Most developers I have met don't know how to use a profiler, so not sure any of this discussion matters at all anyway, but I feel like it's important to note when the default in Rust is safe, whereas the default in C++ is fast (and more or less unsafe)
Rust had the opportunity to iterate on C++’s default choices and improve on them in many cases. I’m not convinced that naive C++ is particularly efficient.
(That said, lots of rust beginners make heavy use of Box and clone() and write inefficient programs that way. It’s hard to actually measure average, beginner code.)
There are so many situations where something is guaranteed to be safe but there is no way to express that in the Rust typesystem, so the only thing you can do is to wrap everything in Arcs and Mutexes, which introduces allocations, pointerchasing and locks
Unless maybe you mean tokio’s work stealing executor, but you can just not use it.
Going from Rust to C++ seems a strange choice, since you get most of the same problems just without memory safety. Zig, Odin, C3 or even plain old C though? At least those languages have things to offer that neither Rust nor C++ provide (and if it's just compilation speed).
LLVM is not going to take PRs written in Rust to fix the Rust compiler backend, for example.
Neither are OpenJDK, .NET, V8, JSCore going to take such PRs.
By the way, LLVM and GCC would also not take PRs written in C.
As for PRs: well yeah of course, when in Rome...
Same applies to C++ with CUDA, SYSCL, HPC/HFT frameworks, LLVM/GCC, AI compilers,...
They don't look dead to me, even if we as industry could be much better by now.
Could you help me understand with an example or two? My understanding is that well written Rust and C++ are often identical in performance thanks to relying on the same compiler backend (both clang and rustc use LLVM).
What? Rust generally doesn't have worse performance to C++, so this argument makes no sense at all to me.
> and I also relate with just personal preference, but beyond those it seems a no brainer to me.
That's another argument altogether.
It's totally fine to have preferences and decide to go with them of course.
Rust isn't perfect but it solves a lot of the pitfalls of C++ (not just UB, package management, horrible cmake files, linker errors etc.)
Isn't ironic for a project that successfully killed all past and future human involvement?
I would love that Java and .NET would provide all layers like several managed languages in the 90's,
However it has taken a quarter century to get back features we already had in Modula-3, Mesa, Oberon and co.
For two, there are plenty of reasons to use C++ unfortunately: compatibility with existing code based and availability of developers to name two.
Forgetting all the predictions about singularity etc, at the very least AI as it is now, is going to make it very hard to justify hiring a SWE for 200k. I will say, at the very top for a software heavy company like Google or Anthropic, they will still hire excellent engineers to create new software that AI is not very good at.
But for companies where software is simply a cost center. Like Walmart, or Target, companies that were already outsourcing software development, or using cheap H1bs, now they have the alternative of AI which is much better than even hiring an average software engineer for 200k. This is a sea change in the job market, it’s going to have a pretty big effect as it is right now. US has around 1.6 Million software developers, this number is going to get cut drastically, the very top, say an L6 quality in FAANG will be fine, the average in a no name Bank, or the guy building the website for McDonalds is out, he needs to learn something else or he’ll end up without a job soon.
I would not have predicted this a year ago, now it seems clear that this will happen. Just shows how much of a sea change we have witnessed just like that.
I suspect rather than hire less people we will just produce more code changes.
> I suspect rather than hire less people we will just produce more code changes.
Why? Towards what end? Code changes are output, not outcome. It also needs to be connected to someone willing to pay you hard cash. That is the hard part, a race to the bottom, and the reason I also believe there will be downwards pressure on salaries and even employment.
I largely agree with your comment, but is this sentence typo'd or something? "Rewrite it in Rust" happens so often that it's become a meme (and that was so before the rise of agentic coding).
Maybe you meant to say that nobody would have entertained the idea of rewriting this project in Rust?
All that really proves is that you’d be an astoundingly poor choice to hire. If you’re spending $200k on someone that doesn’t know at least two out of three (context, Rust, or Zig), you’re just burning money.
That’s not to say that experienced engineers familiar with the stack would or wouldn’t be able to do it in a year, but they’d certainly have a better shot at it.
It’s also not that this project sprung into thin air from a quick prompt and LLM magic… it was driven by a dedicated, highly talented, subject matter expert with extensive SWE background and extensive support from the leading experts in the world. You’ll continue to need someone to steer the ship, even in the Wal-Marts and Targets. An LLM is only ever as good as the input it’s given.
There's a few significant facts here:
- They had an existing functional Zig implementation
- They had an existing test suite for the Zip implementation
- They had a separate JavaScript compliance test suite with ~ 1 million tests
- The person overseeing the rewrite was responsible for a huge portion of the existing codebase and was very familiar with the existing architecture and problems
I don't think that middle management at most companies is going to be starting from that same point when it comes to building or updating something. Generally, I don't think there are many projects out there that have such robust existing tests and specifications.
In this case, the engineering behind the tests and specifications need to also be considered part of the process, since without those you wouldn't be able to build a control loop in the same way.
Also I'm pretty sure Walmart directly hires software engineers and doesn't just outsource everything - https://careers.walmart.com/us/en/results?searchQuery=softwa...
and btw, if it can write code, it can write tests and test suite
Ai has come a long way, 2 years ago, this task would be impossible ! 1 year ago, it would hit a dead-end in the first hours of the execution
This blog post is nothing short of an amazing and fascinating tale, yet in some aspect very very scary ...
It would help if you actually engaged at all with any of the points they just made. They were good points!
We have known for decades that the hard part in programming isn't programming, it's making sure you have the right target! Bad designs and bad specifications have caused more bugs and broken software than anything else, and the time taken to develop the testing and validation suites (that is - the javascript spec tests) this rewrite used must be considered part of the engineering effort.
> and btw, if it can write code, it can write tests and test suite
You must be very early in your career. Again, the hard part is not writing the test, it's knowing which tests to write.
As for your other arguments, I’m not certain we won’t just Jevon's Paradox into more work.
Also it is debatable they got any value at all from this. Anyone who wrote unsafe rust and also wrote zig would know that unsafe rust is much much more unsafe in comparison
> At the time of writing, about 4% of Bun's Rust code sits inside an unsafe block (~13,000 unsafe keywords across ~27,000 lines / ~780,000 lines), and 78% of those blocks are a single line — a pointer that came from C++, or one call into a C library.
I don't know much about Rust but I imagine this is safer than 100% 'unsafe' code in Zig or C++.
Not sure if they have 100k or a million lines of code
Not only is the time and dollar spent lower than a lot of people expected. We could now foresee a lot of these human interaction, mistakes, time and cost could be further reduced by a factor of 2, 5 or even 10+ in the not far future.
Also worth taking into account what is stated in the blog post is also acting as PR piece for Claude and LLM in general.
jevon's paradox just says people will consume more software as software gets cheaper, which is likely true, not that there will be more high paid jobs for programmers.
as programmers are not the resource, the program is the resource, programmers were the means of production of said resource.
stated differently: in a world where you dont need programmers to make programs, there is infinity of programs, and no jobs for programmers.
A more realistic scenario is that programmers are still needed in the foreseeable future (for some definition a what a programmer is, even if it's "understands software enough to be able to write good prompts and judge the results"). So the question is how the productivity boost compares to the increase in the amount of software being made.
experiencing the gpt3->4o->sonnet4->opus4.6->fable trajectory I am fairly confident what we call programmer today will not exist, in the same time regular people will finally be able to tell computers what to do and this will unlock the next stage of complexity in manufacturing/material science/medicine and so on.
of course I am likely wrong, people say: its never as good or as bad as you think.
While there are a lot of resources being poured into improving and integrating tooling and use cases and increasing model sizes, the leap between the models you notes are not as much as you think it is. It's the same tech that's scalable to a certain limit of diminishing return. It's a recurrent pattern in technology.
> yes, I assume we will have the 'replicator' from Star Trek, and programs will just appear as you want them.
