Lies we tell ourselves to keep using Golang (2022) | Hacker News Reader

Lies we tell ourselves to keep using Golang (2022)

fasterthanli.me

342 pointsreillyse511 comments

This article makes a lot of great points about the shortcomings of Go. I don’t think explicit error handling is one of them however. I’ve previously spoken about my loathing of exception handling because it adds a “magic” layer to things which is way too easy to mess up. From a technical standpoint that isn’t necessarily a good argument, but from a pragmatic standpoint and decades of experience… well I will take explicit error handling which happens exactly where the errors occur every day. You can argue that Rust does it in a more elegant way, and I prefer it for personal projects. For big projects with a lot of developers of various skill level joining and leaving I think Go’s philosophy is one of the sanest approaches to error handling in the modern world.

Staying in that lane. In my part of the world Go is seeing adoption that no other “new” language has exactly because of its simplicity. It’s not the best language, but it’s often the best general purpose language because it has a lot of build in opinions which protect you from yourself.

When I started trying to teach myself Rust, the error handling story fell apart on me very quick.

Like as soon as I wanted to try and get sensible reporting in their, suddenly we were relieving libraries, adding shims and fighting mismatched types and every article was saying the same thing: haha yeah it's kind of a problem.

I'm very, very unsold on explicit error handling compared to exceptions for practical programming. The number of things which can error in a program is far larger then those that can't.

I felt the same but after switching to anyhow and thiserror in pretty much every Rust project I work on I find it quite painless. It's not ideal to rely on crates for a core language feature but I never find myself fighting error types anymore. Have you tried those crates? Do you still hold that opinion?

You don’t need crates for it, anyhow is basically a better Box<dyn Error>, if you just want the error signal you can use that. The main thing missing from the stdlib fur this use case is I don’t think there’s anything to easily wrap / contextualise errors built in.

The problems you're describing don't exist in go. There is exactly one standard type that is used by everyone, at least in public API's, you can always just return the error to the caller, if you don't want to handle it in place. The main difference with exceptions in my practice is the fact that it's a lot easier to mess up, since it requires manual typing. This is probably my main problem with everything being as explicit as possible: it requires people to not make mistakes while performing boring manual tasks. What could possibly go wrong?

I think that sort of nails it: the problem with errors as values is errors become part of the type signature and put under user control, and the user can't really be trusted with that power.

Even the simplest functions tend to have error states - i.e. floating point math can always wind up handing back NaN.

So where I end up is, the default assumption is every function is effectively of a type MightError(T)...so why even make us type this? Why not just assume it, assume the unwrap handling code, and so you basically wind up back at try-catch exception handling as a syntactic sugar for that whole system.

The drawback, on the other hand, is that all the Go code I've read (including the stdlib and all the dependencies of my current project) is using `fmt.Errorf` or `errors.New`, which means that you can't use `errors.As`, which means that you generally cannot handle errors at all.

I would certainly argue against the claim that explicit error handling is far overkill.

Where I agree: It forces you to think about all of the possibilities your code might generate. (This is more of a C question than it is with other languages)

However, when abstracting blocks of code away, you don't always need to handle the error immediently or you may want to handle it down the stack.

You're giving up a lot of readability in order for the language to be particular.

Rust handles this much better.

Error handling is still explicit, but it gives you the tools needed to make it less tedious.

> It forces you to think about all of the possibilities your code might generate.

Except it doesn't actually. You can totally just ignore it and pretend errors don't exist. Lack of sum types/Result, and pointers as poor mans optional, really hinder's Go's error handling story.

I agree with you that the sum types are much better and it's intentional that they wanted null support. (Which was a frustrating pick considering the modern dev languages out there).

The coding style encourages the behavior you're talking about. We're back in the C days where you have a result and an error. It's not a good pattern.

> It forces you to think about all of the possibilities your code might generate

I’ve seen way too much Go code which never even tested the err value to believe that until something like errcheck is built in to the compiler.

I do agree that this is a plus for the explicit model but that’s been a better argument for Rust in my experience since there’s better culture and tooling around actually checking errors. I’m sure there are plenty of teams doing a good job here, but it always felt like the one lesson from C they didn’t learn well enough, probably because a language created by experts working at a place with a strong culture of review doesn’t really account for the other 99% of developers.

It's actually very rare that it should be the caller who has to handle the errors.

Go, however, forces you to spread your error handling over a thousand little pieces with zero overview or control of what's happening.

Rust eventually realised this and introduced try! and ? to simplify this

More importantly, Rust has the notion of a result type and it is designed to be both generic and composable.

A problem I often face in Go and TypeScript code is code that ignores errors, often unintentionally. For instance, many uses of JSON.parse in TypeScript do not check for the SyntaxError that may be thrown. In Go, it is common to see patterns like

  _ := foo.Bar()
  // assume Bar() returns error

This pattern exists to tell the reader "I don't care if this method returns an error". It allows one to avoid returning an error, but it also stops the caller from ever being handle to the error.

Also, the position of the error matters. While the convention in the stdlib is to return errors as the final value, this isn't necessarily followed by third party code.

Similarly, errors are just an interface and there is no requirement to actually handle returned errors. Even if one wants to handle errors, it's quite awkward having to use errors.As or errors.Is to look into a (possibly wrapped) chain of errors.

The benefit of Rust's Result<T, E> is that

- position doesn't matter

- there is strong, static type checking

- the language provides operators like ? to effortlessly pass errors up the call stack, and

- the language provides pattern matching, so it's easy to exhaustively handle errors in a Result

The last two points are extremely important. It's what prevents boilerplate like

  if err != nil {
    return nil, err
  }

and it's what allows one to write type-safe code rather than guess whether errors.As() or errors.Is() should be used to handle a returned error.

I am pretty sure if it were for the Typescript creators they would not allow exceptions in the language, but they had to work within the confines of Javascript. Heck they even refused to make exceptions part of the type-system.

It is unfortunate that many of Typescript developers still rely on throwing exceptions around (even in their own typescript code). Result types are totally doable in Typescript and you can always wrap native calls to return result types.

Why would you "check" for TypeError being thrown? Just let exceptions in general propagate until they reach one of the few places in the program that can log, display, crash, or otherwise handle an exception. No need to "check" anything at call sites.

90% of the criticism of exceptions I see comes from the bizarre and mistaken idea that every call needs to be wrapped in a try block and every possible error mentioned.

Unsure if this is the right place to ask, but this conversation inspires me this question:

Is there in practice a significant difference between try/catch and Go's "if err" ? Both seem to achieve the same purpose, though try/catch can cover a whole bunch of logic rather than a single function. Is that the only difference ?

Try/catch can bubble through multiple layers. You can decide/design where to handle the errors. If you don't `if err` in Golang, the error is skipped/forgotten, with no way to catch it higher up.

You can decide not to catch a thrown exception, it travels upwards automatically if you don't catch it.

I think that's the biggest difference.

With Go you need to specifically check the errors and intentionally decide what to do, do you handle it right there or do you bubble it upwards. If you do, what kind of context would the caller want from this piece of code, you can add that too.

I'd argue that at least checked exceptions also require a conscious decision from you. You either need to add the Exception type to your throws clause, or your catch clause.

Compared to Go, this is actually better because the type system can tell you what kind of errors to expect, instead of just "error".

> With Go you need to specifically check the errors and intentionally decide what to do, do you handle it right there or do you bubble it upwards.

Is this really all that interesting or worth the LOC spent on error handling when 99.9999% of the time in practice it’s just bubbled up?

And any “context” added is just string wrapping. Approximately nobody types golang errors in a way that lets you programmatically know what went wrong, to be able to fix it in-line.

I think I would be more empathetic to the arguments defending golang here if I’d ever worked or seen a project where people actually handled errors instead of spending 2/3 of their time writing code that just punts on any error.

“if err” doesn’t catch all types of errors. Some errors are colored different from others and instead cause the program to immediately crash, sorry panic. But don’t worry! it’s just very rare errors, like nil dereference and index out of bounds, that throw unrecoverable errors like this!

> Rust eventually realised this and introduced try! and ? to simplify this

That was prototyped around Rust 0.4, so I wouldn't say "eventually" :)

Having programmed for over 30 years, including nearly a decade of C#, I would say exceptions are one of the worst ideas in all of programming.

They are just horrific gotos that any library can invoke against your code. They are pretty much never, ever handled correctly. And nearly always, after an exception is “handled”, the application is actually in an unknown state and cannot be reasoned about.

Even junior engineers have a trivial time debugging most go errors, while even experienced principles struggle with figuring out the true cause of a Java exception.

> never handled correctly

I’ve seen this argument, but if you look at real golang code and examples, it’s just a bunch of “if err <> nill” copy pasta on every line. It’s true that handling errors is painstaking, but nothing about golang makes that problem easier. It ends up being a manual, poor-man’s stack-trace with no real advantage over an automatically generated one like in Python.

Which could be solved in one swipe by adding a Result<T, Error> sum type, and a ? operator to the language. This is more a self-inflicted limitation of Go, then a general indictment of explicit error handling.

Nothing prevents explicit error handling in Python either. Forcing explicit error handling just creates verbosity since no system can functionally prevent you from ignoring errors.

Many people against Go's error handling do not advocate for exceptions, but for a combination of an Either/Result type (for recoverable errors) and fully aborting (for unrecoverable errors).

Abort on the other hand is used WAY to liberally in Rust.

How I hate it, that every second function call can break my program when it's clearly not a "halt the world, it's totally unrecoverable that the user sent us nonsense" type.

Return a Result and get on with your life!

If a Rust function can panic, there's generally a non-panicking alternative. For example, `Vec` indexing has `vec[n]` as the panicking version and `vec.get(n)` as the version that can return `None` when there's nothing at that index.

I do wish this is something Rust had done better though - the panicking versions often look more attractive and obvious to developers, and that's the wrong way round. Vec indexing, IMO, should return Option<T>.

While that is true, there are clippy::indexing_slicing, clippy::string_slice for that:

  https://github.com/rust-lang/rust-clippy/issues/8184#issuecomment-1003651774

  error: indexing may panic
     --> src/main.rs:100:57
      |
  100 |             rtmp::header::BasicHeader::ID0 => u32::from(buffer[1]) + 64,
      |                                                         ^^^^^^^^^
      |
      = help: consider using `.get(n)` or `.get_mut(n)` instead
      = help: for further information visit https://rust-lang.github.io/rust-clippy/master/index.html#indexing_slicing

One of my dream projects is creating a Rust stdlib based solely on panics for error handling. Sure, it'd be incompatible with everything, but that might be a feature, not a bug.

I think someone's already created this.

One of my dreams is for someone to create a Rust stdlib, full stop.

I love Rust the language, but the current bazar of little bits of functionality scattered around in the form of a zoo of crates is such a mess compared to, say, Java's class library (I/O, data structures, std. algorthms for searching, sorting etc., arranged in a logically and hierarchially named way).

I'm not against alternative implementations, but I'd rather have one "official" implementation that everyone knows that covers most cases and makes for the idiomatic reading of source code.

No!

I mean, yes, but not an official one.

The stdlib is where good ideas go to die.

Waiting for "pub struct RDBMSInferfaceV17ThisTimeWeGotItRightForSure"

Firm disagree.

Sylvain Kerkour described the problems Rust faces by having a limited standard library:

"The time has come for Rust to graduate from a shadow employment program in Big Tech companies to a programming language empowering the masses of engineers (and not just "programmers") wanting to build efficient and robust sfotware.

What crypto library should we use? ring, RustCrypto, rust-crypto (don't use it!), boring, aws-lc-s or openssl?

Which HTTP framework? actix-web, axum, dropshot or hyper?

What about a time library? time, chrono or jiff (how I'm even supposed to find this one)?

You get it, if you are not constantly following the latest news, your code is already technical debt before it's even written. Fragmentation is exhausting.

I just looked at the dependencies of a medium-sized project I'm working on, and we have 5+ (!) different crpyto libraries: 2 different versions of ring, aws-lc-rs, boring, and various libraries from RustCrypto. All of this because our various dependencies have picked a different one for their own cryptographic usage. This is insane, first because it introduces a lot of supply chain attack entry points, but also because there is no way that we will audit all of them"

Source: https://kerkour.com/rust-stdx

My moderate-sized Rust web service project requires 587 third-party crates which seems ridiculous.

Fortunately, cargo makes it easy to manage the dependencies, but unfortunately I don't know how well supported or maintained the dependencies are. How well will they be maintained in five years from now? Will I need to find newer libraries and rewrite portions of my project to provide the same features that I have now? I don't know.

in the stdlib, yes. In 3rd party crates? Depends!

Exception and explicit on-the-spot handling are not the only two ways to handle failing processes. Optional/result types wrapping the are a clean way to let devs handle errors, for instance, and chaining operations on them without handling errors at every step is pretty ergonomic.

Rust's error handling evolution is hilarious. In the beginning, the language designers threw out exceptions --- mostly, I think, because Go was fashionable at the time. Then, slowly, Rust evolved various forms of syntactic sugar that transformed its explicit error returns into something reminiscent of exceptions.

Once every return is a Result, every call a ?, and every error a yeet, what's the difference between your program and one with exceptions except the Result program being syntactically noisy and full of footguns?

Better for a language to be exceptional from the start. Most code can fail, so fallibility should be the default. The proper response to failure is usually propagating it up the stack, so that should be the default too.

What do you get? Exceptions.

> what's the difference between your program and one with exceptions

Because errors as values are explicit. You're not forced to use ? everywhere; you can still process errors however you like, or return them directly to the calling function so they deal with it. They're not separate control flow like exceptions, and they're not a mess like Go's.

No, because you end up with a function coloring problem that way. A function that returns something other than Result has to either call only infallible code or panic on error, and since something can go wrong in most code, the whole codebase converges as time goes to infinity on having Result everywhere.

Yeah, yeah, you can say it's explicit and you can handle it how you want and so on, but the overall effect is just a noisy spelling of exceptions with more runtime overhead and fewer features.

I very much care about whether a function can fail or not, and I encourage all the function colouring needed to convey that.

As almost always, we programmers / software developers / engineers, forget to state our assumptions.

In closed-world, system-software, or low-level software you want to have your kind of knowledge about everything you call. Even more: can it block?

In open-world, business-software, or high-level software it is often impossible or impractical to know all the ways in which a function or method can fail. What you need then, is a broad classification of errors or exception in the following two dimensions: 1. transient or permanent, 2. domain or technical. Those four categories are most of the time enough to know whether to return a 4xx or 5xx error or to retry in a moment or to write something into a log where a human will find it. Here, unchecked exceptions are hugely beneficial. Coincidentally, that is the domain of most Java software.

