Hacker News Reader

Comment by jimbob45 | Hacker News Reader

These articles seem like they're playing code golf more than proving anything as a viable alternative to C.

The goal is to analyze your needs and pick the language best suited to your task. The goal is not to find one language that excels at everything.

View on HN

I actually disagree. If you look at the code written in the article, it's hardly unreadable or very code golf-y as you suggest. I think the author is really just trying to say that programmers shouldn't dismiss Go as a possible systems language in favor of C just because C has a reputation of being faster in all cases.

Are we really saying a garbage collected language should be considered for a systems language? I know the GC is fast but surely memory managed languages are going to have less issues with pausing during execution, no?

Re: pause times, Go's garbage collector is actually really good here; you're looking at 10s of microseconds.

I'm convinced the term "systems language" doesn't have a coherent meaning at this point. See:

https://zenhack.net/2018/07/14/three-funerals-in-the-name-of...

I always thought the GC pause depends on the heap size... a little searching... http://big-elephants.com/2018-09/unexpected-gc-pauses/.

It depends on how many things need to be collected, which is often a function of heap size.

Right, so when I’m trying to process a network packet in about a microsecond (a reasonable target for financial trading software), Go’s GC rules it out. Many low-level, high-performance systems language tasks aren’t feasible in Go for similar reasons.

That doesn’t make it a bad language, but it’s not universally applicable either.

Strange. I know OCaml is used in HFT (thanks to all those Jane Street ads) and OCaml is also GC'd. What does OCaml have here that Go doesn't?

> but it’s not universally applicable either.

Nothing is universally applicable.

But yeah, I certainly wouldn't use it for hard real-time tasks. If you can't tolerate missing deadlines ever, there is a very short list of acceptable tools.

But (and correct me if I'm wrong; it's not my area) HFT doesn't strike me as hard real-time? See also a sibling comment that asks about Jane Street's OCaml use.

It's not like GC pauses are happening constantly; Go programs don't allocate that much, and a well tuned program can go a long time between collections. It likely is appropriate for many soft or firm real time systems. And you can shut the GC off if there are sections where GC really must not happen:

https://golang.org/pkg/runtime/debug/#SetGCPercent

That's how those who use Java do it: write their code to minimize allocations, and then configure the JVM with a GC threshold that's bigger than their whole day's allocations, and manually GC at a suitable time.

Last I looked (a while ago) it seemed like there wasn't quite enough control on the Go GC to do this effectively. It's very likely that's been fixed by now.

Just like in hard real time systems using malloc() rules C out, and its use is explicitly forbidden in certification processes.

I guess C isn't universally applicable either.

The difference is, you can write perfectly usable programs in C without littering the codebase with malloc() and free().

Just like you can in Go, as proven by TinyGo.

Or in dozen of other GC enabled systems programming languages for that matter.

Tip: Give up dude.

Exactly my tip to the anti-GC luddites.

The difference being evolution, takes care of sorting that out, like in all anti-technology groups throughout mankind history.

In the sense that the Go authors used "systems language", i.e. for writing things like servers and command line tools, sure why not? Go's GC pauses are extremely short (under 1ms). That shouldn't really affect much.

Source? People write CLI tools in Ruby and Node.js these days, so I really doubt that anybody would use the term "systems language" to refer to languages well-suited to building CLI tools.

https://www.wired.com/2009/11/google-announces-a-new-program...

> As a systems language, Go is intended to be used for developer applications like, for example, web servers.

Couldn't find a citation for CLI tools, but that should satisfy you (since people write web servers in Ruby and Node.js too).

As others mentioned "systems language" isn't well defined. If you want to define a "systems program" as one with hard real-time requirements, then no, Go isn't a very good choice; however, the article defines it differently (such that `wc` satisfies), and demonstrates that Go satisfies that definition.

That said, it's better to use terms that map better to a set of requirements, such as "hard-realtime" or "soft-realtime".

Yes, as proven multiple times, and being pushed forward by the likes of Apple, Google and Microsoft.

Regarding using Go as a real systems language (in the same meaning as C):

- gVisor hypervisor on Google Cloud and Linux sandbox on Chromebooks

- Android GPGPU debugger

- Fuchsia TCP/IP stack and volume management

- Baremetal TinyGo on Arduino Nano33 IoT, Adafruit Circuit Playground Express, BBC micro:bit among many others

- Coreboot firmware

Coreboot is C/ASM though.

Esp. as a systems language. Safety first.

I think there are already a few gc/jitted components in mainstream kernels (bpf, lua drivers). So in some points, it's not a crazy idea. Now I don't advocate for Java drivers.

I'm not sure this proves that Go is a possible systems language, since it could be equally like that the explanation is just "wc doesn't need to be written in a systems language to be fast". For an executable that runs, processes some stuff, and then terminates in the span of a second or so, you're not going to incur much slowdown due to a GC counting references and cleaning stuff up, but for something long-running, it could make more of a difference.

I could be wrong, but it seems like it wouldn't be difficult to outperform these implementations with a similarly hacky C implementation. I expect that blog post in the coming days.

Why wait?! Do your "Beating C with C" blog post and enjoy your HN fame!

It's mentioned in the post https://github.com/expr-fi/fastlwc/

Ah, I didn't see that. Thanks!

Hey, author here. You are absolutely right - Go would have a hard time outperforming finely tuned C. Rather, the article was more focused on exploring concurrency mechanisms in Go. The (admittedly clickbait) title was more of a reference to the articles before it than anything else.

That said, I think you'll find this SIMD-enhanced wc interesting: https://github.com/expr-fi/fastlwc/

Author here. Why do you feel it is code golf? My primary focus when writing it was readability - I'm sure I could do it in much less than 70 lines if I had to. If you have any suggestions for improving the readability of my implementation, do let me know!

The version you are comparing against is ancient and assumes a low-memory footprint. It only uses something like 10KB of data. It also has a ton more features -- like command line parsing, the ability to read from a file on the command line or stdin, and locale-specific recognition of whitespace.

It's really not solving the same problem.

Argument parsing is hardly the bottleneck when you're scanning 1 GB of data. The reason I didn't implement the rest of the features was to keep the implementation light and readable, and I don't think a fully compliant wc implemented in Go with these patterns would be significantly slower.

As for having a low-memory footprint, 2 of the 4 implementations I mentioned (one single core, one multi-core) consume less memory than wc, and still run much faster.

How much memory does your program uses VS the C program? Did you profile that?

FTA: ”Our parallelized implementation runs at more than 3.5x the speed of wc, while matching its memory consumption”

Yes, it's the last column in all the result tables.

Reply on news.ycombinator.com