Yes, perhaps one day. Probably no other "job" as they are today will have the same shape and form. Cancer will be cured and fission will be solved. We may even crack teleporting. Let's hope these all happen and the outcome is not detrimental the livelihood of humans.
of course, my point was that gpt3->4o was smaller jump than opus 4.6->fable
Ai won't work on it's own until we get agi.
Writing code is only one part of the job. I use Claude Code every day and I love it. It has made me much more productive. But I still have to guide it carefully, review every change, and fix the bugs and poor design decisions it introduces.
Personally, I'm looking forward to retirement. I expect there will be no shortage of consulting work helping companies clean up the AI-generated mess generated by inexperienced developers and overconfident middle manager with zero software development experience :)
Sure, if you're wasting money in silicon valley. They could have hired in Europe and got three people for $300k, which is only double what they spent.
I think the time is the really significant factor, not the money. I bet if they had the option of paying $300k to have it done by humans rather than AI, but magically in a week instead of a year, they would have gone for that instead. $300k is nothing to Anthropic.
It won't be possible to HIRE anyone doing this job confidently, they won't even know if they f*ked up.
I am not too worried anyway, one person with no context, with or without AI, can't do this job. Just like with all the AIs you have, giving you 10 millions, you still can't build a AAA game, it is that simple.
And no, since they don't know they don't know about it, there are no signs of fixes around these real issues. They had posted something about the unsafe usage, but lol most of the UBs are casually swept away (not even treated as unsoundness), if you actually look at the code and the summary.
- They've written two articles, one by claude doing the migration another by the human behind it, neither acknowledge these problems. Nor does the incorrect SAFETY comments in code.
- I only spent ten minutes and found two code patterns that are wrong with little exceptions. Also randomly checked a few samples and there was no exception. This basically means that failure patterns not explicitly picked and disallowed will repeat themselves over and over. Try imagine the more subtle ones hiding in this huge codebase.
- Trivial to find != Trivial to fix. 107 of the problems (one of the categories) are borrow issue and requires large refactoring (rearrange and rethink whole modules of code) to be removed completely.
So yes, huge regression by my standards. And I'm not questioning Rust the language. I'm questioning Claude, Jarred the person, their dev & marketing practice and the Bun project.
Btw these categories are separate from the (still unfixed) category of problems I mentioned 54 days ago in the comments before.
From my perspective.
1. 128 known bugs were fixed.
2. 237 potential bugs were made grep'able.
That is a massive win to me.
That's not the correct understanding. These are specifically the ones that were not bugs in Zig. They are forbidden in Rust because of reference invariants that don't even exist in Zig, so correct code are made incorrect during the port. That's what I meant by new problems were created and unique to unsafe Rust in my previous comments, if that wasn't clear enough.
Again I'm not sure if you are intentionally ignoring the context. Please clarify that.
I'm not sure that this is true - who says that the same invariants weren't broken on the Zig side? (Aliased pointers are UB either way, for example). Also, "unsafe invariant broken" does not strictly mean that there's a reachable bug. It should be cleaned up though, obviously. But in theory, callers may never actually trigger a bad invariant in an unsafe block.
Regardless, even assuming that these are genuinely new bugs, it doesn't change my point. Going from ~X known bugs to ~2X potential, grep'able bugs, is a win to me. Sounds like it isn't to you. Great.
> Again I'm not sure if you are intentionally ignoring the context. Please clarify that.
I'm obviously not intentionally ignoring anything, if you think I am then you can stop responding.
> ~2X potential
at this point. That's what I've found in 10 minutes. And they are actual bugs that I've take extra time to (statistically) verify. Please worship me as the new machine god if you think I've found all the bugs with a glance.
And I'm not going to debunk each of your false claims, when you are doing that at the expense of appearing either dishonest or ignorant. I'm allergic to apologism.
Additionally I'd admit that I've intentionally avoided talking about the specifics of these problems, giving links to each of them or something. That's for a concrete reason - I can't bear it if Bun gets (even in the slightest bit) better because of my contribution. It would be one of the worst things I can do, morally.
I'm confused. You said that you've just grepped and found some undefined behavior. What do you mean that you've statistically verified them?
> And I'm not going to debunk each of your false claims, when you are doing that at the expense of appearing either dishonest or ignorant. I'm allergic to apologism.
My false claims? Like that pointer aliasing is undefined behavior in both languages? lol okay?
> Additionally I'd admit that I've intentionally avoided talking about the specifics of these problems, giving links to each of them or something. That's for a concrete reason - I can't bear it if Bun gets (even in the slightest bit) better because of my contribution. It would be one of the worst things I can do, morally.
I'm taking all of your claims at face value, I haven't said that you have to provide evidence of them. I'm not saying that your claims are false, I'm saying that I can grant them and still disagree with you.
I don't know what to tell you. Even if you said that those 2x bugs were 100% verified, I'd still say that this is a win because you can grep for these bugs now. That is obviously better to me.
> pointer aliasing is undefined behavior in both languages
That included and also basically every other claim in that comment. I'd be happy to learn otherwise if you have concrete evidence.
[Edit: erased information that can help people find the bugs without prior knowledge]
> I haven't said that you have to provide evidence of them
Sure, but if I'm going to it would be a lot easier to convince everyone how bad it is (just like I did in private message groups). I can't take the risk that "everyone" also includes Jarred though.
> you can grep for these bugs now
You can grep for the greppable new bugs. I raised them as an argument because these issues are glaring. What about the ungreppable ones? More than half (again, randomly sampled) unsafe usage in the 10,413 unsafe blocks in current Bun looks iffy to me. I just haven't taken the time to verify them because many requires reasoning across the code base (see? the exact problem Rust ought to prevent). From a Rust point of view it's a huge failure.
All these while considering that the Bun agents (no people at this point I'm afraid) don't even know how to grep for the easy ones.
Please READ. Including the one I've added at slightly higher layers of replies.
> who says that the same invariants weren't broken on the Zig side
There are no such invariants in Zig.
> "unsafe invariant broken" does not strictly mean that there's a reachable bug.
Broken invariant literally means you are reaching it. The word you are looking for is "unsoundness".
> callers may never actually trigger a bad invariant in an unsafe block.
There's a neat class of UBs called "immediate UB"s. They can ruin the code just by being compiled, instead of reached at runtime. And guess what? They have a lot of these.
> potential, grep'able bugs
We are indeed going in circles but I have to ask: you still don't understand that? Grep works for violated lints, unsafe blocks and all things explicitly exist in the code. Not for unsoundness in safe functions that can be called incorrectly and ruin everything, can't be reasoned about by any existing tools. They also have thousands of those. It's unfair to say the unsoundness is freshly introduced by the port, but Rust also doesn't help in this way where no unsafety is encapsulated and calling anything can be a wild ride.
Zig does have invariants. It doesn't have an aliasing invariant like rust though, but it does have its own invariants, naturally. Zig is a memory unsafe language.
> Broken invariant literally means you are reaching it. The word you are looking for is "unsoundness".
A reachable bug means one that can be triggered. Undefined behavior does not mean a bug can be triggered.
> There's a neat class of UBs called "immediate UB"s. They can ruin the code just by being compiled, instead of reached at runtime. And guess what? They have a lot of these.
Can be, yes.
> Not for unsoundness in safe functions that can be called incorrectly and ruin everything, can't be reasoned about by any existing tools
This is only possible if unsafe is used. `unsafe` is grep'able, hence all memory safety bugs are grep'able. That's the whole point of rust.
Not these in particular. You are again ignoring the context.
> Undefined behavior does not mean a bug can be triggered.
Undefined behavior are bugs by themselves. Let's check if you are holding any of the wrong assumptions:
https://predr.ag/blog/falsehoods-programmers-believe-about-u...
> `unsafe` is grep'able, hence all memory safety bugs are grep'able.
Wrong implicit premise: "all memory safety bugs resides in unsafe functions/blocks".
No. Even assuming everything is done idiomatically, memory safety bugs can still originate from unsafe blocks and all the safe functions in the same module. So we want the module containing unsafe to be small and closed. Otoh with this code base unsafe lives in every module.