Of course, these two groups of software systems are not distinct, there is a grey area in the middle.

Monadic error handling (such as Rust's) is a good compromise because it allows you to handle errors precisely when you want to _and_ bubble them up with minor boilerplate when you don't. You can even add context without unwrapping the whole thing.

That's why it's often considered one of the better approaches: it has both the "errors as values" benefits (encoding errors in the type system) and avoids many of its problems (verbosity when you just want to bubble up).

> A function that returns something other than Result has to either call only infallible code or panic on error

...Or solve the problem. A library function that can be a source of issues and can't fix these issues locally should simply not be returning something that is not a result in that paradigm.

> since something can go wrong in most code

That is not my experience. Separating e.g. business logic which can be described cleanly and e.g. API calls which don't is a clear improvement of a codebase.

> the whole codebase converges as time goes to infinity on having Result everywhere.

As I said previously, it is pretty easy to pipe a result value into a function that requires a non-result as input. This means your pure functions don't need to be colored.

> Or solve the problem. A library function that can be a source of issues and can't fix these issues locally should simply not be returning something that is not a result in that paradigm.

People "solve" this problem by swallowing errors (if you're lucky, logging them) or by just panicking. It's the same problem that checked exceptions in Java have: the error type being part of the signature constrains implementation flexibility.

In my experience an unwrap, expect or panicking function is a direct comment in code review and won’t be merged without a reason explaining why panicking is acceptable.

Function "coloring" is good! It's not a problem here and it's overblown as a problem in general. If something fails recoverably then it should be indicated as such.

I think your conclusion is on the right track. Syntax sugar propagating results is in a way equivalent to exceptions. What you are missing it isn’t just equivalent to ordinary exceptions. It’s more equivalent to checked exceptions. A very powerful concept, that unfortunately got a bad rap, because the most widespread implementation of it (Java) was unreasonably verbose to use in practice. You might claim Rust is still too verbose and I don’t disagree with that, it’s a big improvement from Java at least, while at the same time providing even more safety nets.

Yes, Java is too verbose, but Kotlins cleaned up much of that boilerplate and runs on the same VM.

I'd be curious to see examples for where you think Rust is still to verbose.

You misunderstood. They meant that Java's checked exceptions are very verbose and that Rust's approach to errors is similar. While Rust's approach is less verbose, it's still more verbose than "regular" (unchecked exceptions).

Kotlin isn't relevant here because all exceptions are unchecked.

"Function coloring problem"

Function coloring isnt a problem, it is just approach

> Once every return is a Result, every call a ?, and every error a yeet

The Try operator (`?`) is just a syntax sugar for return. You are free to ignore it. Just write the nested return. People like it for succinctness.

Yeet? I don't understand, do you mean the unstable operator? Rust doesn't have errors either.

> what's the difference between your program and one with exceptions except the Result program being syntactically noisy and full of footguns?

Exceptions keep the stacktrace, and have to be caught. They behave similar to panics. If panics were heavy and could be caught.

Rust errors aren't caught, they must be dealt with in whatever method invokes them. Try operator by being noisy, tells you - "Hey, you're potentially returning here". That's a feature. Having many return in method can both be a smell, or it could be fine. I can find what lines potentially return (by searching for `?`).

An exception can be mostly ignored, until it bubbles up god knows where. THAT IS A HUGE FOOTGUN. In Java/C# every line in your program becomes a quiet return. You can't find what line returns because EVERY LINE CAN.

I'm not sure why you bring up Rust here, plenty of libs/languages use the Result pattern.

Your explanation of what bothers you with results seems to be focused on one specific way of handling the result, and not very clear on what the issue is exactly.

> what's the difference between your program and one with exceptions

Sometimes, in a language where performance matters, you want an error to be handled as an exception, there's nothing wrong with having that option.

In other languages (e.g. Elm), using the same Result pattern would not give you that option, and force you to resolve the failure without ending the program, because the language's design goals are different (i.e. avoiding in-browser app crash is more important than performance).

> syntactically noisy

Yeah setting up semantics to make users aware of the potential failure and giving them options to solve them requires some syntax.

In the context of a discussion about golang, which also requires a specific pattern of code to explicitly handle failures, I'm not sure what's your point here.

> full of footguns

I fail to see where there's a footgun here? Result forces you to acknowledge errors, which Go doesn't. That's the opposite of a footgun.

The '?' operator is the opposite of a footgun. The whole point of it is to be very explicit that the function call can potentially fail, in which case the error is propagated back to the caller. You can always choose to do something different by using Rust's extensive facilities for handling "Result" types instead of, or in addition to, using '?'.

In most languages with exceptions:

• they may propagate automatically from any point in code, potentially breaking atomicity invariants and preventing forward progress, and have to be caught to be transformed or wrapped – Result requires an explicit operator for propagation and enables restoring invariants and transforming the error before it is propagated.

• they are an implicit side-channel treated in the type system like an afterthought and at best opt-out (e.g. "noexcept") – Result is opt-in, visible in the return type, and a regular type like any other, so improvements to type system machinery apply to Result automatically.

• try…catch is a non-expression statement, which means errors often cannot be pinpointed to a particular sub-expression – Result is a value like any other, and can be manipulated by match expressions in the exact place you obtain it.

Sure, if you syntactically transform code in an exception-based language into Rust you won’t see a difference – but the point is to avoid structuring the code that way in the first place.

> they may propagate automatically from any point in code, potentially breaking atomicity invariants and preventing forward progress

A failure can propagate in the same circumstances in a Rust program. First, Rust has panics, which are exceptions. Second, if any function you call returns Result, and if propagate any error to your caller with ?, you have the same from-anywhere control flow you're complaining about above.

Programmers who can't maintain invariants in exceptional code can't maintain them at all.

> try…catch is a non-expression statement,

That's a language design choice. Some languages, like Common Lisp, have a catch that's also an expression. So what?

> they are an implicit side-channel treated in the type system like an afterthought an

Non-specific criticism. If your complaint is that exceptions don't appear in function signatures, you can design a language in which they do. The mechanism is called "checked exceptions"

Amazing to me that the same people will laud Result because it lifts errors into signatures in Rust but hate checked exceptions because they lift errors into signatures in Java.

Besides, in the real world, junior Rust programmers (and some senior ones who should be ashamed of themselves) just .unwrap().unwrap().unwrap(). I can abort on error in C too, LOL.

> A failure can propagate in the same circumstances in a Rust program.

What do you consider failure? A result or a panic? Those are worlds apart.

> First, Rust has panics, which are exceptions.

That's not how exceptions work in Java. Exceptions are meant to be caught (albeit rarely). Panics are meant to crash your program. They represent a violation of invariants that uphold the Safety checks.

Only in extreme cases (iirc Linux kernel maintainers) was there a push to be able to either "catch panics" or offer a fallible version of many Rust operations.

The Rust version of Java Exceptions are Results. And they are "checked" by default. That said, Exceptions in Java are huge, require gathering info and slow as molases compared to returning a value (granted you can do some smelly things like raising static exceptions)[1].

[1] https://shipilev.net/blog/2014/exceptional-performance/

> Non-specific criticism. If your complaint is that exceptions don't appear in function signatures, you can design a language in which they do. The mechanism is called "checked exceptions"

And everyone hates checked exceptions. Because rather than having a nice pipe that encapsulates errors like Result, you list every minute Exception possible, which, if we followed the mantra of "every exception is a checked exception" would make, for example, Java signatures into an epic rivaling Iliad.

> Besides, in the real world, junior Rust programmers (and some senior ones who should be ashamed of themselves) just .unwrap().unwrap().unwrap().

If this is a game who can write worse code, you'll never find Java wanting. As a senior Java dev, I've seen things... Drinking helps, but the horrors remain.

> Rust has panics, which are exceptions.

Panics are fatal errors that the program is not expected to recover from, and it might as well terminate on the spot – as if you cut off power, which you have to be ready for no matter what. In fact, abort-on-sight-without-unwinding is one of the ways they can be configured to work. It’s the case where an operation fails, but the program attempts to continue running that is the most fragile. This is where you’re likely to enter an invalid state.

> you have the same from-anywhere control flow you're complaining about above

Not from literally anywhere – from the places where the error-propagation operator is used. I will never have myself thinking “oh wait, F could have thrown before G is invoked and break my invariants!” months after writing that code. C++ is chock-full of such problems: https://www.youtube.com/watch?v=b9ZYM0d6htg

> Some languages, like Common Lisp, have a catch that's also an expression. So what?

So more power to them! That is actually a major improvement compared to your typical dynamic ALGOL, where you find yourself stuck in situations like:

    def foo():
        try:
            return d1[k1] + d2[k2]
        except KeyError as e:
            """ now what? which lookup failed? """

and have to resort to clumsy workarounds.

> Amazing to me that the same people will laud Result because it lifts errors into signatures in Rust but hate checked exceptions because they lift errors into signatures in Java.

I don’t think you can write a generic over a checked exception list in Java, which means higher-order functions cannot deal with them. Modern implementations just don’t bother. Meanwhile, when errors are values, improvements to the type system in the happy path will apply to the error path as well, because they don’t use distinct language facilities that have to be separately considered and specified.

One practical benefit of Rust’s approach that hasn’t been emphasised enough yet is the consequences of Option<T> and Result<T, E> being just values, same as anything else.

It means you can use things like result.map_err(|e| …) to transform an error from one type to another. (Though if there’s a suitable From conversion, and you’re going to return it, you can just write ?.)

It means you can use option.ok_or(error) or option.ok_or_else(|| error) to convert a Some(T) into an Ok(T) and a None into an Err(E).

It means you can .collect() an iterator of Result<T, E> into a Vec<Result<T, E>>, or (one of the more spectacular examples) a Result<Vec<T>, E> which is either Ok(items) or Err(first_error).

It’s rather like I found expression-orientation, when I came to Rust from Python: at first I thought it a gimmick that didn’t actually change much, just let you omit the `return` keyword or so. But now, I’m always disappointed when I work in Python or JavaScript because statement-orientation is so limiting, and so much worse.¹ Similarly, from the outside you might not see the differences between exceptions and Rust-style Result-and-? handling, but I assure you, if you lean into it, it’s hard to go back.

—⁂—

¹ I still kinda like Python, but it really painted itself into a corner, and I’ve become convinced that it chose the wrong corner in various important ways, ways that made total sense at the time, but our understanding of programming and software engineering has improved and no new general-purpose language should make such choices any more. It’s local-maximum sort of stuff.

Exceptions are values in C++, Java, and Python too. They're just values you throw. You can program these values.

As usual, I find that opposition to exceptions is rooted in a misunderstanding of what exceptions really are

Exceptions are values, but normal-value-or-exception (which is what Result<T, E> is) isn’t a value. Review my remarks about map_err, ok_or, &c. with the understanding that Result is handling both branches of the control flow, and you can work with both together, and you might be able to see it a bit more. Try looking at real code bases using these things, with things like heavily method-chained APIs (popular in JS, but exceptions ruin the entire thing, so such APIs in JS tend to just drop errors!). And try to imagine how the collect() forms I described could work, in an exceptions world: it can’t, elegantly; not in the slightest.

Perhaps this also might be clearer: the fuss is not about the errors themselves being values, but about the entire error handling system being just values.

It’s about where the error value ends up versus where the success value ends up. With sum types they both end up in the return value and are handled in the same place. Typically you will now try to let the value flow further with the typical functions (bind/bimap/map/or_else and what else we can come up with).

With exceptions they are thrown and end up somewhere else in a catch block, so the path of the success value is now distinct from the path of the error value.

Honest question: syntactically noisy as opposed to what? In the context of this post, which is a critique of Go as a programming language, for me this is orders of magnitude better than the "if err != nil {" approach of Go.

The big difference is API stability, one of the primary focuses of Rust.

Hard disagree. Exceptions are actually good. They make code clear and errors hard to ignore. I've written a ton of code over decades in both exceptional and explicit-error languages and I'll take the former every day. There's no function color problem. No syntactic pollution of logic with repetitive error propagation tokens.

Also, exception systems usually come with built in stack trace support, "this error caused by this other error" support, debugger integration ("break the first time something goes wrong"), and tons of other useful features.

(Common Lisp conditions are even better, but you can't have everything.)

You can't just wave the word "goto" around as if it were self-evident that nonlocal flow control is bad. It isn't.

> And nearly always, after an exception is “handled”, the application is actually in an unknown state and cannot be reasoned about.

That's not been my experience at all. Writing exception safe code is a matter of using your language's TWR/disposable/RAII/etc. facility. A programmer who can't get this right is going to bungle explicit error handling too.

Oh, and sum types? Have you read any real world Rust code? Junior developers just add unwrap() until things compile. The result is not only syntactic clutter, but also a program that just panics, which is throwing an exception, the first time something goes wrong.

Many junior developers struggle with error handling in general. They'll ignore error codes. They'll unwrap sum types. They might... well, they'll propagate exceptions non-fat ally, because that's the syntactic default, and that's usually the right thing. We have to design languages with misuse in mind.

Bah, no, I hated that you had to wrap basically every code block in a try/catch in Java, because the underlying lib could change and suddenly throw a Runtime-Exception.

At the same time Checked Exceptions were a nightmare as well, because suddenly they were part of the contract, even though maybe wrong later.

Checked exceptions are more trouble than they're worth. That doesn't make exceptions in general bad.

As said, I don't like the wrapping of about everything with try/catch

Sure, you can only do it way up the stack, but that's not enough quite often.

If you can only do it all the way up, I find it ergonomic.

Maybe I should experiment more with catch unwind in Rust.

What does make exceptions bad in my opinion (and shared by Go developers?) is a few things:

1. Exceptions are expensive (at least in Java / C#), as they generate a stack trace every time. Which is fine for actually exceptional situations, the equivalent of `panic()` in Go, but:

2. Exceptions are thrown for situations that are not exceptional, e.g. files that don't exist, database rows that don't exist, etc. Those are simple business logic cases. The workaround is defensive coding, check if the file exists first, check if the row exists? that kind of thing.

3. The inconsistency between checked and unchecked exceptions.

4. Distance - but this is developer / implementation specific - between calling a function that can throw an error and handling it.

But #2 is the big one I think. Go's error handling is one solution, but if it's about correct code, then more functional languages that use the Either pattern or whatever it's called formally are even better. Go's approach is the more / most pragmatic of the options.

> e.g. files that don't exist, database rows that don't exist, etc. [...] The workaround is defensive coding, check if the file exists first, check if the row exists?

Ugh NO. Please don't. You should never "check if the file exists first". It can stop existing between your check and your later attempt at opening the file (the same with database rows). That can even lead to security issues. The name for that kind of programming mistake, as a vulnerability class, is TOCTOU (time-of-check to time-of-use).

The correct way is always to try to do the operation in a single step, and handle the "does not exist" error return, be it a traditional error return (negative result with errno as ENOENT), a sum type (either the result or an error), or an exception.

Totally agreed, but as the previous poster wrote:

an exception is meant for EXCEPTIONAL behavior.

So it may be that the file access throws an exception but generally, I wouldn't agree.

Not having checked exceptions is a huge problem, because then you never know when something might throw and what it might through, and in the .NET world the documentation on that is pretty awful and absolutely incomplete.

But then over in Java world, your checked exception paradise (which it of course isn't because the syntax and toolkit for managing the things is so clunky) is easily broken by the number of unchecked exceptions which could be thrown from anything at any time and break your code in unexpected and exciting ways, so not only do you have to deal with that system you also don't get any assurance that it's even worth doing.

But this doesn't actually mean checked exceptions are a bad idea, it means that Java didn't implement them very well (largely because it also has unchecked exceptions, and NullPointerException is unchecked because otherwise the burden of handling it would be hideous, but that comes down to reference types being nullable by default, which is a whole other barrel of pain they didn't have to do, and oh look, Go did the same thing wooo).

> in the .NET world the documentation on that is pretty awful and absolutely incomplete.

Depends on the area you look at. Language documentation is pretty good and so is documentation for the standard library itself. Documentation for the frameworks can be hit or miss. EF Core is pretty well documented and it’s easy to find what to look for usually. GUI frameworks are more of a learning curve however.

FWIW many in Java community consider checked exceptions to be a mistake. While I don’t find writing code that has many failure modes particularly fun with exception handling - Rust perfected the solution to this (and the Go way is visually abrasive, no thanks), I don’t think it’s particularly egregious either - Try pattern is pretty popular and idiomatic to use or implement, and business code often uses its own Result abstractions - switch expressions are pretty good at handling these. Personally, I’d write such code in F# instead which is a recent discovery I can’t believe so few know how good it is.

> the underlying lib could change and suddenly throw a Runtime-Exception.

And what would you do in that case? Since this is a future change your existing code presumably wouldn't know what else to do but throw its own exception, so why not just let that one propagate?

Always depends on which level it happens

> Have you read any real world Rust code? Junior developers just add unwrap() until things compile.

If you really don't like unwrap[1], you can enable a linter warning that will let you know about its uses to flag it during code review. You know exactly where they are and when they happen. Exceptions are hidden control flow, so you rely on documentation to know when a function throws.

> Writing exception safe code is a matter of using your language's TWR/disposable/RAII/etc. facility. A programmer who can't get this right is going to bungle explicit error handling too.

Rust has RAII, so you don't have to worry about clean-up when returning errors. This is a Go problem, not Rust.

[1] https://blog.burntsushi.net/unwrap/

> Oh, and sum types? Have you read any real world Rust code? Junior developers just add unwrap() until things compile.

Junior developers will write suboptimal code in any language. So I'm not sure what your point is.

But with Exceptions you can easily implement multiple return types in e.g. Java ;)

I shocked my Professor at university with that statement. After I started laughing, he asked me more questions... still went away with a straight A ;D

As you should, it shows a deeper insight in the language beyond the base course material and out of the box (and cursed) thinking.

When I took the compiler course at university, the professor would have a new coursework theme every year, and the year I took the course, the coursework compiler was exception-oriented. So exceptions were the only control flow mechanism besides function calls. If/else, while, return were all variations of throw.

To me this proved that there's nothing inherently wrong about exceptions.

It's how you structure your code and the implied assumptions you share with your colleagues.

Some people are way too optimistic about their program will actually do.

Happy path programming.

Sounds interesting - any link to the control flow implementation online?

Or how does one optionally throw and exception without an "if" statement? What's the "base" exception call?

Ie if "if" is implemented via exceptions, how do exceptions get triggered?

And is "while" done via an "if exception" and recursion? Or another way?

In OCaml, that's actually a common use for exceptions, as they let you unwind extremely quickly and carry a result (helped by the fact that OCaml doesn't have finally or destructors).

Well, I've also programmed for over thirty years and I wouldn't use a language without exceptions and even wrote a whole essay defending that position:

https://blog.plan99.net/what-s-wrong-with-exceptions-nothing...

> Even junior engineers have a trivial time debugging most go errors

Not my experience at all. I had to do this once. An HTTP request to a production server was yielding a 400 Bad Request with no useful information for what was bad about it. No problem, I'll check the logs and look at the source code. Useless: the server was written in Go and the logs had no information about where the error was originating. It was just getting propagated up via return codes and not logged properly. It ended up being faster to blackbox reverse engineer the server. In a language with exceptions there'd have been a stack trace that pinpointed exactly where the error originated, the story of how it was handled, and the story of how the program got there.

Diagnosing errors given stack traces is very easy. I've regularly diagnosed subtle bugs given just logs+stack trace and nothing else. I've also had to do the same for platforms that only have error codes that aren't Go (like Windows). It's much, much harder.

> Diagnosing errors given stack traces is very easy.

This is the most important aspect of exceptions in my view.

The line that threw the exception isn't even the part of a stack trace that I find most interesting. The part that is most valuable to me when working on complex production systems are all of the call sites leading up to that point.

I remember in my junior years I wasn't a big fan of exceptions. A stack trace would make my eyes glaze over. I would try/catch at really deep levels of abstraction and try to suppress errors too early. It took me a solid ~5 years before I was like "yes, exceptions are good and here's why". I think a lot of this boils down to experience and suffering the consequences of bad design enough times.

Exception usability is definitely an area that needs work. If you work support for a dev platform for a while, it's a really common experience that people will post questions with a copy/pasted stack trace where the message+stack actually answers their question. You can just copy/paste parts back to them and they're happy. There's too much information and not enough tools to digest/simplify them and people get overwhelmed.

Still, better to have too much data than too little.

I can't agree with you about C++ exceptions being worse than useless. Exceptional C++ is worth it. Safety isn't that hard with RAII and ScopeGuard..

In your map example, just add a scope guard that removes the just-added element using the returned iterator if the rest of the procedure doesn't succeed. It's no different in Java.

Haven't seen ScopeGuard before but it looks like an implementation of defer() in C++?

That sort of thing can help yes. But it's still way harder to get exception safety right in a language with manual memory management. In a GCd language you only have to be careful about cleaning up non-GCd resources, whereas in C++ you have to be ready for the expected lifetimes of things to be violated by exception unwinds at many different points and if you get it wrong, you corrupt the heap. Very few C++ codebases use exceptions vs all of them for Java, and I think that's why.

A significant number of C++ codebases use exceptions. Google famously doesn't, but Meta, alike in dignity, does. GDB, now C++, does. At least some AI labs do. C++ exceptions are normal and common.

> Useless: the server was written in Go and the logs had no information about where the error was originating. [...] In a language with exceptions there'd have been a stack trace that pinpointed exactly where the error originated

Go has support for exceptions, not to mention providing runtime access to stack trace information in general. They are most definitely there if your application requirements necessitate, which it seems yours did. Unfortunately, language support only helps if your developers actually speak the language. Go in particular seems especially prone to attracting developers who are familiar with other languages, who insist on continuing to try to write code in those other languages without consideration for how this one might be different and then blame the technology when things don't work out...

Is “goto” just used to mean “bad and evil” here? Because exceptions are not a goto anymore than a return is a goto. The problem with goto is it can jump to any arbitrary place in your code. Exceptions will only go to catch-blocks up the call stack, which presumably you have written on purpose.

> Because exceptions are not a goto anymore than a return is a goto

Not true at all

* goto goes to a specific hard-coded address

* return looks up the previous address from the stack and goes there

* exceptions are a complex mess that require branching logic to determine where to resume execution

Very well put.

I agree about implicit exceptions, but I think that there is a sweet spot with explicit exceptions like Swift (and maybe Java): where you cannot not-handle one, it is part of a function's signature, and the syntax is still compact enough that it does not hurt readability.

I think checked exceptions are noise. They usually never provide value, and are often just caught and rethrown as runtime exceptions to avoid the boilerplate. Scala did right to remove them.

There are several shortcomings with go's error handling. The author heavily lies onto rust, so the alternative is not exceptions but a `Result<T, Error>` sum type.

No stacktraces and error wrapping forces you to not only invent unique error messages. You must also conceive a unique wrapping message at every call-site so that you can grep the error message and approximate a stacktrace.

The weird "return tuple" , which obviously just exists for errors because there is not a single other place where you can use tuples in the language, and the awkward variable initialization rules, make it so that you use the wrong `err` var at some point. E.g. if you want to reassign the result to an existing var, suddenly you have to declare `var err error`, and if `err` already exists then you have to reuse it.

There should be an enum type in go, or instead of the bizarre "return tuple" mechanics exclusive for errors, they should have added a better syntax sugar for errors like rust's `?` sugar. Instead we have something extremely tedious and quite error prone.

> it has a lot of build in opinions which protect you from yourself

It does have opinions, but too often they seem to be there to protect the language from being criticized. Sadly, this works, as marketing (lying) is an important factor towards making a PL popular in today's market.

Exceptions are sum types, they just have different syntactic sugar.

Not really. Exceptions usually imply unwinding the stack, and the ability to catch at any point throughout the callstack. Result types are just 'dead' data.

These are fully equivalent in outcome, though often not low-level implementation. You can use try...catch (called panic...recover in Go) to pack a normal and abnormal return case into the equivalent of a Result<> type. Or just pass an abnormal Result<> back to the caller to manually unwind a single "layer" of the call stack.

> These are fully equivalent in outcome

They are so different in DX, ergonomics, implementation and traceability that I'm not sure this is true other than in the most abstract sense

There is some DX similarity between checked exceptions and Result types.

Because the compiler will fail if you don't explicitly mention each possible exception.

But checked exceptions are coming out of style: They're unchecked in C#, and frameworks like Spring Boot in Java catch all checked exceptions and rethrow them as Spring Boot flavored unchecked ones.

For unchecked exceptions and Result types:

The DX is very different in one critical way:

With Results you constantly have to differentiate between error and ok states, before you proceed. With unchecked exceptions you generally assume you're always in an ok state. It's equivalent to wrapping your whole function body in 'try { ... } catch (Exception e)'. And you can get that with Result types in Rust by using '?' and not worry about doing something half-way.

Ultimately: Are you a happy-path programmer?

>Because the compiler will fail if you don't explicitly mention each possible exception.

But only the first time. Once you add `throws FooException` to the caller signature, the compiler won't complain about any future callees that also happen to throw FooException, even if you did care about handling their exceptions yourself. With callees that return Result you do get to make that decision for every callee.

That's a fair distinction.

I would say "adding Result to the caller signature" provides a similar cascade, but it's not entirely true: Every call must separately be unwrapped. So consistently wrapping your Result calls with the '?' operator is similar to a gigantic try-catch block or adding 'throws CheckedException' everywhere.

So both the '?' operator and adding 'throws CheckedException' everywhere let you accidentally neglect proper error handling: You have a default that is syntactically almost invisible and frees you from thinking. The checked exceptions give you a little more freedom from thinking than the '?' operator.

>So both the '?' operator and adding 'throws CheckedException' everywhere let you accidentally neglect proper error handling: You have a default that is syntactically almost invisible and frees you from thinking.

No. You're choosing to write the `?` there to bubble up the error. It's a compiler error if you forget it. Whereas with `throws FooException` you only get forced to remember it the first time, and after that you're allowed to forget it.

> You're choosing to write the `?`

...but you're not choosing to write the `throws FooException`?

I can only say that if the `?` weren't around and you had to write more verbose handling each time, having the question "How should this one particular error be handled?" might come up more often and might prevent more bugs than when you can carelessly `?`-annotate your way through a dozen lines without any equivalents of `finally { ... }` anywhere.

>...but you're not choosing to write the `throws FooException`?

Now read the rest of the comment after where you stopped reading. Or reread https://news.ycombinator.com/item?id=42249525 I'm tired of repeating myself.

And different control flow, and different or sometimes non-existent types (Java's throws).

Checked exceptions may be implemented as a sum type. Traditional exceptions are more likely to be a single type that wraps up a context object alongside stack trace information.

> The weird "return tuple" , which obviously just exists for errors because there is not a single other place where you can use tuples in the language

MRV, go does not have tuples.

Go is not the only language with MRV (as a special case) and they’re not necessarily bad, iirc Common Lisp uses them as auxiliary data channels and has a whole host of functions to manipulate and refit them.

Go is singularly incompetent at MRVs though, in the sense that only syntax and builtin functions get access to variable arity (e.g. if you access a map entry you can either get one return which is a value, or two which are the value and whether the key is/was in the map). So MRVs mostly end up being worse tuples infecting everything (e.g. iterators needing Iter and Iter2 because you can’t just yield tuples to for loops).

> MRV, go does not have tuples.

> MRVs mostly end up being worse tuples

I think you noticed yourself that you’re getting too hung up on terminology. Multiple return values are a half-hearted, non-reified version of tuples.

No, MRVs can actually offer useful properties and features, that is what they do in Common Lisp. That Go does not do that has nothing to do with MRVs.

f(g(), h()) can't work with MRVs, only f(g()). multiple-value-list, multiple-value-call, and quietly ignoring unneeded values makes them much more usable.

Which is what they said. I'm not sure what point you're making

> MRV

Acronym for Multiple Return Values https://gobyexample.com/multiple-return-values

I've noticed that many people make this mistake with Go and other languages that use MRV (multiple return values). It seems that people who come from a language with tuples have a hard time dealing with the difference or realizing they are different.

It's a plus versus minus type of situation, where both have their advantages and disadvantages, along with a number of workarounds and alternatives that can be used.

In many cases, people can return arrays, hash maps, or structs in place of tuples. That looks like the reasoning in Go, where if the person needs something beyond MRV, they can use the other options available. Go and various other languages, didn't need or didn't want to lean on tuples as much.

> It seems that people who come from a language with tuples have a hard time dealing with the difference or realizing they are different.

The problem is that Go only has the drawbacks of MRV and leverages essentially none of the advantages.

> Go and various other languages, didn't need or didn't want to lean on tuples as much.

As far as I’m concerned go would be a strictly better language if it had tuples (or even just the ability to unpack structs).

All the nonsense about foo/foo2 to handle both single and multiple results would go away, and pretty much the only thing which would be lost is individual addressing of MRVs for named return values, which is easy enough to replace by naming the return struct and going through it.