Making them encapsulated is the whole point of Rust, in this code nothing is encapsulated so no, they missed the whole point of it.
Aaand it's not like you can just remove the unsafe blocks one by one trivially. It's as hard as rewriting the codebase from ground up without reference (unlike the current migration) as every referenced object in the call graph needs to be reordered. Well partly because the original code sucks I assume.
What context? I just said afterwards that aliasing is not one of those cases.
> Undefined behavior are bugs by themselves. Let's check if you are holding any of the wrong assumptions:
Nope and your post wouldn't justify the statement. Not all UB leads to a reachable bug. The blog post doesn't say otherwise, and in fact it indicates exactly this (that UB can compile "correctly" one day and "incorrectly" the next). Check the errata on the post.
> Wrong implicit premise: "all memory safety bugs resides in unsafe functions/blocks".
No, I'm right other than l-unsound. Saying "but safe code that references unsafe code" doesn't change that.
> It's as hard as rewriting the codebase from ground up without reference
Not necessarily. At least multiple cases I've seen in the rewrite would be a matter of just moving `unsafe` from one spot to another and then placing the assertions at the caller.
This conversation is increasingly dumb. I've made my point repeatedly - moving to Rust means memory unsafety violations can be grepped for and that's not up for debate. You can say that that's not sufficient as a win for the project, I really am ambivalent.
That's untrue. It has already been maximally dumb since your first comment with the various logical fallacies you managed to carry out in a single sentence. The only reason I'm still making my replies is that you are making so many false claims that may mislead someone else.
https://doc.rust-lang.org/nomicon/working-with-unsafe.html
I'm not even trying to prove any point since the beginning. Can't do that when you have a different definition of bug from others.
Please if you are going to reply to this ever again try stop making any claims. Otherwise I'm afraid that it will continue into infinity.
I just haven’t found another language that just makes sense. Zig doesn’t hide anything from you
[1] For example, as a random sample, https://github.com/oven-sh/bun/blob/bun-v1.3.14/src/css/medi... -> https://github.com/oven-sh/bun/blob/4924862cffbf671792d47c92...
> I rewrote Bun in Rust using about 50 dynamic workflows in Claude Code run continuously over the course of 11 days.
> Excluding comments, Bun is 535,496 lines of Zig.
> How do you review a PR with +1 million lines added? How do you start to build the confidence needed to responsibly merge large quantities of LLM-authored code? A language-independent test suite with a million assertions, adversarial code review and when something does go wrong, fixing the process that generates the code instead of hand-fixing the code.
That’s vibe coding. This blog post is an ad for Claude, nothing more.
The code was generated by a LLM, and the output wasn't even read by its user. That's definitely vibe coding.
Translation does seem to be a strength of LLMs, and as they said in the post, the code at the function-level all still feels familiar to the team. They've also already moved users to the codebase without anyone noticing; that's a better result than typical vibe coding.
I'm not trying to spin up some kind of conspiracy theory here, but I'm not sure to what extent Anthropic does have any vested interest in this project (in fiscal terms at least) because the reputational fallout could be significantly delayed and might just not be big enough to matter.
That's a very big reason not to screw this up.
This is not an assertion you are qualified to make
> Go read any analysis of the Claude Code leak for proof
You seem to be implying that Claude code is unmaintainable. Yet they appear to be maintaining it just fine. Did I misunderstand your implication?
And the fact is that they are maintaining it and it is one of the most successful software products of all time and is earning them mountains of cash. By any metric it is a successful product. So obviously whatever they are doing is working.
They rewrote the entire thing with extensive LLM use.
It is abundantly clear that their idea of maintainable and yours probably don't match up.
It's apparently out there, shipped in the real world, with people saying it's good. I think it's a pretty clear win for them.
There are different definitions of "maintainable". It seems the prevailing culture at Anthropic is to not touch code manually anymore. So for them "maintainable" means the LLM can fix it.
Did you compare the code before/after? It's a mechanical line-by-line port, and most of the code is identical to the old version, just with Rust syntax. They have an example in the blog post.
Not sure why people use it.
That's because closely following PL theory isn't always the right goal for a language. Ergonomics and pragmatism are far more important.
Though ignoring all the PL lessons for the sake of "pragmatism" is just plain ignorance.
Rust has a quite solid theory backing. Zig has nothing interesting really: it's "vibe coded" by someone who was doing game dev or something like that.
More precisely speaking: GC languages are said to delay memory problems far beyond the horizon, which is often unreachable throughout the project's history. Zig can be a similar case.
But they wouldn't get a change to the structural issues that created the issues in the first place. They'd end up "ke[eping] fixing these kinds of bugs one-off in perpetuity".
Let me save you time and tell you that C# and it's ecosystem is where you'll be happy.
I expect if he'd spent $165,000 running Fable against the Zig version he could have got a 5% performance improvement, too.
It's not great for Zig if you have to put in more work to end up at the same place efficiency-wise, especially for a language marketed at people who like to get the most out of their metal.
They could have done the same in Zig, even though it probably shows cargo is better at this than build.zig.
I've been working on a new rewrite that's focused on beating Postgres on performance. As of this morning I got to 100% of the tests passing and have meaningful performance gains over Postgres.
Like, you have now created your own new database engine that happens to be compatible with Postgres. Wouldn't you take an absolute pride in giving it your own name? Why call it "Postgres rewrite in <programming language>"?
[1] https://github.com/malisper/pgrust I guess
After having used 2 full weeks of 20x Max plan tokens on Fable over the weekend (coding all day Saturday and Sunday on a non-trivial project, tasks across full stack, mix of adding features, reviewing code, and fixing bugs), I’m confident if he’d spent $165,000 in Opus tokens the port would have gone more or less just as well (and probably for less than $165,000). Especially so with the system they set up with all the custom workflows, adversarial reviews, extensive test coverage, etc.
But I get your point is probably more about Jarred’s experience level and the high cost than the specific model used other than it being SOTA. I’m just being pedantic and feeling a bit disappointed with Fable’s real world performance after all the hype.
> I expect if he'd spent $165,000 running Fable against the Zig version he could have got a 5% performance improvement, too.
Totally agree and in fact I’m sure it could be done with significantly less cost even if they stuck with Fable instead of Opus which I’m sure could also do it.
But a lot of people have been saying this for sometime, Rust and LLM is a great match. A lot of friction of the language were smoothed out by LLM assisted programming.
That said, I think the parts around wanting to properly have memory safety guarantees rather than try hard & patch as issues are found is a more serious concern for Zig as those speak more to the design goals than the current implementation. "better safety than C while maintaining C compatibility" may not be a very compelling reason to chose Zig if other languages are able to do that portion better anyways, even ones without a GC.
Zig was the right tool to start, Rust is the right tool to finish.
Making something possible and refining something until it is high security and reliability are different problems. Zig is great, but for a JS runtime, I just don't think that's the best long-term fit.
This is impressive from a technological standpoint, but it does gloss over the fact that it would have cost $165k in tokens were Bun not part of Anthropic.
The comparison here isn’t completely fair - it would take a small team a year to port it if they spent $0 extra on it.
I’d be interested to see a comparison between spending $165k in 11 days on Claude vs splitting that between 50 people over 11 days for a line-by-line rewrite of the Zig code. I suspect Claude might be faster and therefore cheaper, but maybe not by a lot.
Even if this is not the right answer today, it can at the very least serve as a herald of a possible future, no?
Salary info: https://www.levels.fyi/companies/anthropic/salaries/software...
[0]The maths I used (posting because I'm tired and prone to mistakes):
$336,000 / 260 (working days of the year) = ~$1,292.
$1,292 * 11 * 50 = ~$710,769Of course, then you can also ask, could it have been done with a cheaper model. Probably yes. But then you wouldn't get free marketing.
That said, I don't think "rewrite from one language to another" with inexpensive engineers is a pattern that works. Happy to be proven wrong.