I forgot:

The tenet "accept interfaces, return structs" is violated all over by returning the `error` interface.

IMO it's okay to make behaviour-exceptions specifically for error handling. Rust for example doesn't really have builtin behaviour exceptions specifically for errors, they're generic to sumtypes and just happen to work well for errors. But then in practice you must resort to thiserror or anyhow helper crates to deal with errors in anything but tiny programs.

If you do make behaviour exceptions for error handling, be honest and upfront about it. Don't say "errors are just values so just use them like regular vars" in docs, if then there are several huge exceptions (tuple-returns and breaking a core tenet). If you make exceptions then you might as well do them right, instead of half-assing them. I believe zig does it right, but I haven't gotten around to try it.

"Do what I say, not what I do" is almost a design guideline of Go at this point, there are so many inconsistencies like this in the language design.

> The tenet "accept interfaces, return structs" is violated all over by returning the `error` interface.

To be fair, that expression came from a blogger who often wrote about Go. It is not a tenet held by the Go project. In fact, Rob Pike has made clear that he considers it to be misguided advice. It is only violated in the same way returning Result in Rust violates the assertion I made up for this comment: Do not return Result.

https://go.dev/wiki/CodeReviewComments#interfaces

How on earth is that violated by error?

error is implemented by types all over the standard library and beyond and consumed by functions that wrap errors in the errors package. It's exactly an example of what you claim is violated.

Even more obviously your link isn't talking about functions but packages. There are some violations out there but generally when included packages define interfaces they are ones that get consumed like in io or it's dissolution.

> The implementing package should return concrete (usually pointer or struct) types

It's in the first paragraph. It goes on in the second:

> Do not define interfaces on the implementor side of an API “for mocking”; instead, design the API so that it can be tested using the public API of the real implementation.

And yes, io.Reader/Writer violate that too, because either the tenet is wrong or the design of interfaces in go is wrong.

> Even more obviously your link isn't talking about functions but packages.

It doesn't matter: your exporting errors behind the error interface, not your concrete error implementation. If you're just using one of the many ways to create stringly errors (!) like fmt.Errorf, you maybe don't notice but you are in fact returning interfaces all the time.

The actual first paragraph states:

> This page collects common comments made during reviews of Go code, so that a single detailed explanation can be referred to by shorthands. This is a laundry list of common style issues, not a comprehensive style guide.

The document does not assert that you must not return interfaces or that it is incorrect to return interfaces. It only indicates that returning interfaces at inappropriate times has been a recurring issue found during code review. Sometimes returning an interface truly is the right choice, but when it isn't...

Like most adages in programming, the aforementioned tenet holds validity in many cases, but, as always, "use your noggin" applies.

It does matter that packages and functions are different.

It also matters what the io package actually does. https://pkg.go.dev/io . The io package has a very limited number of functions that return Readers. The vast majority of its functions take Readers or Writers as arguments and do useful things with them: e.g. Copy or LimitedReader. Most of the interfaces it defines (ReedSeeker, ReadWriteSeeker, etc) aren't instantiated by anything in it.

os implements Reader and Writer for filehandles. net does the same for sockets etc.

It doesn't matter with `error` because it's returned everywhere, both functions and from packages by proxy.

I'm not arguing that the tenet should be held true, to be clear. I'm saying that this tenet is misleading. If you can, return a concrete type. If several packages consume the same interface, then you it's not reasonable to define the interface at the consumer because you'd just have to copypaste it.

> I'm saying that this tenet is misleading.

Doesn't that go without saying? There is no tenet that isn't misleading when presented to a general audience. Fair that if you come from a position where you understand the full context and nuance under which the tenet was built then you should be able to free yourself from being mislead, but, of course, this time is no exception.

> If several packages consume the same interface, then you it's not reasonable to define the interface at the consumer because you'd just have to copypaste it.

Where several packages find interface commonality, there is no doubt a set of "primary" functions that roll up shared functionality around that interface. The package of shared functions is understood to be the consumer under that scenario.

Where several packages stumbled upon the same interface without any shared functionality, copy/pasting is warranted. In this case, while the interfaces may look the same, they do not carry the same intent and that needs to be communicated. Another oft-misunderstood tenet, do not repeat yourself, does not imply avoid repetitive code.

I don't quite see when a package is "understood to be the consumer" of... itself? We're talking about other packages importing an interface.

I can give you a concrete example.

I have a "storage" package, that exports multiple storage implementations/backends. Files, in-memory, S3, zip...

Some other packages pick one of the implementations to instantiate, depending on the use case (this is NOT a case of mocking for testing or anything like that).

Most other packages work with a "storage" interface, as they're only concerned with reading/writing to "storage".

So the storage package, or in any case some package, has to export the interface. Otherwise, every consuming package would have to copypaste that interface, which is NOT warranted.

Actual code is always better, but based on the description it seems a bit strange that there would be a single package that exports multiple implementations. Presumably each distinct implementation should be its own package. None of these packages would export the interface.

An additional package would export the interface and provide the common functionality around that interface. Those common functions would be considered the consumer. In fact, the standard library offers an example of exactly this: io/fs.

No, I cannot agree that this would be called the consumer.

Yes, you have technically moved the interface type away from the implementation, but just for the sake of it, without any other upsides. The consumer is still the package that is using and importing this interface type, just from another package now.

That is the beauty of engineering: There is no universal truth, just different tradeoffs. Meaning that you don't need to agree, nor should you even seek agreement. You can and should forge your own path if different tradeoffs are warranted for your unique needs.

But, this is the "idiomatic" approach. The upside is consistency for future readers. Code is for humans to read, after all. Most codebases follow this pattern, so it will be familiar when the next person encounters it. If you have a reason to do things differently, go for it. Nobody knows your problem better than you, so you cannot listen to others anyway.

I am quite curious about what you see in the different tradeoffs you are accepting, though! What has you weighing them in favour?

Sorry but I haven't really seen this pattern anywhere, care to give some examples?

All libraries that I recall ever using always export a single package, including interfaces. I just took a look, and even io exports a bunch of structs along the interfaces they implement. And `error` is like a basic type of the runtime.

> Sorry but I haven't really seen this pattern anywhere, care to give some examples?

The standard library provides examples, including a "storage" example.

> All libraries that I recall ever using always export a single package

And now there are questions around the language being spoken, so to speak. That's the power of idioms – it avoids the reader needing to ask questions when encountering language that is new to them. But idioms are not the be all, end all. Sometimes they just don't fit. And if that's the case in your situation, no problem. Nobody knows your problems better than you. To listen to someone else tell you how to solve your own problems is foolish.

That is why I earlier wished we had seen some real code. Perhaps then we would understand the nuance that has lead to you choosing these particular tradeoffs. We have no sense of what problems you are actually trying to solve, and questions to try and glean greater insight have been left unanswered. As a rule, though, an overarching interface package with consumptor functions along with multiple implementation packages is preferable because then it avoids a lot of the questions developers are going to start asking.

For example, with the alternative suggested, what if I want to add a new storage adapter that conforms to your interface? Do I need commit rights to your package or should I create my own package that satisfies your exported interface? If I create my own package, why is it the lone implementation in its own package while the others are all rolled up in one package? Is it that you don't want third-party implementations for other services not covered by your package? Why is that? On and on...

If you stick to the idioms, those questions are already answered by convention.

I prefer the structure of Rust errors as it’s fully typed, I don’t like that you can chain them though. It’s a cool feature but it leaves you with some of the same issues exception handling does when the freedom is used “wrong”.

> The weird "return tuple" , which obviously just exists for errors because there is not a single other place where you can use tuples in the language

Go functions also accept a tuple on input. While theoretically you could pass an error, or a pointer to an error for assignment, it is a stretch to claim it is for errors.

Yes exactly, that rather useless feature just makes the whole thing even weirder.

A function that passes tuples, or multiple arguments as they are more commonly referred to as, is not particularly weird, and has pretty much been the norm since the addition of functions to programming languages. But you are right that languages whose functions only accept tuples, but not return tuples, is a strange curiosity. In practice, you end up with developers packing multiple, unrelated items into a single variable to work around the limitation, which then questions why not do the same for input?

I meant weird/useless in the sense of that it's used virtually nowhere in practice.

There is a nice func Must(value, err) { if err { panic(err) } else { return value } that you can use in tests to get around the inane regular error handling. But that's about it, to my knowledge. Sad that there isn't more to it. If literally everything returns error as last return value, there could well be some syntax sugar there.

I see it used often, even in languages that don't formally support multiple return arguments – with some hacked up array/object single value return to try and emulate what would be better represented as multiple return arguments.

Like I mentioned in another comment, Go does seem especially prone to attracting developers familiar with other languages who insist on trying to continue to program in those other languages even after working in a Go codebase. That may create a condition whereby the developers you regularly come across don't have a good grasp of where to use multiple return arguments effectively, and thus end up avoiding them, even where they would be appropriate.

With respect to the syntax sugar, there have been a number of proposals for exactly that. While the proposed syntax itself has been well received, nobody has come up with a good solution for the rest of the problem. Syntax is only about 10% of what is needed, of course. Rust, for example, has well defined traits and other features to go along with its '?' operator to fill in the remaining 90%. Presumably there is a good solution out there for Go too, but until someone proposes it...

The pragmatically "right" choice is to have some tuple type built-in. Because they are not only very good for function in/out, they can be used in many more places. Lists of tuples, tuples in structs, etc.

If the language doesn't have tuples, then you have to "roll your own" and emulate them every single time, but it's not a functor so you can't do all the useful stuff.

Go didn't do the pragmatically right choice, because it's only right from a type perspective. But go doesn't care about types at all, they're just a side effect. Go only needed to return multiple values, so that you don't have to pass in one (or more) "out" pointer as argument, and check an errnum. So go is right but only from its narrow perspective: it's better than C.

> The pragmatically "right" choice is to have some tuple type built-in.

Perhaps, but in practice we end up with these ill-conceived languages that support tuples but end up not embracing them. Consider, for example, this Rust function:

   fn process_tuples(tuple1: (i32, i32), tuple2: (i32, i32), tuple3: (i32, i32))

If it made the "right" choice a function would only accept a single input value, which could be a tuple:

   fn process_tuples(tuples: ((i32, i32), (i32, i32), (i32, i32)))

But maybe that's not actually the "right" choice? What if there is a pragmatic reason for functions to have their own concept for tuples that is outside of a tuple type? I will admit, I would rather work with the former function even if it is theoretically unsound. It does a much better job of communicating intent, in my opinion.

And once you accept the theoretical unsoundness of a function having its own tuple concept for input, it naturally follows that it needs to also extend that concept to returns. After all, the most basic function is an identity function. What would be super weird is not being able to write an identity function any time the number of inputs is greater than one.

Go needed to allow multiple outputs because functions can accept multiple inputs.

> Go needed to allow multiple outputs because functions can accept multiple inputs.

I don't think that's the right conclusion, unless you have any source or insight?

It would explain why you can directly cast the multi-return-values into parameters when calling, but... that doesn't seem to fit go at all.

What I would think is that "go needed multi-return to be able to return errors along values", and the mentioned feature is just some vestige. It's not like go ever implements anything for the sake of consistency or correctness.

> I don't think that's the right conclusion

It is the "right" conclusion in general. Multiple input arguments without multiple output arguments is frustratingly awkward. I can understand the appeal of your suggestion of only accepting one input and returning one output, where the passed value might be a tuple, but I expect there is good, pragmatic reason why virtually no languages, even those with a proper tuple type, haven't gone down that road. Once a language accepts multiple arguments, though, it needs them in and out. Otherwise you can't even write an identity function. If you can't write an identity function, is it even a function?

> "go needed multi-return to be able to return errors along values"

If there was some specific reason for its addition, beyond the simple fact that it would be crazy not to, it was no doubt for map presence checks. Without that, it is impossible to know if a map contains a given key. Errors can be dealt with in other ways. It wouldn't strictly be needed for that purpose, although it turns out that it also works for that purpose, along with a whole lot of other purposes. Not to mention that we know that the error type was a late-stage addition, so there is strong evidence that errors weren't a priority consideration in the language design.

With conclusion I meant you claiming that go returns multiple values because a function also takes multiple values. I did not refer to judging this as worse or better than having tuples as general types instead.

> it was no doubt for map presence checks. Without that, it is impossible to know if a map contains a given key

Are you just making stuff up on the go? A map access is not even a function call. And if you need a presence check in the early language development stage, it's much simpler to add support to check presence and then get the value, since it's not threadsafe anyway.

> With conclusion I meant you claiming that go returns multiple values because a function also takes multiple values.

That remains the most reasonable explanation. There is no legitimate reason for functions that accept multiple arguments but only return one argument. That defies the very nature of what a function is.

C made a mistake, but that is not a good reason to copy it. It also made a mistake around memory management. You wouldn't copy that in a modern language either, would you?

> A map access is not even a function call.

Theoretically it is, albeit with unique syntax. However, importantly, it establishes the concept of multiple returns in the language. Once the syntax for multiple returns is worked out, there is no reason to not extend that to functions proper.

But the most likely reason it was added was not because of some specific reason, but because it would be strange without. It is not like Go is the only language that has that feature. It was already an accepted paradigm.

Rust, and other modern languages, have tuples. So a single return value is effectively sufficient.

I refuse to believe that go added MRV for some esoteric reason. In practice it just makes up for the lack of tuples and sumtypes (errors mostly). It also would stand sharply against every other design decision (or accident) that go has made.

MRV also lacks functors in contrast to tuples.

Rust and Go's lack of stack traces are basically equivalent in that you need to call an additional function to add the stack context to the error result. For go you use fmt.Errorf, in Rust you use .context from anyhow (bad practice in many contexts IMO) or .inspect_err + log. It's rather unfortunate that neither has an easy way of capturing a line number + file easily and appending it to the context. Go could easily do it, I think. Oh well.

I agree that Go should really have an analogue to Rust's `?`, but you can't really do that in a sane way without sum types to represent your conditions. The very multiple-return style error propagation makes it impractical to do.

IMO Go should add really basic tagged unions and rework the stdlib to use them, but such a change would probably necessitate a v2.

RE: Golang v2, they clearly said they will not do it and will double down on backwards compatibility with exceptions powered by env vars and/or CLI switches.

I'm fully aware of that.

Technically, Go v2 signified the transition away from Google control to the project being directed by the community. That happened several years ago. Go v2 is already here and has been for a long time. The stdlib is also at v2 now (e.g. math/rand/v2).

You must mean the language? They said that a language v2 (go2) is probably unnecessary – that any future additions could be added without breaking the existing language. I don't expect simple tagged unions would need to break anything. A v2 (or even v3, perhaps) stdlib would be necessary to take advantage, like the parent suggests, but that has never been ruled out and is already the status quo.

This was what I was referring to. The stdlib is what would need to see backwards-compatibility-breaking changes, not the language itself.

I am not sure how would that work? F.ex. how would you introduce tagged unions and make the language 100% backwards-compatible... but not the stdlib?

I admit I have no idea.

The stdlib would remain 100% backwards compatible, but the implication was that he would want to see certain existing features of the stdlib amended with modified versions that leverage the new tagged unions. He imagined that modification would necessitate v2 stdlib packages to maintain sensibility.

Oh. Silly me, I am ashamed for failing reading comprehension so badly.

Thanks for clarifying, that makes sense.

> Go should really have an analogue to Rust's `?`, but you can't really do that in a sane way without sum types

It could be just some simple (hey, that's what go wants to be, right?) macro thing, that just does the everyday `if err!=nil{return ..., err}` for you without having to juggle (and think about) vars.

    b := g(f()?)?
    // var b B
    // {
    //     var a A
    //     var err error
    //     a, err = f()
    //     if err != nil {
    //         return *new(A), *new(B), err
    //     }
    //     var b B
    //     b, err = g(a)
    //     if err != nil {
    //         return *new(A), *new(B), err
    //     }
    //     // no accidental reuse of err
    // }

I mean look how much utterly useless noise this is, and count all the opportunities for mistakes that wouldn't get caught by the compiler.

For a more corporate driven and controlled language, once it goes down a certain direction, it appears really hard for it to stop the train or get enough momentum to change directions. Easy examples are generics, enums, sum types, etc...

Therefore other languages (influenced by Golang), like V (vlang.io), Odin (odin-lang.org), Jai, C3... made such improvements. One could argue that it's more a matter of if people are going to be brave or open-minded enough to try something different or go in a different direction.

> I agree that Go should really have an analogue to Rust's `?`, but you can't really do that in a sane way without sum types to represent your conditions. The very multiple-return style error propagation makes it impractical to do.

There is always the Odin style 'or_return' operator, which is defined for a similar situation.

https://odin-lang.org/docs/overview/#or_return-operator

What is quite interesting (after looking at their documentation), is that V lang[1] has all that is mentioned: `?`[2], `or`[2], sum types[3], and can return multiple values[4].

[1]: https://vlang.io/

[2]: https://github.com/vlang/v/blob/master/doc/docs.md#optionres...

[3]: https://github.com/vlang/v/blob/master/doc/docs.md#sum-types

[4]: https://github.com/vlang/v/blob/master/doc/docs.md#returning...

There was a proposal for a `try`, which I still think should have been adopted.

Under that proposal, `someComplicatedExpression(try(functionReturningError()))` would be converted to `foo, err := functionReturningError(); if err != nil{return zeroValue, zeroValue, err}; someComplicatedExpression(foo)`

You would need some form of compile-time reflection/specialization to implement that properly (what if the second return value isn't an error? What if there's only 1 return value?). Further, you would lose the ability to add context to the error branch via fmt.Errorf, which seems rather critical to understandable error conditions.

I'm not sure I would be satisfied with any implementation of try as the language is now, I assume the Go language team would probably feel the same.

https://github.com/pkg/errors provides stack traces support. This unfortunately did not get included in the Go1.13 release when error wrapping was introduced.

Alef actually has/had real tuples, as does Limbo.

Looking at Alef, apart from using tuples to provide multiple return values in a C-like language, they don't seem to actually add much functionality over what Go has without them. One of the few extra things is the ability to pass a tuple over a channel, however passing a struct is more or less equivalent.

I've not looked in much detail at what Limbo could do with its tuples.

So maybe they don't really add that much, hence why they were not carried over when Go was created?

Alef had enums, Limbo has something like the Go constant scheme and iota. Limbo also had tagged unions / varient-records, in a sort of Pascal like fashion - the "pick adt".

> I’ve previously spoken about my loathing of exception handling because it adds a “magic” layer to things which is way too easy to mess up.

I kind of see your point. In this very moment, it doesn't matter whether I agree. What I don't understand, though, is why (typically) people who abhor exceptions are among the fiercest defenders of garbage collection, which does add a “magic” and uncontrollable layer to object destruction.

Personally, having learned to love RAII with C++, I was shocked to discover that other languages discarded it initially and had to add it in later when they realized that their target developers are not as dummy as those choosing Golang.

How does RAII works in concurrent systems ? It seems to me you need to add compile-time object lifetime evaluation (as in rust) which so far incurs a high toll on language complexity.

The exact same way it works in Rust. C++'s RAII works the same as Rust's Drop trait. The object is released when it goes out of scope, and if it's shared (e.g. Arc), it's released when the last reference to it drops.

Different kind of magic. Needing to account for every single line of code being able to throw an exception is very mentally taxing, whereas the existence of a GC removes mental load of needing to account for every single allocation.

Without fail, every single person I’ve seen rave about go’s error handling compares it only to exceptions as if that’s the only alternative.

On the flip side I have yet to find a person who’s familiar with sum types (e.g., Maybe, Option, Result) that finds the golang approach even remotely acceptable.

You hit the main gripe I have with Go, its types system is so basic. I get people raving type-correctness of Go when they come from Python but the type system in Go is simply pre-historic by modern day standards.

Go’s type system is not even impressive compared to python’s.

Do you have a pydantic equivalent in go? Also modern typing in python is starting to be OK to be honest (well, if you consider typescript typing OK), so it isn't really a knock on Go :)

> Do you have a pydantic equivalent in go?

I've been working on one [1].

But gosh, does go make it hard.

[1] https://github.com/pasqal-io/godasse

Well I was comparing to python codebases before they added type annotations

Which, sadly, is still the case of too many dependencies.

While I much prefer Python as a language, Go wins against Python by having a fresher ecosystem, with a higher baseline for type safety. Still pretty low with respect to Rust or mypy/pyright with highest settings, but much better than any of the Python frameworks I've had to deal with.

I feel that the future for Python people who want type safety will eventually be TypeScript on nodejs. Go was intended as an alternative to C++. It seems that in reaction to the ungodly complexity of C++, the creators wanted to avoid adding language features as hard as possible. If the user could work around it with a little extra verbosity, it'd be ok. I feel they removed too much and maybe not the right things.

FWIW... WebAssembly has Option and Result, and adapters for Go.

What do you mean by this? WebAssembly is a low level bytecode which only defines low level types. WebAssembly doesn't "have" types any more than x86 "has" types right? Or have I missed something?

Ah, sorry. Shoulda said: WebAssembly component model.

Oh, I see. Thanks!

That’s to be expected since it is marketed towards beginner and casual programmers.

I don't agree that that's what it's marketed towards, but it was designed with those in mind. That said, experienced developers can enjoy it too, as code is just a means to an end and code complexity or cleverness does not make for good software in the broader sense of the word.

It's a Google solution to Google scale problems, e.g. codebases with millions of lines of code worked on by thousands of developers. Problems that few people that have an Opinion on Go will ever encounter.

Here, you've found me.

I don't LIKE it, but it's acceptable.

I have been working with Rust since 2015 (albeit not professionally, but a lot of side projects) and love it.

But I also dabbled into go the last couple of months and while it has its warts, I see it as another tool in the tool-belt with different trade-offs.

Error handling is weird, but it's working, so shrug

Sorry, balanced opinions are not welcome in discussions about favourite programming languages.

I dislike sum-type based error handling. It is annoying syntactically and only really doable with lots of high-level combinators, which in turn hinder debuggability.

Have you tried the approach that Zig has, or the approach that Rust has? They are easy to debug and do not use any crazy stuff, just simple syntax like `try x()` (Zig) or `x()?` (Rust)

Yes and that syntax sucks.

That sounds like your opinion rather than a "it sucks because it encouraged bad practices" or something

if err != nil { return err; } is an antipattern. Codifying it through builtin constructs does not change this, but instead encourages such poor error handling. That's why try sucks.

I'm dealing with Rust based error handling now and nesting `match` statements is not exactly very happy. Or having to pull in `anyhow` and dealing with its awful awful documentation.

I think it’s because Go is an alternative to Java and C# more so than an alternative to Rust. It is for me at least. As I said, Rust isn’t seeing any form of real world adoption in my region while Go is. Go isn’t replacing C/C++ or even Python though, it’s replacing Typescript, C# and Java. Now, there are a lot of good reasons as to why Go shouldn’t be doing that, a lot of them listed in the article, but that’s still what is happening.

As I pointed out I think Rust does it better with its types error handling. That isn’t too relevant for me though as Rust will probably never seem any form of adoption in my part of the world. I think Zig may have a better chance considering how interoperable it is with C, but around here the C++ folks are simply sticking with C++.

> Now, there are a lot of good reasons as to why Go shouldn’t be doing that

I disagree. Typescript, C# and Java are terrible languages (as are Python/Ruby/etc. in other ways). Golang is bad by OP's standards but there's nothing wrong with it gaining ground on those languages.

Besides it's also easier to convert a codebase to Rust from Golang than Typescript or C#/Java.

Rust and C# have far more overlap than Go could ever hope for. Go is limited (and convoluted sometimes due to "solutions" devised to cope with it) so it is easily expressible in languages with better type systems and concurrency primitives.

It was meant more as an observation than my opinion. I would pick Go over Java/C# any day of the week, but it’s not like talented JVM engineers won’t run in circles around you as far as performance goes.

I’d frankly pick Python for most things though. It’s a terrible language, everyone knows it’s terrible but it gets things done and everyone can work with it. I view performance issues a little different than most people though. To me hitting the wall where you can no longer “make do” with C/Zig replacements of Python bottlenecks means you’ve made it. The vast majority of software projects will never be successful enough to get there.

C# is the best designed lang out of the top10 most popular langs

Technically, if you read the origin story around Golang, it was made to be an alternative to C++. Some departments at Google were going crazy with a myriad of problems involving using C++, so wanted something less troublesome and easier to use.

However, I do agree that it could be argued that Golang has become much more of a Java and C# alternative. Looks like Rust will be shoehorned into C++ and C spaces, but it remains to be seen how successful it will be as an C alternative. Seems like many fans of C are quite resistant or even outright dislike Rust.

For me, the issue with error handling is that while errors are explicitly stated, they are often poorly handled. Rarely have I seen the handling of multiple reasons for why an error might occur, along with tailored approaches to handle each case. This is something very common in older languages like Python or Java

As a regular Go user, I agree with this take. Though the tools exist, error wrapping and checking (with errors.Is and so on) is actually pretty rare in my experience.

Positive example of good and appropriate usage here: https://github.com/coder/websocket/blob/master/internal/exam...

This is down to developer style and agreements though; Go has typed errors and a set of utilities to match them [0]. Not using those is a choice, just like how in Java you can just `catch (Exception e)` after calling a dozen methods that might each throw a different exception.

[0] https://pkg.go.dev/errors

Interestingly, every time (and I mean _every_ time) that I've tried to use `errors.As` on errors raised by lib code, I found out that the lib just went with "nah, I'm just going to use `errors.New` or `fmt.Errorf`", which makes the error impossible to match.

So... I'd say that this is a fumble in the design of Go.

%W exists to solve this

How? Stringly matching? That's not typesafe at all.

No it wraps the underlying error

But the underlying error stays unmatchable. Doesn't sound like a solution if you have to duplicate every error type, and worse, they don't even map 1:1 but now you have the same underlying error wrapped to god knows how many different errors.

For example, the lib produces some an error "bad file descriptor". You'll be wrapping it when you call fileOpen, fileDelete, etc etc 20 times. So you will be wrapping it in "open error", "delete error", etc, 20 times. You cannot try to match it to "bad file descriptor", that information is lost, you now have 20 relatively useless errors.

Except if you stringly match.

https://pkg.go.dev/errors#Is and https://pkg.go.dev/fmt#Errorf clearly state that there is a way to match these errors if the package exposes the values, which the stdlib does.

Ah okay, it seems to work because errors are pointers, and errors.Is just checks for equality.

Which only works when the error itself doesn't contain any information.

It's better than nothing, of course.

It would, if people actually defined new errors. But, as I mentioned in my message, people just use `errors.New` or `fmt.Errorf`, all the way down, so you end up with errors that cannot be matched. Which means, in turn, that if these are errors that should be handled properly (and not just propagated/logged), they cannot.

this part of error handling is pure religion. it goes even against one of the most basic go tenents. that code should be easy to read not write. Try reading and understanding the logic of a particular method where 75% of the lines are error noise and only 25% are the ones you need to understand what the method does. yes it's noise because whenever read a codebase for the first time you are never interested on the the error edge case. first glance readability needs to tell you what you are trying to accomplish and only after what you are doing to make sure that is correct.

on this point go's error handling is a massive fail. Notice that I'm not saying explicit error handling is bad. I'm saying the insistence that error handling needs to be implemented inline interleaved with the happy path is the problem. You can have explicit error handling in dedicated error handling sections

Do you have examples for the latter?

the most basic example was the declined proposal https://github.com/golang/proposal/blob/master/design/32437-...

Some people didn't like the "try" keyword it reminded them too much of exceptions, some people didn't like that they couldnt see a return inline (which was the purpose of the proposal in the first place).

it's not that there are no solutions. the main problem is the go team's insistence to have "one true way" (tm) of doing something and unfortunately this gap between people who want to see every return inline and people who want to see the clean solution separate from the error handling is not something that can be bridged by technical means. the only solution is to implement both ways and lets see which one wins.

This doesn't look meaningfully different from current error handling in Go.

It's basically the same syntactic sugar as `try!` in Rust, isn't it?

> yes it's noise because whenever read a codebase for the first time you are never interested on the the error edge case.

maybe this has something to do with how bug-prone it usually is for a new hire to modify a codebase for the first time in most orgs

you could also just do things fail-fast style and panic everywhere if you REALLY wanted to stop inlining error conditions. or ignore error checks until some sort of guard layer that validates things.

IME you usually don't want to do either of those things, and the go approach at least encourages you to think about it closer to the site than checked exceptions (which you can more easily toss up and up and up and auto-add to signatures of callers). unchecked exceptions are arguably less-bad than "just ignore go return errors" - they'll get seen! - but terrible for a reliability/UX perspective.

optional-esque approaches are nice but just a different flavor of the same overhead IMO.

I see a lot of people say this about exceptions, and I don't have that problem. The exception bubbles up the stack until something catches it. Ok it's a different code path, but it's a very simple one (straight up). So you either catch the exception nearby and do something specific with it, or it bubbles up to a generic "I'm sorry there was a problem please try again later" handler.