There are excellent engineers in India, but the system they operate in unfortunately doesn't allow them to shine.
On Teams, channels related to AI are flooded with daily support requests from supposed engineers from India who clearly are not competent enough to set up GitHub Copilot or properly report issues they encounter during the setup.
And don't get me started on the shared libraries some teams located in India work on. If the library I need to use is full of obvious bugs where I wonder how any competent engineer could have shipped this to production, and then I see that the work has been moved to India, how am I supposed to feel about this?
Do we really need to sugarcoat this?
USA has amazing software engineers and less competent ones.
Europe has amazing software engineers and less competent ones.
etc
You get what you pay for, pay top end in India and they will be same as top end in Bay Area (and many in Bay Area/USA are migrants...)
$1200 per day.
Your estimation is 50*11 days so $660,000. That’s 4x what Claude cost.
That’s assuming that you actually get those 50 people to work without blockers, stepping on each other, or other coordination issues. The coordination complexity alone is astounding.
I don’t like it necessarily, but Claude wins here, easily. It’s not close.
Which takes us to a point of future US dev salaries if this thing with agents gets better more and more
What does the math look like with 25 devs making ~100k and doing it in 22 days? I’m sure you could find a reasonable combination which costs less. And if you’re already paying the devs the salary, it’s basically free (minus the opportunity cost of them not working on other things).
Even a junior is going to cost you $200k by the time you're done paying payroll taxes, healthcare, etc.
This is a question of exceptional management, which needs to be present both in the Claude and human cases, and is scarce. Not everyone given the Claude tokens would be able to deliver the same result.
While Jarred used Mythos-class model, some open weights, if they were as capable (certainly, GLM 5.2 looks the part), would have been way, way cheaper than professionals.
Approx costs:
DeepSeek v4 Pro & Mimo v2.5 Pro $3,426 ($2,567 / $600 / $259)
Tencent HY3 $3,892 ($1,180 / $552 / $2,160)
GLM 5.2 $30,016 ($8,260 / $3,036 / $18,720)
Qwen 3.7 Max $37,925 ($14,750 / $5,175 / $18,000)
Claude Opus 4.8 & GPT 5.5 xhigh $82,750 ($29,500 / $17,250 / $36,000)
5.9 billion uncached input tokens, 690 million output tokens, 72 billion cached input token reads.Opus 4.8 still found some real issues afterwards and spent about an hour fixing things, before the code was good enough to ship. Overall promising, wrote about it here: https://blog.kronis.dev/blog/z-ai-s-glm-5-2-is-a-great-model...
The GLM Coding Plan seems to have lower token limits than the corresponding Anthropic Max subscriptions, but if you had to pay API rates for some LLM to do work somewhat reliably, it's a no brainer (unless you're swimming in money that you can give away and value your time more).
HINT: those 50 people must be coordinated...
PEOPLE make stuff, people use stuff, and people are ultimately the ones who are going to pick and choose which stuff gets made, used, adapted, enhanced, and carried into the future.
AI is an inherently anti-social, anti-human technology, and this rewrite is the perfect example of that.
Assessed from the perspective of "technology in a vacuum", of course. it's a success. He did the thing that transformed the thing from one kind of stuff to another kind of stuff. It still does all the things it did before, and in many cases with better stats than it did before.
Assessed from the human angle, and especially the angle of Bun as a community, I would bet money that this rewrite -- executed by nobody for nobody, built and maintained by machines, maintainable only by machines -- has killed the entire project.
Maintainable only by machines, because anybody with any knowledge, experience, or investment into Bun as a platform, or who contributed patches themselves, or whoever had a question about how it works and went "Hmm, I'm gonna go into the codebase and take a look at how that happens", they all got slapped in the face and summarily kicked out of the tent with the rewrite.
Not really. LLMs are a tool, and they do possess some interesting properties, but they are, in essence, a very surprising auto-complete.
No, what sucks is the AI companies marketing them as if they are The Holy Grail. And people are assuming that somehow LLM -> AGI -> ASI, even though there isn't really a causal link to be had, and acting like it's The Singularity.
I don't blame autocomplete for fucking up RAM, labor and god knows what else. I blame the idiots (at OpenAI and Anthropic) buying all RAM stocks and pretending this shit is any way good for anyone.
TBH in hindsight it seems that the Bun acquisition was also for the PR stunt they just did with the Rust porting.
> TBH in hindsight it seems that the Bun acquisition was also for the PR stunt they just did with the Rust porting.
The events went roughly like this:
- Anthropic used Bun. So they acquired it.
- Bun made some LLM derived patches for Zig. They were rejected due to Zig no-LLM stance.
- Bun couldn't get patches in, so they switched to Rust.
Was it a PR stunt, an attempt to pressure Zig maintainers into changing AI stance, or just doing the fastest thing possible? It's difficult to tell. I'm inclined to believe the last option.
As easy as it is to blame Bun for a PR stunt, it's also possible that you can't really associate with someone who rallies against your employer/owner.
This is a very weird issue to be mad about, and can be extended to multiple technologies - if you think about it, from computers to cellphones.
[0] https://github.com/oven-sh/bun/blob/fc865b398e51de8a95ddde4b...
But second, you're right, this is an easy thing to search for (especially with LLMs). And yet, the example I linked has been there since may, surviving multiple rounds of review. From this, we can draw a few conclusions: 1) claude (including apparently mythos/fable) fundamentally doesn't "understand" how unsafe rust works, and 2) no one on the bun team is aware that they need to tell it to fix this (if they're even aware that this isn't allowed in the first place).
The reason such trivial defects in a codebase are a red flag isn't so much those specific issues themselves, it's what they reveal about the authors. If you find a lot of obvious defects in some code, it's very likely that there are also many more subtle harder to detect issues as well.
And for the emotional aspect. I'm working as an R&D engineer and one of my recently assigned experiments is to evaluate replacing workers with LLMs, quality software with slops. Every bit of exaggeration they make is making my coworkers lose jobs and I'll be no exception. The entirety of me consists of reasons going against this kind of marketing compaign.
>Many projects opt to answer these kinds of questions through a style guide. TigerBeetle's TigerStyle is an example in Zig and Google's 31,000 word C++ style guide is another. The challenge with style guides is enforcement.
TigerStyle[1] is a bit more than just a style guide. The key rule for this discussion, uplifted straight from of NASA[2], is *static memory allocation*: all memory is allocated in the startup phase, and there's absolutely zero `alloc`s afterwrads . This plus crash only[3] design means that we never call `free`.
This rule is self-enforcing and compositional, in Zig. There's no global memory allocator, so the code after startup simply hasn't the API to allocate. You can't circumvent this by accident. Of course, if the programmer is byzantine, they can stuff allocator in the global, or just directly `mmap` and `unmap` pages of memory, but, at our scale, we don't have problems with that. This is a similar in kind (not degree) to Rust, where untrusted code generally can circumvent safety guarantees, even without literally spelling `unsafe`.
And, naturally, never `free`ing goes a long way towards solving many memory errors by construction. Empirically, they just haven't been a problem for TigerBeetle. It's hard to untangle contribution of static allocation in particular from everything else we are doing, but it would make sense for it to play a leading role.
(As a footnote, we aren't actually do static allocation to avoid memory errors, we use it as a linter to check that every quantity has a known _logical_ static limit, the main property we care about)
[1]: https://github.com/tigerbeetle/tigerbeetle/blob/main/docs/TI...
[2]: https://spinroot.com/gerard/pdf/P10.pdf
[3]: https://www.usenix.org/legacy/events/hotos03/tech/full_paper...
A huge benefit is having to think and be explicit about the limits up front. To take your example of arbitrary sized JSON and flip it around: how would you do it normally?
Maybe, you'd allocate a buffer to hold your entire object when it comes in - but now you'll end up crashing when you get a large document that exceeds your available memory.
Maybe you can do things in a fixed amount of memory, using streaming or chunking - which would be pretty simple to turn to static allocation!
Static allocation forces understanding of those limits upfront.
What has pushed me back to Node is seeing how amateurish the transition has been handled.
- No LTS support for the Zig version regarding CVEs etc.
- Huge bugs like the 3MB memory leak mentioned in the blog post abandoned in the Zig version to basically force people into the Rust version to fix their apps in production.
- Zero involvement with the Bun community about such a major decision. One day it was "stop the drama I'm just playing with this" and a couple of days later "yolo merged to main".
Jarred basically keeps operating as if he was a lone hacker working on his personal project.
> Or are you expecting open source maintainers to do free work for no particular reason?
Free work? Last I heard Anthropic had acquired Bun.
I don't know it makes sense to try to make the an LTS version of where the Zig version left off, particularly if they know they are shipping a different solution & codebase to tackle the bulk of their security and memory bugs. Let this settle a few more months with the intent of releasing an LTS of it by the end of the year, if there really is demand for it. All of the benefits of having an LTS version like Java, Node, and .NET without needing to jump to a pre-port version of the codebase which was never targeted to be a good version to LTS in the first place.
A final Zig-based LTS would have made great sense if LTS releases had already been in the picture though.
If they are not it would be a good reason to move away from it.
To my understanding, Java and .NET offer public LTSes because they’re financially buoyed by huge companies using those LTS versions, and Oracle and Microsoft are fundamentally selling that LTS support as a product. I have no idea why Node does, maybe it’s similar. But Bun’s position appears to be different than all three.
In fact, he had two adversarial reviewer Claude instances on every code change, every line. I don’t know a single human team that does two independent reviews of every line, except maybe the people that wrote space shuttle software.
Also they fixed the memory leak. How does it matter what language it’s written in? At the end of the day, people use it to run their typescript code among other things.
How many bun users care that’s it’s written in zig? I certainly don’t. I’ve been using bun for 2 years and I think I looked up zig once. It’s just not relevant.
Did it get more stable? Yes. Slimmer? Yes. More performant? Yes. Is there any proof that it got LESS secure? No. The code has been out for two months. By now all the nay sayers would’ve found the smoking gun. They haven’t. How much more proof would you like that this was a resounding success?
This is our new reality. The agents are so good that projects like this are in the realm of possible. That’s exciting.
Do we know 100% for certain that this isn't the case? No.
In fact it would be naive to think a rewrite of this magnitude wouldn't introduce new bugs and/or unexpected changes in behavior.
> Nothing to me suggests they’re being careless here.
Plenty of reasons suggest this including the lack of an LTS or any kind of thought put into such a massive transition.
There were initial analyses done on the port. And things continue to get refactored. But is there any slam dunk article where someone actually found any regression in functionality or stability? We’re seeing the opposite. Dozens of bugs fixed. We don’t have to theorize. They’ve been running this experiment for 2 months, with all the code out in the open.
It just feels like after two months, people want to cling to the _idea_ that this was reckless, without evidence of any meaningful negative impact.
1) My (and possibly other people's) last impression of this was when it was merged just based on all tests passing. In many projects, relying just on tests would definitely be reckless (not sure about coverage/quality of Bun tests).
I didn't follow much what where they doing later, maybe it is indeed good enough. For example, Claude Code using it and being fine is reassuring.
2) Making such a big decision that quickly and not (even having the time to) consulting community doesn't really inspire confidence.
3) You can be reckless even if everything ends up being perfect in the end.
Going 80mph in a city, you probably have like 80% chance of not crashing. And if you don't crash, you just had a much faster and more fun trip. Doesn't mean it wasn't reckless.
This. Something I expect any person who think they are "engineer" to understand without thinking.
Every release would have tons of CVEs and would take so much effort. E.g. the example from blog with memory issues. Better just think that Zig version was not there what comes to security. Use at your own risk.
> Jarred basically keeps operating as if he was a lone hacker working on his personal project.
They have right to do it, however. It is expected, especially if company owns it.
Yeah. The human aspect of the transition was just incredibly bad. The person behind Bun has just demonstrated how much he values the community.
But I'm sure he will build another community around this rewrite. After all, there is an abundance of people cheering "Rust rewrites".
Until the new favorite language comes up.
yolo!!!
Yeah, that was beyond ridiculous.
I think Rust is a locally optimal target for LLM coding, we might see a better language in the future, but I think Rust will dominate for quite some time.
Faster iteration, maybe? Rust's safety guarantee isn't exactly free (while still being very excellent) and does affect iteration time. I have a private project (>300K LoC) that has been translated from Python to TypeScript and the reason we couldn't use Rust was definitely the iteration time.
What costs rust in iteration time in my opinion is the low level (by default) nature of it. There's a faster-to-iterate language that has yet to be created which is rust but we sacrifice performance (and memory fiddling ergonomics for the odd person who does that) so we don't have to worry about things like whether a variable is stack or heap allocated. Which is in the direction of a GCed language but retains the mutable-xor-aliasable semantics.
Between rust and current GCed languages though... I guess I agree with "maybe" in both directions.
It's really the only systems language in its exact niche.
I'm suggesting a language where there's no difference between Box<u32> and u32. &Vec<u8> and &[u8] are the same thing. I don't need to write Box::new(...) around my closures to pass them to functions that take a function pointer. This comes with overhead, but in exchange we get simpler less verbose code. I.e. a language that isn't systems level, and isn't particularly machine-empathetic. But still has all the lightweight-formal-methods power of rust with lifetimes and mutable vs shared borrows (and thus references to references) and so on.
My impression of Hylo is that it's purpose is to be a similarly low level systems language to rust, just with a less complicated, and as a consequence less expressive, lightweight formal methods system for proving correctness.
I agree I don't expect rust to be displaced anytime soon. It creates a lot of time to create a good compiler, and a lot more to create the ecosystem of code, tools, and community around it.
-- Andrew Kelley
https://andrewkelley.me/post/my-thoughts-bun-rust-rewrite.ht...
> Two, I actually don't have any personal criticisms of Jarred
The author says two things that really popped out to me that you could say are "professional" and not "personal" criticisms, but I think they're still rude and contrast this statement.
> Jarred was already writing slop well before he had access to LLMs
> The grapevine was large and healthy and full of juicy grapes, and all those grapes contained the juice of the same message: Jarred was a stinky manager. Poor communication, unrealistic expectations, low empathy, no experience
Now, both of these may be true. I don't have any evidence though, so I don't know how to take it.
All this to say, I'll take both of these posts without a ton of salt when it comes to the non-technical parts.
Why not just port Claude code over.
But my guess is that maybe it doesn’t have as robust a test suite?
This might embolden them to do it…
Not sure about what's going on over there, but over the last year Claude Code has gotten way better, I sure hope that they're working on a good enough test suite to avoid a crapload of regressions (and pave way for more refactoring, should they need it).
This changed for me over the last 5 years.
The first scenario was joining a company where a software product barely worked. We did the traditional incremental refactoring / rewriting, but eventually learned how rotten the core was that rewriting from first principles was the best path forward.
The lesson learned here is that the conventional wisdom probably only applies to rewriting complex but working systems.
Then multiple scenarios in the agentic coding age. Between day jobs and hobbies I've reproduced major chunks of complicated software like Salesforce, Gmail, Pioneer Rekordbox with very lean teams.
Much like the blog post, the trick is to get an excellent verification loop with a compiler, linter, and test harness / test suite around the core behaviors.
It's feeling more and more that designing and implementing comprehensive test harnesses is the real work, once you have that let the LLM cook.
However, I've been skeptical of using Bun, because I want a project whose first and foremost goal is to build good tools that achieve the objectives of the project.
It reminds me of asking game developers: Do you want to build a game, or do you want to build a game engine? Building a game engine is fine, but if you're goal is to make a game, then building an engine is a poor way of achieving your goals.
Likewise, I've wondered if the creators of Bun wanted to build better JavaScript tools, or if they wanted to use Zig.