Honestly makes me wonder what I'm missing. Maybe it's because I don't deal with state much? Do the problems start to mount up when you get into writing transaction locks, rollbacks etc? But I don't see why you wouldn't have the same problems with Go's mechanism.

Hoping to gain enlightenment here.

[copied from a comment below]: They are just horrific gotos that any library can invoke against your code. They are pretty much never, ever handled correctly. And nearly always, after an exception is “handled”, the application is actually in an unknown state and cannot be reasoned about.

Maybe this is it? I prefer a "fail early and often" approach, and I tend to validate my data before I perform operations on it. I don't want to attempt recovery, I want to spew log messages and quit.

> They are pretty much never, ever handled correctly. And nearly always, after an exception is “handled”, the application is actually in an unknown state and cannot be reasoned about.

I think that’s misdirected but illustrates the emotional reasons why people develop a negative impression of the concept. Usually it means someone had bad experiences with code written in a poor culture of error handling (e.g. certain Java frameworks) and generalized it to “exceptions are bad” rather than recognizing that error handling isn’t trivial and many programmers don’t take it seriously enough, regardless of the paradigm. As a simple example, C and PHP code have had many, many security and correctness issues caused by _not_ having errors interrupt program execution where the users would have been much better off had the program simply halted on the first unhandled error.

If you write complex programs with lots of mutable shared state, yes, it’s hard to reason about error recovery but that’s misattributing the problem to the mechanism which surfaced the error rather than the fact that their program’s architecture makes it hard to rollback or recover.

Exceptions are difficult to discuss because different languages implement exceptions differently, each with their own downsides. That said, I don't think anyone has an issue with bubbling. Even sum type proponents love Rust's ? shorthand, because it makes it easier to propagate Results up the stack.

The big issue with exceptions in C#, Python and JS is that they're not included in function signatures, which means you have to look up the list of possible exceptions from documentation or source code. This could be amended with checked exceptions like Java, but it allegedly doesn't mesh well with the type system (I haven't personally written Java to confirm this). And then there's the C++ crowd that slaps noexcept on everything for possible performance gains.

Personally, I like the way Koka does exceptions with algebraic effects and type inference. It makes exceptions explicit in function signatures but I don't have to rewrite all the return types (like in Rust) because type inference takes care of all that. It also meshes beautifully with the type system, and the same effect system also generalizes to async, generators, forking and other stuff. Alas, Koka is but a research language, so I still write C# for a living.

> Go is seeing adoption that no other “new” language has exactly because of its simplicity

Yes - for me, the simplicity is essential. As a part-time programmer, I don't have months to spend learning C++ or Rust.

If my project needs to compile to small(-ish) standalone binaries for multiple platforms (ruling out Python, Ruby, Java, C#, etc) what simple alternative language is there? Plain C?

C# can compile standalone binaries for multiple platforms.

Basic Rust doesn’t take months to learn, especially when you’re not trying to do things like distributing crates to other people. I found the compiler to be enough more helpful than Go’s to make them roughly time-equivalent for the subset of common features, especially for a simple CLI tool.

Go's error handling is not explicit because error is an interface. So, a nil MyError is not the same as a nil error, and error comparison is inconvenient because the interface constraints are so loose that people often resort to checking errors with strings.Contains.

Rust and Go are very different and I feel people want a middle ground that just doesn't exist currently.

A garbage collected relatively simple language that compiles into a statically linked binary but has a type system similar to rust, rest types etc.

Syntactically, Gleam and Kotlin come somewhat close but not really. I like Rust but I do believe it is too complicated for many people who are capable of creating something but are not CS grads nor working as programmers. If you're only gonna use the language every once in a while you won't remember what a vtable is, how and when things are dropped etc. I understand that "the perfect language" doesn't exist but I think both Go and Rust brought amazing things to the table. I can only hope someone takes inspiration from both and creates a widely usable, simple programming language.

> A garbage collected relatively simple language that compiles into a statically linked binary and has a [good] type system

Yeah! Pattern matching too. What are currently available languages closest to this? AFAIK, Gleam relies on a virtual machine, but otherwise seems promising.

Stretching 'currently available' a little, there's Roc lang [1]. Though still in development, you can use it for small personal projects. Another caveat is that it's a functional language, which could potentially hinder its wide adoption

[1] https://www.roc-lang.org/

Try F# with the new AoT compilation option and publish single file switch.

> A garbage collected relatively simple language that compiles into a statically linked binary but has a type system similar to rust, rest types etc.

Swift.

If they'd drastically improved their tooling then yes.

But sadly it's not that easy to create a statically liked binary in swift. The last time i did it it also included the whole runtime, and the resulting "hello world" binary was 50mb large. Annoying at least.

For years I wished they got their stuff together, but at this point I'd be very suprised. They probably have too much technical dept already due to the support of XXX edge cases Apple need for iOS/MacOS development.

> For years I wished they got their stuff together, but at this point I'd be very suprised.

This is a thing that's having work done, but still too large depending:

https://www.swift.org/documentation/articles/static-linux-ge...

See here for sizes of various languages as of August 2024 when I'm commenting, but this is Swift 5.10 for Windows:

https://github.com/MichalStrehovsky/sizegame

Build actions here:

https://github.com/MichalStrehovsky/sizegame/blob/master/.gi...

What you want, instead, is Swift 6 for Linux, while e.g. Alpine, as of summer, still too large:

https://mko.re/blog/swift-alpine-packaging/#:~:text=Package%....

In this thread on Swift for AWS Lambda, a dev gets a Linux static build to 5.9 megabytes:

"Then, I stripped it and it's now 5.9Mb. That's an impressive 86% reduction. Given that musl libc.a is 2.4Mb and libc++.a is 10Mb, I find that 5.9Mb for an executable that contains both libc and the Swift runtime is not that bad :-)"

https://forums.swift.org/t/aws-lambda-functions-and-the-linu...

> They probably have too much technical dept already due to the support of XXX edge cases Apple need for iOS/MacOS development.

See also embedded Swift:

"Apple explains that Embedded Swift is a subset of the Swift programming language with a much smaller footprint (the binary can be around 10KB..."

https://www.cnx-software.com/2024/06/13/embedded-swift-esp32...

Good that they work on it. And the efforts to refactor foundation into smaller modules go in the same direction.

I continue to root for them but they need so many improvments to be a realistic alternative to python/typescript/go/java for regular backend development.

At least in my field where we need to sure that other developers can continue the development without much hassle.

Just as an example, with those languages it's almost trivial to setup a basic projekt with dependencies. In swift however, it takes careful reading ot the swift package manager documentation to understand their concept of "products", "targets", etc. And I'm pretty sure i'd have to start from zero when i use swift for the advent of code in a few days.

There are many problems with devepdencies in Python, or package.json/gradle files. But at least they are (almost) foolproof to get started.

> relatively simple

Kotlin is interesting as a middle ground, but I still find it much less productive than Go for most tasks, and unsuitable for tasks where you'd reach for Rust.

In practice, Kotlin is extremely complicated, and you end up spending time being clever. There are 1000 ways to do things. Operator overloading. Proxies. Properties. Companion objects. Exceptions AND result types...

The build system (assuming you use Gradle) is tantamount to torture for anyone used to "go build".

The coroutines APIs feel simultaneously more complicated and yet more restrictive than Goroutines. More structured but less flexible and more effort to use.

Access control feels awkward. There's no way to make a type package-private -- it's file-private or available to the whole module. This leads to either a larger API surface than desired, or the inability to break up complexity into multiple files.

Kotlin/Native and Kotlin/JVM really are two different beasts too.

Kotlin/JVM is mature, but then you are running on the JVM, so that cuts out a whole class of use cases you might bust out Rust for.

There is a very weak ecosystem for Kotlin/Native, and it's poorly documented. There are some scary bugs in the bug tracker.

You can't publish source-only libraries for Kotlin/Native either, so you need a complex CI setup to build binaries for every OS and arch under the sun. Or just don't publish libraries at all, which probably feeds in to the weak ecosystem...

Don't forget that it's made by someone trying to sell you an IDE!

> A garbage collected relatively simple language that compiles into a statically linked binary but has a type system similar to rust, rest types etc.

You just described Ocaml and ReasonML (which is Ocaml with Go-like syntax).

Last time i tried to install ocaml on windows few months ago i failed to do so, it's a well known thing that it is not 100% supported on windows therefore it wont have the adoption that go/rust has (as it's been the case forever now)

This doesn't count because the idioms are very different from Go and Rust. And I suspect there's a high learning curve for features like Functors.

If you check the Wikipedia page for OCaml to find out where it gets used, you'll see why it's ocaML. That is, you'll notice that it's mostly a MetaLanguage, or a language for writing other languages. The same observation applies to other languages in the ML family.

(author here) in which ways does Gleam come short of that? Because I'm also looking for that middle ground and I was very curious to get a look at Gleam.

IMHO it's just that it's a beam VM language, which is a fatter runtime/ecosystem than is really needed to achieve the goal stated above can bring it's own bag of problems (but also it's own superpowers).

Also to be productive you have to utilize the rest of the erlang ecosystem, so at least some superficial knowledge in elixir & erlang is helpful for for some use-cases.

Syntactically I actually don't think it's that for off, but I dunno what GP was thinking, maybe that it leans more into functional patterns & sugar for those whereas rust/go can also be used in a procedural style. (Though at least personally I am using way more functional patterns in rust than I expected)

Trying to come up with a reply made me realize something. I was initially going to write something like "it's just not ready yet" but paused for a moment and reflected on that. That's actually not a real reason why it would come short of that. And instead of talking about how something "is not quite ready yet" we can instead choose to invest into it and use it anyway, with the goal of contributing and helping make it "ready" and more widely used.

Out of everything mentioned, I think Gleam is the most likely to succeed and what I think I will invest in as well. But that's not just because of the language design but rather because of their approach to community and community building! It's a very very nice extra that the language/syntax also happen to be what I like and have described in my previous comment.

There is Crystal and Nim. With especially Nim, there is GC and generates c in the end.

I love Crystal, but the lack of a proper LSP and tooling makes it hard to just jump in and adopt for bigger projects.

Unfortunately, Nim's BDFL is... not known for playing well with others. Perhaps the Nimskull fork will grow into something widely useful?

https://github.com/nim-works/nimskull

That's total BS btw, I've been part of the Nim community for over 7 years now and never had an bad interactions with the project's lead. But I did witness behavior from certain individuals that was out of line. Some of them went on to blame Araq in other places, I have read their criticism, which I found very hypocritical. You should take their views less seriously.

> people want a middle ground that just doesn't exist currently

https://borgo-lang.github.io/

Rust syntax, compiles to Go.

I think Inko (https://inko-lang.org/) has the potential to be that language with some tooling/adoption/maturation

Imho, most features of Rust that people would like to see in Go would still fit into the "concept" of Go. Just like they added generics, they could add just three things: A generic container for errors (Result), one for saner Nil handling (Optional) and a small sprinkling of syntax sugar to make these comfortable to work with (something like an elvis operator equivalent).

Go has the one big advantage that is almost solely responsible for it's success: It was created and directly used by a giant company that could afford to create amazing tooling around it and develop great opensource libraries for it. Already being in use and having libraries feel like the biggest determinants of a languages success.

> A garbage collected relatively simple language that compiles into a statically linked binary but has a type system similar to rust, rest types etc.

So OCaml then (ocamlopt to do native code compilation)

C# and F# will be by far the closest. Other options lack sufficiently good type system or tooling to match either of the two.

Compile to static native binary with 'dotnet publish -p:PublishAot=true' (or add this property to .csproj to not specify on each publish). In the case of F#, you will need to use Console.* methods over 'print*' because print has unbound reflection inside for structural output on "%A" format specifier (it will work most of the time but negatively impacts binary size and causes the compiler to complain).

I can especially recommend F# as "easier more business-focused Rust alternative" because it is expression-oriented, has discriminated unions, full HM type inference and gradual typing is a joy to work with. Data analysis and domain modeling are very pleasant to do in it too.

For systems programming C# is going to be the option to use - it will give you great concurrency primitives, fast (sometimes even zero-cost) native interop, smaller than Go native binaries and a lot of low-level APIs including portable SIMD. Go is often poorly suited for these tasks or can't do them at all (at least without workarounds). There are many new high-performance libraries focused on this domain as .NET gains popularity in non-gaming communities in this area. And of course you benefit from a huge existing ecosystem and won't have to do the all the heavy lifting by yourself unlike in niche languages suggested in sibling comments.

So here is my take on this, once again:

Start with JavaScript. The basic syntax is delightfully direct, it has massive adoption already, the ecosystem is large, the runtimes are getting better all the time, compilation is here with WASM,

Now remove the weird parts (e.g., too much flexibility to redefine things, too much use of the global scope, too much weirdness with numbers, etc.), and add:

- Types (including sum/product types, Result<T>, Maybe<T>, decimals, etc.)

- More functional features (everything-is-an-expression, pattern matching, currying, etc)

- A comprehensive standard library.

Already this starts to yield a language that has the best chance to be what a lot want.

The other major advance in developer tools that I'm wanting to see is revamping HTML to have proper sophisticated controls built-in, controls that can be easily styled with inline CSS. This will reduce the amount of JS needed on the client.

These two things will yield a massive advance in programming productivity, at least as far as web-related development is concerned, IMO

I think this is a really wise take, and one that I honestly haven't seen before. I would certainly try a language like this.

The one thing I would add: Provide a well-trodden path to easily drop down into Rust/C/C++ for performance critical functionality. I have found this to be a big point in Python's favor over Go; you can write things the slow way (ie. executing in the Python runtime), then profile and figure out where to push things down into an extension. Often you're doing something that already exists in a library like numpy or pandas or polars, but if needed, you can write your own extension.

Does WASM enable FFI on its own?

I think here you're pointing at something along the lines of Roc/Gleam?

  Go: "I'm a simple language!"
  User uses Go for some time.
  User: "I hate you, you're a simple language!"

Perhaps it's because I'm 50+, I love a simple language.

I feel the "critique" is not very balanced, and I view judgements that are not balanced as weak, as everything in technology is about tradeoffs.

I of course come to a different conclusion: https://www.inkmi.com/blog/why-we-chose-go-over-rust-for-our...

I wish the discourse that Go is a "simple" language would die.

Despite its veneer, once you start writing Go it quickly becomes apparent that it isn't simple. Hidden complexity and footguns are abundant (e.g., https://archive.ph/WcyF4).

It's nevertheless a useful language, and I use it quite a bit, but it's not "simple".

I have no idea why anyone would say it's not simple, it's super-simple. Learning how duck typing works with interfaces and how to use it is perhaps the only hurdle. In my experience, only certain BASIC dialects like VisualBasic are simpler.