Clearly the model itself doesn't completely change the narrative, but at least as a note to myself, I would like to be more careful with assuming the capabilities of the models used internally by Anthropic and affiliated orgs.
I thought the same thing. Looking back, I was probably mislead in May when Jarred was explaining the pattern to "Rewrite every .zig file to .rs" as if it was something I could have done in May following his pattern. What he wasn't telling us was he was using pre-release Fable. [1]
A possible signal for next time is when we see an Anthropic owned company disabling the Claude Co-Authored-By trailer. [2] In an IPO year they have to take every chance to promote Claude unless it was something (Fable) that we weren't supposed to know about back in May.
[1]: https://xcancel.com/jarredsumner/status/2060050586024743376#...
[2]: https://github.com/oven-sh/bun/commit/23427dbc12fdcff30c23a9...
People who are surprised by this probably has not seen what Zig code actually looks like. Zig's explicitness and lack of abstraction have a real cost that it is basically one of the most verbose programming languages I've ever seen, it's somehow even more verbose than Go. Basic features of modern languages like pattern matching and generics, and as you can see, having to manually clean up everything means that if you forget once, it's a memory leak. Having SOME abstraction is actually good if it prevents you from making mistakes.
Ironically, Zig is a programming language that's probably best written by LLMs, since they can actually tolerate the verbosity.
Rust in my opinion feels the same.
There may be some prompting that can help with this but I suspect there is a fundamental tension between writing working code vs good code in LLMs. Go is popular for being simple, making it easy to jump in and write something fast and stable - minimizing the gap between working and good code probably helps out the LLMs a lot.
Sounds like a good time to tell Claude to create a system wide memory of how to implement lifetimes and borrowing and what to NEVER do.
Claude will then make the memory file. Usually memories help a lot. I see it think about it when it reasons through code.
#[derive(Copy, Clone)]
enum Expr {
Int(i32),
Add(i32, i32),
Neg(i32),
}
fn eval(expr: Expr) -> i32 {
match expr {
Expr::Int(x) => x,
Expr::Add(a, b) => a + b,
Expr::Neg(x) => -x,
}
}
#include <stdint.h>
typedef enum {
EXPR_INT,
EXPR_ADD,
EXPR_NEG,
} ExprTag;
typedef struct {
ExprTag tag;
union {
struct {
int32_t value;
} Int;
struct {
int32_t left;
int32_t right;
} Add;
struct {
int32_t value;
} Neg;
};
} Expr;
int32_t eval(Expr expr) {
switch (expr.tag) {
case EXPR_INT:
return expr.Int.value;
case EXPR_ADD:
return expr.Add.left + expr.Add.right;
case EXPR_NEG:
return -expr.Neg.value;
}
__builtin_unreachable();
}
I am saying that I do not believe there is a correlation between source code length and binary length. If that's what benced meant by their question, then yes, I agree :)
cmp ecx, 1
je .LBB0_3
cmp ecx, 2
jne .LBB0_2
Oddly, both sources seemed to be quite sensitive to match switch and enum reordering, resulting in very different generated code. Possibly something to look into further.
To properly answer this you'd need to compare a large number of identical implementations written idiomatically in several languages and see if there is a correlation.
If I were to throw my 2 cents in I'd say "a very weak correlation" is probably right. Not because verbose languages HAVE to result in more bloated code but because it seems to me languages fine having a lot of bloat in the syntax also tend to be languages fine having a lot of bloat in the implementation or attracted to abstraction (which never does seem to actually compile away fully in large projects, even though it often largely does).
> it seems to me languages fine having a lot of bloat in the syntax also tend to be languages fine having a lot of bloat in the implementation or attracted to abstraction
Which is why I chose an example of the exact opposite: a language not known for bloat, taking way more code to produce the exact same thing as one that's more succinct.
It's not as good as some sort of scientific survey of a wide variety of options, but if you can find examples in all directions, assuming there's no correlation until proven otherwise is a pretty solid bet, I think.
E.g. one could seek to find a 6' preteen and 6' adult to construct a counterexample to the idea height is in some way correlated with age. Doing so gives just as little evidence of what the strength of correlation is as seeking to find a 5'9" preteen and a 6' adult to show the correlation is positive or seeking to find a 6'1" preteen and a 6'0 adult to show the opposite. I.e. it doesn't follow one can filter the search as they please and then assert that's what the correlation of the unfiltered searches should be assumed to look like. In all 3 cases of positively correlated, negatively correlated, and not correlated we'd expect to be able to construct an example which says whatever we want to say - that isn't the same thing as sampling what the actual correlation usually is.
What I was gesturing at, badly, was more that Zig’s low-abstraction / explicit-by-default syntax tends to have you write more boilerplate-y code in general that are more annoying to write and maintain, while not buying you enough over a language with better tooling and ecosystem and compiler optimization like Rust.
Also note that the larger percentages were against already smaller binaries. That smells like there was a single large constant number that got saved somewhere rather than general improvements.
> After that initial shrinkage, the team explored more opportunities for binary size reduction using linker optimizations like Identical Code Folding, removing unused data from ICU, and lazily decompressing small parts of libicu with a zstd dictionary on-demand.
I'd be VERY interested in seeing what the individual effects of those parts were.
The main reason why Zig is verbose in some aspects is the main goal of Zig is program performance. It is a worthy tradeoff.
Whether or not a language is verbose or obscure is very much about your coordinate system. Not unlike safety.
I think C is a reasonable zero for both things.
Zig is more succinct and safer than C while still being comparably ergonomic. Rust is (mostly) safer and more succinct than Zig while being dramatically less ergonomic (take it up with Wadler memory chads, no one likes affine types).
I like lean4, which is dramatically safer, more succinct, and more ergonomic than Rust.
But I can see why some would say it's a bit too succinct.
This is just your opinion.
It wasn't until 2014 that Orchard formalized the coeffect discharge calculus via indexed monad that makes a binary ownership semantic irretrievably sunsetted as a degenerate case.
It's my opinion. I'm not concerned about how informed that opinion is.
It seems the reports of Bun's death have been greatly exaggerated.
I'm guessing you mean probabilistic? Nevertheless, you have an indeterministic variable here which is what the LLM generates.
The most significant revelation for me was that Claude Code has been using the rewrite without much fanfare since June 17th.
Some of the world's more important software - from browsers to the JVM - mix high-level languages with a GC and low-level languages, and it works not because of a style guide (even though one may exist). As someone working on the HotSpot JVM, I can say that it's done with a lot of thinking about constructing the right primitives that make this work well. Zig doesn't lack the features to construct the mechanisms required for getting good results in that domain, and Rust doesn't have features that could save you the thinking about such mechanisms.
All the ones that opted to switch language after less than a decade were those with serious shortcomings in their technical decision process, and those problems, unsurprisingly, persisted after the language change. After all, the very decision to switch so soon is an admittance that they'd made a very serious misjudgment, but these projects never properly debrief why they'd made such a big mistake and how they can avoid making one again.
First, technically respected products like the ones you describe either 1. plan or expect to switch in advance (e.g. they start with, say, Python/Ruby, expect that if they grow they'll switch to, say, Java) or 2. they improve their chosen language runtime (e.g. Facebook with PHP/Hack or Shopify with Ruby). Projects that switch a language without expecting to always show a pattern of bad decisions (clearly, if they thought their chosen language will carry them through growth and then they're convinced that it won't, that means that they don't know to judge languages' merits).
Second, this is clearly not the situation here, is it? There is absolutely no new information that Bun learnt in the past year that they didn't have five years ago, and certainly this has nothing to do with growing workloads on some service. They say they believe the language they have chosen lacks the features needed for the very core of the domain, which is dealing with JS objects. As someone working on the HotSpot JVM, I can tell you this is not true, but fine - that's what they believe. What could have taken them five years to come to that conclusion? Again, this happens to be a domain close to my own, only much simpler, and I seriously doubt they made some novel discoveries in the past year. And if it's taken them five years to acknowledge what they now think are fundamental limitations with the language they had chosen, how can they be so confident they've made a right choice now after a few weeks? It looks like they chose Zig on a whim and then chose Rust on a whim, and neither of these choices is the source of their problems and neither is the solution to them.
Why is your unsupported speculation about the guy (or the people above him) supposedly deciding that "Rust being a better fit for an Anthrophic-owned project" a critical insight exactly? You not being convinced by the rationale presented does not mean he himself wasn't. Do you imagine critical thinking as just coming up with and preferring (to you) plausible-sounding alternative explanations?
Because respectfully, that's very much not how I understand the word, and I'd hope most people do not either. Maybe that's the issue. Post-hoc rationalization being a rather shaky gate with people (and AI) anyways.
Worse still, neither of the options you present are trustworthy summaries necessarily in the first place. From the blogpost:
> What if, instead, I spend a week testing if Anthropic's new model can rewrite Bun in Rust?
That sounds to me like something of an excited but cautious dare, not an arbitrary technological decision, nor some hyper-strictly rationale-supported one. It's the same kind of exploratory excitement anyone who used these things at all would be familiar with: you have an unlikely project idea, with some contentious details that are fair (see their blogpost), and some that would be simply annoying to justify (see your comment). Buuuut you also have heaps of corpo money, a lingering suspicion and interest, and AI tokens to burn. And so "fuck it, we ball". Not much to believe on it when it's a shared experience. And it's not because of just having read so somewhere, despite your assertions otherwise.
Strong subjective judgements about an inexact but large sounding fraction of people, without any actual details, is a red flag the size of a skyscraper. And to top it all off, your conclusion is extremely convenient too: you're right, people are wrong, despite the blatant facts otherwise. Not exactly the hallmark of a particularly robust position, not for me anyways.
Any languages with LLVM backends get cross-language LTO for free.
Have they ever tried Zig Bun with LTO?
This is the bit I was really curious about. Definitely not something within reach for us mortals.
For now though it is a bit disappointing trying something like this is relegated to project proposals at work rather than my personal hacking.
The coolest outcome was being able to run a redis comparible store on an a cloudflare durable object so you do I.e. rate limiting for free with little infra.
However, an open-sourced tool like that would've greatly harmed the Zig ecosystem and community.
People looking to abandon the ship first chance are unlikely to contribute much to the ecosystem and community.
That's what underlies most of the vitriolic reaction to the events, it was done really in a really rash way.
All they had to do was a)not gaslight people about your intentions when they found the branch b)publicly post the intent to do this, and then c)publish a doc like this one right before merging, ideally leaving the branch open for like a week in case anyone in the "community" finds things to fix.
Then those UB/Miri issues others found would have been "yay collaboration" boosts instead of negative issues that prove that the approach was risky/unthoughtful.
If you're going to cross a rubicon, maybe tell people a)that you actually want to do it, b)why you want to do it and c)what it looks like after you've done the rewrite.
I feel like people will make the wrong comparison with the cost to complete. $165000 should be compared to not the cost of a programmer going line by line by hand but someone designing a transpiler from zig to rust. The time to complete is impressive though, if you could spend $165000 and a year of time to find out the rewrite project worked, or instead spend that in a month, you'd probably take that month now that this proof of concept exists out there.
(2) The rewrite itself is a massively successful marketing move. It shows what Claude code can do and how little it costs compared to human engineering. But the question remains whether someone else, not knowing zig, rust, and TypeScript can pull this off.
My biggest issue currently, is I can't seem to get a code review that's about the simplicity of the code, and no /simplify ain't it. Removing certain bugs and generally working seems to be doing alright, especially if it's following either an example code (like in the Bun rewrite case) or a well defined "spec" of how to proceed.
He makes it sound like Claude did a fantastic Rust rewrite, and "the work continues."
But when the Rust port merged to main, the state of the code was very, very bad. There were 13,000 instances of `unsafe`, no Miri tests at all, and, sure enough, it exposed UB in safe Rust. https://github.com/oven-sh/bun/issues/30719
Observers could see this coming from a mile away, objected strongly to using AI to RIIR before the code merged. Rather than incorporate feedback and get the code ready for production, Jarred gaslit us all, right here on HN. https://news.ycombinator.com/item?id=48019226
Just 9 days before he merged the Rust rewrite to the main branch, Jarred wrote:
> This whole thread is an overreaction. 302 comments about code that does not work. We haven’t committed to rewriting. There’s a very high chance all this code gets thrown out completely.
It's plausible that Bun's Rust rewrite is now in much better shape than it was in May. But a blog post like this would have been a place to apologize, to accept that it was a very bumpy rollout, to acknowledge that public messaging was extremely poor, and to earn back our trust.
As it stands, I guess I'll have to run my own tests to try to evaluate whether Bun 1.4 is ready for prime time, because I just can't trust Jarred to give us a straight answer.
Like have you run into a specific bug or seen a regression - that's the cause of your reaction?
God forbid an engineer express uncertainty.
Based on that, the bun rewrite messaging was fairly misleading.
Either they estimated poorly, or it ended up the lesser portion of their estimate after all. After all, unless the estimate is 100%, there's always a chance it'll fall into the other portion.
Maybe Jared just broadcasted uncertainty and was wrong, but given his position he's not being given the normal grace you might extend to an engineer you trust.
I mean yeah, that's what this whole post is about. It's about the process of going from that original state to something that's now shipping in production.
It'd be interesting if Anthropic became a general software company just because they have access to models that aren't yet released, possibly export-banned.
This was done by someone who has essentially already rewritten Node once. Bun itself is a reimplementation of Node, so the author was walking in knowing exactly what the correct behavior is. And an exhaustive amount of test suite to verifiy the changes?. On top of that, there is a reference from Node and V8 to validate more throughly. So the $160K is simply the price of translating knowledge that already lived in one engineer's head in a newer syntax.
From the article
> Pre-merge, this took 5.9 billion uncached input tokens, 690 million output tokens, and 72 billion cached input token reads — around $165,000 at API pricing
> It's hard to believe that there have been no problems/downsides since the port.
A significant portion of the article was dedicated to the 19 regressions they've found. Starting here: https://bun.com/blog/bun-in-rust#porting-mistakes
So essentially this whole re-write was about making Bun LLM compatible.
I clearly don't get the value proposition of bun? And even if I accept that you want to bundle your run time, package manager, test runner and bundler, why do you want to include things like a custom S3 client?
This is also not a memory issue but an actual logic issue, so Rust doesn't help there.
I don't think this is a value anymore for them.
I wonder if the delay of Fable has also been a factor and maybe they didn't want to release this blog post while they couldn't allow customers to use Fable and waste the advertising opportunity.
Great for the Bun creators, but now we will have major runtimes that are optimised to work with one companies models...
I'm sorry but that is insane, how was this never fixed before the rewrite?
Furthermore, there's no mention of an LTS plan for the Zig version. It seems that if a CVE is discovered in the future, Bun users will no have no option than to update to the Rust port.
This is not how you run a project that others depend on and enough for me to not touch Bun ever again.
It would be naive to think there aren't new bugs or changes in behavior introduced in 1.4.
But yes, of course there will be new bugs. But that's why 1.4.x for x > 0 is interesting. If the branch is being used and people are not reporting _more_ bugs, and the bugs you care about it are being fixed (successfully) on it, and it passes your tests, etc., ... I dunno. This is an application domain where you can do some pretty solid testing of it, comparative fuzzing, etc., so it doesn't strike me as entirely mad to jump over after a few minor releases where you can see the bug trajectory.
Anthropic is not exactly the hallmark of engineering excellence... quite the contrary.
> But yes, of course there will be new bugs
Obviously, which is why more thought should have been put into the transition.
Not everyone will want to yolo their production projects into such a massive rewrite overnight.
if you want superior developer experience - Deno.
if you want stability over everything - Node.