I think the sticking point is what people mean when they say simple. To me, and likely to many saying Go isn't simple, simple is not a synonym for easy.

Go is easy, but it is not simple. For example, solving the problem of generics in a generic way from the start so the same problem can be addressed in the same way everywhere, would be simple, but maybe not (as) easy. Contrast that to giving the runtime/standard library a special exception with maps and lists. That's easy, but not simple. People used to literally use code generators or weird UTF-8 characters to fake generics, that's not remotely simple.

Frankly code generators to fake generics aren’t remotely easy either.

This 100%, I was just about to type a long rant up about this. There are so many weird parts of the language that took me forever to grasp, and in many cases, I still don't have an intuitive grasp of things.

And plenty of other examples that aren't in that article:

- You have a struct with an embedded interface. Does the outer struct satisfy the embedded interface? And can I type assert the outer struct into whatever embedded struct is fulfilling the inner interface?

- When should I pass by value and when should I pass by reference? Like I generally know when to choose which, but do I really know without performing a benchmark? And what about arrays? Should I store pointers in them? But it also seems that people just don't care and just randomly roll the dice on when to return a pointer or a value?

- Shorthand variable declaration. How does it work when you shorthand declare two variables but one of them already exists?

Don't both answering the questions, that's not the problem. The problem is that it's just not intuitive enough such that I'm confident I know the correct answer.

Your thinking is too complex for Go. You might be better of with Rust.

Same about benchmarking, if you want and need the fastest code, or the best memory management, use Rust.

If you need something faster than Python in general but not the fastest, use Go.

But that's the thing right? Like I come from Java. In Java, we have objects. They are pointers. That's it. You don't get to decide on whether you want a pointer or a value (I guess primitives are an exception lol). But it was so simple!

And same in JavaScript. Everything is a pointer except primitives. That's it. End of story.

And I have written Rust too, and while the situation is definitely more complicated there, the guidance is extremely simple and straightforward: If the struct implements Copy, then it is very cheap to copy and you should pass by value. Otherwise, you should pass by pointer/reference.

And meanwhile in Go, I just see pointers and values being used seemingly interchangeably, and seemingly at random.

> but do I really know without performing a benchmark?

Not really. But that’s one of Rob Pikes rules [1], I think the intention is to write whatever is simplest and optimize later. The programmer doesn’t need to remember 100 rules about how memory is allocated in different situations.

[1] https://users.ece.utexas.edu/~adnan/pike.html

I mean it's a great idea, and I fully agree that I do not want to worry about memory allocation. So then why is `make` a thing? And why is `new` a thing? And why can't I take an address to a primitive/literal? And yet I can still take an address to a struct initialization? And why can't I take an address to anything that's returned by a function?

https://go.dev/ref/spec#Address_operators

I am not going to answer the questions, but this is a very strange complaint, to be honest.

For example, passing by value/passing by reference is something covered immediately in the Go FAQ document once and for all. Everything is passed by value in Go, that is it. There should be no confusion at all. If you spend 15 minutes reading Russ Cox's post on the internals of the most common data types, you will also understand what data structures have implicit pointers under the hood.

Well yes obviously I know everything is passed by value, just like in literally every other popular language. I'm talking about the difference between pointer parameters/receivers vs value parameters/receivers.

The first is simply a question about what an embedded interface is. Any way in which you learned about this feature should also answer the question.

The second isn't related to Go.

The third I can see as being a bit confusing, but isn't it something you try once and then remember forever?

Here's one: Creating a new variable via shorthand that shadows a variable in an outer scope - that can be confusing and an easy mistake to make.

But broadly, I would strongly advocate that it is simpler than most other languages, even if it has some quirks. In the same vein of "If I had more time, I'd write a shorter letter", sometimes it takes a bit of time to understand why something is simpler than the alternatives. No language has zero ramp-up. Go doesn't exist in a vacuum, you have to compare it's learning curve and complexity to other languages.

Yes, like "Opening Brace Can't Be Placed on a Separate Line" (from your link).

Everyone can read Go code and understand what happens. There are some minor difficulties like func (*A) vs func (A).,

RiscV assembly is even easier to read by that metric.

My assembler days were 4 decades ago, but

"Everyone can read Go code and understand what happens."

There seems to be a difference between "easy to read" and "understand what happens" - or what happens on what level. The challenge is that there is a tradeoff between the two. Assembler is too low to understand what "really" happens, on the other hand Haskell for example with Monad stacks is again very easy to read + understand what happens "most of the time", but hard to understand all the abstracted away side effects.

In Haskell with

   add 3 5

everything can happen beside what you see.

In assembler

  ld a, 3
  add a, 5

nothing happens except these two instructions.

The tradeoff is how much you want to be explicit, with the downside of creating too much noise, and how much you want to abstract away, with the downside of magic happening somewhere.

Yes. And to be less snarky: Go doesn't really sit on the efficiency frontier here:

It requires you to write a lot of stuff by hand and is incredibly verbose, but it also does a lot of magic behind the scenes.

It's easy not simple, but the consequence is that any complexity that other languages handles for you, in go gets forced onto the developer.

There's more stuff to think about, because the language is doing less for you.

Go is "simple" insofar as you're doing simple things. If you've tried writing a KV store or database (as I have) you'll quickly find yourself wanting slightly more modern language features.

"quickly find yourself wanting slightly more modern language features."

Use the tool that works for you.

Feel free to take a look at any more complicated GoLang code (k8s, gorm, etc) and you'll see that the tool/library you're depending on requires a veritable rats nest of bad practices to work around the Go's inherent limitations.

My opinion on Kubernetes as a yardstick example of Go is this:

This was one of the first large systems developed in Go, not too long after the language hit version 1.0 (ca. 12 years ago). What constituted good Go style and package architecture were not well known at that time even. Given that and hindsight being 20:20, I could easily imagine the internal architecture (not even the public surface) for Kubernetes being a lot different and simpler. The package architecture alone makes me cringe. As a hypothetical counterpoint, I wonder what Kubernetes would have looked like had Dave Cheney (https://dave.cheney.net/) and Rob Pike built it in that era. I think that would have been a better yard stick.

I have a unique appreciation for this first large systems perspective as I co-designed an adjacent product in the cloud native ecosystem at that time (Prometheus). There was no good example to follow when it came to large program structure and design in Go at that time. We were all figuring that out for ourselves — organically.

I think this point about organic evolution of the architecture is important to call out explicitly, because developers often look at an existing structure and essentially mimic it with their additions, changes, and refactors irrespective of whether the structure was correct for the problem it was trying to solve or even good. Given that reality, is it really any wonder that one of the first major pieces of software ended up being this metaphorical mess? And the truth is that's not a language-specific problem: it could have happened with any new language. And taking a legacy system of this age and refactoring is difficult from a social perspective, ignoring the gradient/cost of refactoring that is language-specific. You'll have a lot of people who will oppose structural change just because, so probably a significant refactoring to achieve these goals is just not in the cards.

Had there been so much as something similar to https://google.github.io/styleguide/go at that time (ca. 2012) (it's based on the very spartan https://go.dev/wiki/CodeReviewComments), that would have been tremendously useful and impactful in helping preserve simplicity.

Scheme is a simple language, Go just hides complexity until it blows up in the worst possible way. (Of course, most reasonable alternatives to Go are even worse from that POV. See Python, Ruby, JS etc.)

Very common misrepresentation of any critique of Go. "You just don't get simplicity, you got them Java brainwormz"...

There are many examples in the article that point out the annoying inconsistencies in the language, those are the opposite of simplicity.

I love Rob Pike's presentations on Go, some of them were eye-opening to me. However, I just wish that the Go I see in practice would be much closer to the Go language that Go-fans describe in abstract.

> Perhaps it's because I have experience, I love a simple language.

(fixed the statement to focus on your value, not age per se)

I love language ergonomics above all. Python wins. But for runtime bang-for-the-buck, Go wins.

Go is not simple, it's easy.

The difference, for example is: Go invents an abstraction just for one use case, and just for the standard library/runtime itself, instead of taking the time to create a universal version of that abstraction. Generics for maps and lists, and tuples for error handling.

The problem is that Go is not designed to be a simple language for it's users, it's designed to be a simple language to implement for it's maintainers.

In my opinion, a simple language should be highly consistent (have a few rules, but which are universal and consistent everywhere). Instead we have a language with weirdnesses, inconsistencies and workarounds all over the place.

A good example is type elision: it's available when declaring arrays, slices and maps, but not for structs. Allowing that would have several benefits in terms of readability, and also allow named parameters via anonymous struct arguments (which would greatly improve the design compared to the ugly workarounds that are currently used).

I wonder what makes someone go such a great length to bash a language, any language. I say bashing, because even the few valid points in the post are not written in a constructive style. After all is there a language that can't be criticised?

Is the post written to make one feel better having a failed a project the language? (It's not me, it's the language) Or is it the failure to understand that not everyone thinks / works the same and what one finds unacceptably bothersome, others barely notice? Languages that do not cater for a real need would likely vanish on their own, they rarely need help.

As for Go, despite the differences compared to "more sophisticated" languages it works brilliantly for the projects I've encountered. I hope the author is not forced to work with Go though. For the rest of the community, we keep churning out services, giving our feedback to the Go team and seeing the slow evolution of the language without breaking our stuff in production.

> what makes someone go such a great length to bash ...

"""

Inherent complexity does not go away if you close your eyes.

When you choose not to care about complexity, you're merely pushing it onto other developers in your org, ops people, your customers, someone. Now they have to work around your assumptions to make sure everything keeps running smoothly.

And nowadays, I'm often that someone, and I'm tired of it.

"""

> [Go] works brilliantly for the projects I've encountered.

Of course, C, C++, PHP and JavaScript works too! Of course many many many things "work" in our world. Of course just adding one more lane works too, of course police states work too!

Yet something else would work even more brilliantly?

That quote sounds like an argument against Rust, not GO. Rust is a complex language and that complexity doesn't go away if you close your eyes to it either. All complexity adds to project complexity.

The implicit argument is that there's some irreducible non-negotiable project complexity in real-world software, and the explicit argument in the post is that you can either put it into the language/compiler or keep it in wetware.

The supporting argument in the text is that even simple things like downloading a file can get hairy very fast. Networks are complex, HTTP, IPv4/v6, DNS - it's always DNS! - filesystems, permissions, running out of space, computer runs out of battery, etc.

Probably a better argument would be to focus only on application logic (type system, error handling, syntactic-semantic ergonomics) and show that a complex piece of Go [like the k8s persistent volume controller] can be written in a nicer, more maintainable, safer way in Rust.

Of course not all Go code achieves Kubernetes-level complexity. And I think it makes sense to pick the right tool for the job, which is whatever the team/programmer is productive with. (And here productivity is measured based on how well the resulting code/software solves the business case. And basically if there's no need for Rust-level safety/quality/maintainability, or if there's an explicit need for Go-level cheaper hackability, then that's a great result business-wise.)

The important corollary of this is that business requirements tend to change, yet software has inertia and that's roughly how we ended up with insert famously bad software that got wildly popular unexpectedly (for example Windows, Macromedia/Adobe Flash and the infamous plugin, random low-level crap in appliances, and adjacent to that the Bluetooth stack that got rewritten in Rust for Android).

So - of course - for society it would be amazing if software would be better from the start. (Duh!)

Also, one more thing I think worth mentioning, is that the argument about the lack of information in code at callsites in Go (is something passed by reference or by value), how hard it is to keep boundaries (immutability) is basically an argument for making this complexity up-front and visible. Hence Rust looks complex. (And again, the argument is that it's there even if Go hides it.)

https://blog.rust-lang.org/2017/03/02/lang-ergonomics.html

https://github.com/kubernetes/kubernetes/blob/60c4c2b2521fb4...

Perhaps not everyone likes boring clinical reviews and some people like ones that have a bit of passion and humour in them?

Just because this blog post isn't written in a way you like doesn't mean it doesn't have value to others.

It's just some person's blog and they're having a rant. It's okay, it doesn't have to be that deep.

I would guess the 'why' is because OP feels like they have an opinion that they don't feel is sufficiently represented 'out there'. Indeed, as a not-a-fan-of-go, in 2022 I was confused at go's popularity because it always felt to me to have some pretty glaring shortcomings that were seemingly ignored.

Note that people don't really write big blog posts about PHP being a bad language (anymore?) because that's been done to death.

Speaking a someone who has written Go full time since 2014:

Go is used in a lot of places for unglamorous things where the most important thing is explicitness, and where all these features that the PL nerds want, do nothing but make things more implicit.

It's not perfect, but it's a great language. People who like it, though, like me, are pragmatists. The shortcomings of the language that are apparent in complex code are strengths when writing the simple straightforward code that makes up most Go projects, because all of THAT becomes much simpler than it would be in a fancier language.

The reason I've never bothered to write a response to these kinds of screeds against Go is that usually I'm too busy getting shit done with it.

One very subjective, very irrational factor for my borderline hate for Go is that for years the Go zealots gaslighted everyone about every single part of Go.

Anything that Go did, no matter if it was the most basic implementation or if other languages already did it (better), was essential, the best and only way to solve that issue.

Anything Go did not do was superfluous and downright a conspiracy by Big Complexity to keep us unenlightened Non-Goers addicted to the Syntax Sugar: Things like sane null handling, sane error handling, less boilerplate, generics, or not creating special cases for everything (generics and tuples) instead of developing one cohesive solution.

Even now, in this thread, the strawmanning continues: Error handling is brought up, and instead of admitting the unmistakable truth that Gos error handling could be much better (s. Rust), people bring up things like JavaScript. As if anyone criticizing Go that JavaScript was the pinnacle of error handling.

Yes, I've also grown tired of this toxic mindset, as well as the whole "idiomatic Go" dogma that is very often an excuse for poor engineering practices.

“Worse is better” has become a justification for pushing out complete garbage instead of the warning against perfectionism it was originally intended as.

Yeah, this was/is a part big part of my frustration with the ecosystem too. It set a LOT of very problematic patterns in place in the beginning, and many of them are still not unwound.

Stuff like "Go doesn't need dependency injection because it's simple". I heard that literally dozens of times. The opposite is true! It's an even bigger pain without DI because the language is so simple! DI everything or make your users suffer!

Or a personal favorite: "Go doesn't need a package manager". We see where that went, and how much insanely better it is now that we have it.

Or errors. Without wrapping. Oh boy we're gonna pay for that for decades.

There's stuff to like in the language, but the zealots (especially early ones) really did not know which parts they were.

Dear christ yes.