I hate how easy git makes it for llms to catastrophically fuck up repos
> And if I put each Claude into a separate worktree, I would run out of disk space because Bun's git repository is too big and eventually the changes will need to be compiled and seen together.
This seems like a stupidly easy problem to fix.
> So, I asked Claude to edit the workflow to instruct Claude to never run git stash or git reset or any git command that doesn't commit a specific file at once. No cargo either. No slow commands at all.
What's the right way to fix this? I have a Pi extension that blocks any use of the word 'git' with 'stash/checkout/reset/restore' in commands which is a very large hammer approach. Is there a way I can allow a subset of git use in a nicer way without the less nice commands?
It's pretty exciting.
I was looking forward to this blog post too, but in retrospect I don't know why. I could have had an LLM generate a hypothetical of what this blog post might have looked like and it would have probably been able to get close.
I feel like we've replaced unique voices on the internet with the same style / author, which might be more tolerable if the breathless LLM writing style wasn't so jarring. Contrary to the amount of times "But honestly" or "genuinely" is mentioned, nothing about having your LLM speak for you feels honest or genuine.
I know it's not cool to leave responses like this, but I'm really tired of all of this at this point. The ironic thing too is that it might actually be better to have LLM written text be so distinct so that you can still pick out when a human has actually authored something. Again, this is a blog post from Anthropic about having an AI translate 500k+ lines of code in 11 days, so I guess my disappointment is my fault for expecting otherwise.
It doesn't read at all AI-generated to me. What section do you think is?
(Pangram is very good at distinguishing between AI-generated and human text, and assigns a very low score to the article: https://www.salahadawi.com/hacker-news-ai-detector/rewriting...)
> Contrary to the amount of times "But honestly" or "genuinely" is mentioned, nothing about having your LLM speak for you feels honest or genuine.
"Honestly" is used once in that post, in a way that's pretty much the core, self-deprecating human use for it ("It would have been possible to do X, but honestly I didn't want to"), rather than the filler word use-case.
"Genuinely" is not used at all.
> I know it's not cool to leave responses like this, but I'm really tired of all of this at this point.
I think it is cool to flag AI-generated slop and either leave a comment or upvote an existing comment about it being slop. But only if you are sure it's AI-generated. And sorry to say, you don't seem very well calibrated on this. If you can't actually tell the difference and back up your opinion but are just guessing, then it indeed isn't cool.
It's not.
I propose GPTSD (GPT + PTSD)
EDIT: Another one, AIdar (like gaydar but for AI text). EDIT2: GPTSD has prior art (literally) https://ryanthompson.name/project-gptsd.html
I'm very glad to see a holidtic approach to the memory errors and segfaults. I was tinkering on a static webpage just this this weekend, using Bun as the transpiler+bundler since it's so turnkey, and I ran into a few segfaults. E.g. when Bun saw I used an empty data uri for the favicon (avoids the browser trying to ask for one) it'd just crash. It reminded me of my own tinkering with Zig in its current pre-release state where it's usually a good mix of my poor memory management and working around bugs in Zig itself.
This post is also the best ad for AI I've seen yet. Not just comments saying they have 10xed themselves, a small personal project or thing which can (and likely will) be abandoned next month, or a one off dump of unusable code for the world's buggiest C compiler (come on Anthropic?). Instead this is a well thought out way of leveraging LLMs to do something which would otherwise probably not be deemed a reasonable enough effort.
I'm glad they included the rough cost as well. Crazy high, more than I can afford to be throwing at the wall to see what sticks, but still low enough to make sense over trying to hire developers for (even ignoring the timeline).
But it also highlights 2 really key things about current LLMs: the scaffolding can be just as valuable as the model & they still need someone able to figure out the right way to instruct and orchestrate them. Without the scaffolding the current models could never get close to handling something of this scale & quality. With the scaffolding you could probably get this to work okay enough even with a weaker model. On the orchestration side, Claude could help answer what good porting practices would be but it doesn't just get there itself, it requires hours of someone who understands the context of the project from bottom up and a clear understanding of what will/won't work to get the right scaffolding to do the job well.
Finally, it was an interesting dichotomy on presenting the port. On one hand it has been a bit opaque up until now. People saw the repo and there were some comments about testing the waters but then it suddenly shipped into production. On one hand that's awesome and I'm sure the reception here would be very different if there wasn't the "it's already been boringly shipped in Claude Code" shining result. On the other... I think I'll stick to using Bun as a personal tinkering tool for now. The velocity is so high I'm just not sure it'll land in a place I can rely on it st the speed of my org. Of course, that velocity is what has made Bun's success story and they should probably continue with it - I'm just looking forward to an LTS release :).
I think that when you have a $165,000 hammer, all of your problems begin to look a lot like nails.
1. Microsoft don't want to open source .NET Core debugger. 2. I tired of keeping upgrade .NET on my projects. 3. Result type in Rust make me more productive than exception in .NET. 4. async/await in Rust is lightweight and a better than .NET. 5. Thread-safety in Rust is a compiler error instead of figure out by checking the docs if type is thread-safe. 6. Community libraries in Rust has a great quality and docs.
Is it worth $165K? I'm less sure of that but it's honestly a moot point - this will get to 5 then 4 digits of cost pretty fast.
Have European engineers do it for $100k or Asian engineers do it for $50k and the math is already looking a lot sketchier.
> But, we would still be reliant on style guides enforced through code review, and even with ASAN, memory corruption and memory leaks would still happen.
Tell me you didn't even look at C++ without telling me you didn't even look at C++. I don't understand this at all, what's missing? There's clang-format, clang-tidy, cppcheck and so many others, what is missing exactly? Memory safety? Then why bring up C++ and style guides(?) at all?
Clang-format doesn’t save you from all C++ footguns, e.g. using exceptions, macros or templates in the wrong way where „wrong“ is defined by a fuzzy set of rules that requires a lot of experience and vigilance to enforce.
I was hoping it’d end in a “so how much did this cost?” so that others team looking at similar migrations have an estimate on what they can expect
> 11 days (May 3 → merged May 14) · 6,778 commits
> Pre-merge, this took 5.9 billion uncached input tokens, 690 million output tokens, and 72 billion cached input token reads — around $165,000 at API pricing. By hand, I think this would've taken 3 engineers with full context on the codebase about a year, during which time we wouldn't be able to improve Node.js compatibility, fix bugs, fix security issues or implement new features. We never would've done that. The realistic alternative was to do nothing and keep fixing the bugs at the top of this post forever.
It was in the middle. $165k.
The entire point is to get people to spend money on LLMs. Writing a transpiler - even a LLM-coded one - pretty much defeats the purpose.
as I said because it's a much smaller codebase. This is a million lines of code project. A literal Zig to Rust translation that is mostly syntactic is not going to be more than 10-20k LOC for a transpiler. That's two orders of magnitude less work.
Something that would have taken hundreds of developers now took 1 developer with Fable.
Now Claude, rewrite Claude Code from TypeScript to Rust. Make absolutely zero mistakes.
> There are a lot of ways to do a terrible job of this. For example, prompting Claude "Rewrite Bun in Rust. Don't make any mistakes." and then praying it would work is not what I did.
Not really and we both know that is not true.
The last line of the comment was the joke. [0] It is clear that you didn't get it and many others did. That is OK.
Heavens forbid anyone here is allowed to make any jokes on this site.
I will be a lot more excited when this is possible with <10k of api costs.
For a rewrite of this size, the expensive part is deep understanding of the underlying system in order to preserve behavior while keeping performance, and above all that not freezing product work while doing it. Adding more engineers would just end up in managerial burden and review bottlenecks, to say the least.
So even assuming the API cost estimate is high, I don't buy the “just hire engineers for a month” take. A team unfamiliar with the codebase would probably spend a large chunk of that month just building context and deciding how not to break everything. A team familiar with the codebase is even more valuable doing product work, bug fixes, and review of the existing codebase.
So, in short, I do agree with the simple fact that this is still too expensive for most projects, but not with the idea that “a small team would trivially do better in a month”.