“Go is designed as a systems programming language” has been retconned so that “systems” is redefined to mean programs talking over a network?

“Real programs won’t use repetitive error handling, but build on top of it.” I don’t think this one needs further explanation.

“Go doesn’t need xyz.” This is just the slow and painful process of the golang community realizing one at a time why other languages have the features they do.

“Explicit is good.” Explicit and verbose are not the same thing. You can have explicit and terse.

“Golang is simple.” Golang is primitive, not simple. There are tons of footguns and gotchas, not all of which are chronicled in the linked essay, which would have been so easy to avoid. And everyone just collectively internalizes these issues as if “just avoid writing those bugs” is a sane mindset any different than the languages that came before.

“Go doesn’t need a package managers.” Oops, it did. Now we’re like three attempts deep.

Over and over and over I feel like we’ve been gaslit and told we’re crazy, only to later on have the community act like (for example) golang always intended to add generics and of course they’re a good idea.

Bjarne said it best; There are only two kinds of languages: the ones people complain about and the ones nobody uses.

Yet somehow his language meets both criteria somewhat and is only saved by legacy code. Not to say that C++20 or above is not good, I personally think it's one of the best but sadly not every org can afford to upgrade easily.

Every time I work in a different language I'm always wanting to go back to Go even if it's not the perfect language.

I just love the fact that it literally just works. You install Go, you download code, and write code that's it.

No figuring out what version, runtimes, configurations, build tools, package managers to use. Just install and Go.

I think maybe Rust is the only other programming language that provides the same experience.

Maybe these are just lies I'm telling myself, but every time I use Python, Typescript, or Java I dread programming because I just want to write code and I'm often debugging things or figuring out related to configurations, package managers, build tools and versioning.

This is the power of Go, its integrated toolchain. It makes interop more difficult like the article says, but in my personal and limited experience that's not a frequent use case.

This should be the most celebrated aspect of Go, which Rust adopted, along with channels.

Every time I read a critique of Go, I feel the same way: I'm still going to continue to use it anyways. The reason why I am going to continue to use it anyways is because while I understand that it has plenty of easily documented issues in theory (and that people regularly do actually run into in practice), I still find that it is one of the better programming languages in practice anyways. Some of the things that people commonly list as shortcomings I don't agree with (I like the explicit error handling everywhere) and for other things... I often agree, but it doesn't bother me much more than the shortcomings of other programming languages, of which there is certainly no shortage.

I guess I feel bad for people who are particularly sensitive to the areas that Go does not succeed in, because they are probably going to be complaining about it for the rest of their lives. The truth is though, I don't use a bunch of rationale about what is the best programming language to choose which one to work on for a project, I choose a language that I feel works well for me, that I feel I can consistently write good software in, that feels, well, nice to work in. Being a person that values correctness, I do sort of wish that language could be something more like Rust, but for now, it's just not, though it's not like I hate Rust, it's just not what I reach for in a pinch.

Enough has been written about how terrible Go is by this point. At least now I know what it's like to have been a fan of PHP a few years ago! (That's an exaggeration, but it's not that big of one in my opinion.)

> I choose a language that I feel works well for me

Which is the wisest choice for everyone. Golang is only a problem when a manager imposes it on you.

"Manager imposes it on you" just means you work in a team rather than alone. You can pick whatever you like for side projects, of course you're going to use whatever your team uses otherwise.

More generally, when it's not well suited for the problem to be solved. Eager coworkers anticipating Google level traffic may want to write the system in Go and multiple microservices when a simple FastAPI server would do.

Plenty of software engineers don't know any better themselves

Only a problem when a manager imposes it on you for a program where it doesn't fit well.

Exactly. Every programming language is a tool in your toolbox, and you should choose the appropriate one for the job at hand. For me, that's Go around 95% of the time.

I have no need to worry about a 24 byte overhead for a slice allocation, if I did have to worry about that, I'd probably use C or Rust.

And since Go is so readable, theoretically getting the core functionality out and rewriting it in a more specialized language would be fairly straightforward. And while it's an investment in time and effort to rewrite a chunk, at least you know what you're writing already.

But that's a point made in the article, that Go is also good for prototyping. But there's a few languages good for that, e.g. Ruby which powered a lot of the HN startups in their early days until parts needed to be rewritten for performance / scalability.

But writing a new, unproven product in a high performance, high difficulty language from the get-go is cargo cult; you hope you will need the performance and correctness offered by the language. Meanwhile, halfbaked hack jobs like Facebook and Twitter exploded in popularity and revenue, and while their performance issues gave their developers headaches, at least they knew what problems they had and had to solve instead of guessing and hoping.

> And since Go is so readable, theoretically getting the core functionality out and rewriting it in a more specialized language would be fairly straightforward.

You'll have to rewrite your whole program (or at least factor out the "core" part you care about into a separate binary and talk to it via IPC or network) because Golang has terrible FFI as pointed out by OP.

> Go does not succeed in, because they are probably going to be complaining about it for the rest of their lives.

A lot of people really don't like Go because they have experienced other language features. Go has taken an arrogant stance at trying to make the decision about what features you might need and has given you a very small amount of things to work with.

And a lot of people using Go have experienced other language features as well and either decided against them or that the whole tradeoff was not worth it.

I will keep very fast compilation times and decade long backward compatibility over a lot of your features. Because those are features too.

I mean I miss some language features for sure, but the problem with adding language features is that it adds long-term inconsistency. Take a Go codebase from 10 years ago and it should look mostly the same as it would if it was rewritten in modern Go. Do the same with Java and across generations you'd go from straight for loops, to generic iterators and their for-each syntactic sugar, to for-comprehensions, streams and their functional programming style, to whatever Java is up to in 2024, I stopped paying attention years ago.

Counterpoint, other languages - notably Javascript, Scala, PHP, maybe Java - have taken the stance that they adopt other languages' features, not because the language needs it, but because developers were clamoring for it. Which led to added complexity, because they kept adding more and more ways to solve a problem or structure an application, which led to every codebase being so different from the next that the amount of transferable skills and knowledge became less and less. Scala is the worst offender for that IMO.

One frequent praise one hears of Go is that a developer can open any codebase and immediately understand what's happening. Go is readable and stable, which are critical for code for the long term.

The one language feature that I miss in most languages is pattern matching. I wonder if there's any minimalistic language that implements pattern matching well?

haskell?

Minimalistic?

I suppose you mean haskell without extensions.

I like scala and I think it's a great language even with teams.

I think you're overlooking the context of the language. It was a language that introduced a lot of new concepts to most programmers out there on top of that it was flexible to accomidate existing Java developers (for better or for worse). It evolved and grew quite a bit.

The fact that you could develop different styles made the language a lot more useful.

I'm not sure what you mean by transferable skills.. but that sounds like code for overly excessive coding preferences by hiring groups. If you're working with Play you'll be able to learn Finagale, or go to Http4s.

When you attack a codebase, you can't consume it and understand the full context easily, the range of operations that a code base is expected to do is far too vast for that. Even with Go, you can't just pickup and go that quickly. It may (if you're that experienced) become easier to understand. But with a large codebase, it's going to make it much harder as that it encourages large drawn out functions and verbose handling of errors. Also, Go's testing preferences tend to make it harder to write good and effective tests.

I don't even need to open a go codebase to know what's happening. 60% is "if err != nil".

I love the fact that I can pick up a Go project from 5+ years ago and it still compiles with the current toolchain. I might need to do a 'go mod init' first.

It didn't get 67 new features that start shooting deprecation warnings on old functions in the meantime. I don't have to learn new paradigms or switch integral parts of my code to something new.

Generics has been in Go for dunno how long, haven't used it once. Didn't need to.

I used generics once, was kinda useful, but definitely avoidable. The only feature I could see myself using is something Linq-esque for slices and maps. Otherwise I’m content.

I'm not sure that's a good thing. But for your arugement, code from 5+ years ago and keeping the same deps you're mostly able to do that with java*.

(With exception to some java9 and minor deprications).

However, wait long enough that will come to Go, or it'll become less useful.

> I guess I feel bad for people who are particularly sensitive to the areas that Go does not succeed in, because they are probably going to be complaining about it for the rest of their lives.

Well, that’s a stellar endorsement of the article, because that’s literally the point they’re making.

You’ll use go.

…and then regret it.

…but by then it’ll be too late, and you’re stuck with it.

I think the author makes a compelling argument, which is very difficult to counter, that it is a bad choice and you will regret having it in production in many of the explicitly listed cases, and in many professional situations where companies that are not technically competent use unsuitable tech.

Companies should stick to boring tools.

…but, for personal projects? Sure, go for it.

Go is as boring of a tool as it gets. Which is why I will happily use it

There's no tool boring enough to prevent any chance of regret. At the end of the day, it's really, really difficult to anticipate where your pain points will actually wind up in the real world. In practice, I've had lots of really good production success with Go and not too much heartache about the choice. Since adopting it personally (in around 2014) almost every company I've gone to work since has used Go in some capacity and that capacity was usually growing because it was working for them.

Will you regret choosing Go? Maybe. Or, maybe not.

>There's no tool boring enough to prevent any chance of regret.

I'm not so sure. I know C programmers that swear by it, warts and all, with absolutely zero regrets for using it e.g. in the embedded space.

If anyone tells you C is "boring", that's just plain and simple bullshit. C gives you undefined behavior, buggy compilers (yes even for simple C code, MSVC is especially bad at plain C but there are other offenders) and the world's worst standard library for manipulating strings. Using C in embedded development is probably OK, even if you have to suffer with whatever crappy vendor compiler you are stuck with, but that's only considering the pretty severe limitations that very resource-constrained embedded development typically has (e.g. no dynamic allocation.) C is only as boring as you force it to be, and you really have to force it to be.

That said... the thing about the embedded space is that most of it is C, always will be, and may continue to be for the foreseeable future. It's really hard to really know what you have to regret if all you've ever known is C.

>If anyone tells you C is "boring", that's just plain and simple bullshit. C gives you undefined behavior, buggy compilers (yes even for simple C code, MSVC is especially bad at plain C but there are other offenders) and the world's worst standard library for manipulating strings.

Boring doesn't mean "has no issues" or "takes care of everything for you".

It means, stable syntax, lots of mature tooling and compilers, and any problems the language has are well known.

Of the various languages around, C is the least likely one to have "buggy compilers" (issues in MSVC because MS compilers focus on C++ are a different thing, also falling in the "known" category).

> It means, stable syntax, lots of mature tooling and compilers, and any problems the language has are well known.

In case of C what it really means is that the compiler codebases are extremely old and sometimes in rather bad shape.

Speaking of stable, C hasn't just stayed still: The C23 standard adds plenty of fun new things, like an annotation for unreachable code. Modern C has threads and synchronization primitives, #embed, complex numbers, and plenty more.

> Of the various languages around, C is the least likely one to have "buggy compilers"

C compilers are still routinely buggy, because the memory model is surprisingly tricky, there are many edge cases in the specification that are subtle and despite being fairly basic are not stressed often enough to come up, and because optimizing C while adhering to its rules about defined behavior is an unending arms race about what is still "technically" compliant with the standard.

Again, this is especially true if we're considering embedded development where the compiler you have might be a random hacked up old build of some vendor compiler rather than at least being recent GCC or Clang. In that case, even if you only consider a small subset of C99, there's still plenty of room for things to go wrong.

By any standard you can come up with, C is just plain-and-simple not a boring reliable solution. As it turns out, something being old and relatively simple isn't enough to make it boring; you can still have a pathological case that is just an absurd clusterfuck.

I will grant you one thing: it is the most boring embedded development solution. But come on. The best modern competition is Rust. Being more boring than Rust is not exactly an impressive accomplishment, especially when you consider how much practical value Rust has to offer.

>In case of C what it really means is that the compiler codebases are extremely old and sometimes in rather bad shape

Statistically nobody writing C code gets to worry about a compiler error.

>Speaking of stable, C hasn't just stayed still: The C23 standard adds plenty of fun new things, like an annotation for unreachable code. Modern C has threads and synchronization primitives, #embed, complex numbers, and plenty more.

Compared to any other modern language, this is so still that C could make a living as a living statue...

>By any standard you can come up with, C is just plain-and-simple not a boring reliable solution.

C is the de facto language that's considered a boring and reliable solution.

The points made are less substance and more pedantic nit picking ("yeah, it's a language with the most mature and relied upon compilers, but the code is old dawg", "yeah, it's one of the most convervative languages to change, and you can compile decades old code just fine, but they added some stuff in C99, C23, etc").

And Rust is still very niche, single compiler, quickly changing affair. Compared to C and C++ adoption (which is a big yardstick of a tech being "boring") it doesn't even register.

>C is the de facto language that's considered a boring and reliable solution.

Ive spent last year doing C and I disagree, it is not reliable

> Statistically nobody writing C code gets to worry about a compiler error.

lol. (Hello, it is me, I am the statistical nobody.)

> C is the de facto language that's considered a boring and reliable solution.

De facto? Yes. Boring? Citation needed.

> The points made are less substance and more pedantic nit picking ("yeah, it's a language with the most mature and relied upon compilers, but the code is old dawg", "yeah, it's one of the most convervative languages to change, and you can compile decades old code just fine, but they added some stuff in C99, C23, etc").

Incredible. You missed some of the most important points while adding new points that neither of us were talking about. Are you alright?

> yeah, it's a language with the most mature and relied upon compilers, but the code is old dawg

I'm not guessing that the code is bad. You go read GCC code.

> it's one of the most convervative [sic] languages to change, and you can compile decades old code just fine

I don't think you even made this point in the first place, but that just means the changes are all backwards compatible. They have to make almost every change backwards compatible because C doesn't have a versioning mechanism.

However, all of the changes were not backwards compatible. You may find that you actually can't compile code that uses variable length arrays. They were made optional and several compilers never implemented them or no longer support them. Granted, it's good they're dead, but it's something that real codebases really used, including for a long time the Linux kernel, only really removed from it to support compiling with LLVM.

In other cases, compiler updates break things that were never standard but were relied on anyways. For example, if you go and compile old code from the 90s and 2000s, you'll often find they either refuse to compile or crash because they were never actually compliant to standard C in the first place. Because the compilers are buggy. They still are, but they used to be, too.

> And Rust is still very niche, single compiler, quickly changing affair. Compared to C and C++ adoption (which is a big yardstick of a tech being "boring") it doesn't even register.

Something being commonplace doesn't make it boring. You missed the most "exciting" part of C, what I consider the magnum opus of why it is never boring and never will be: Undefined behavior.

(And the standard library. Seriously, fuck C's string manipulation functions.)

Addendum: Until recently, I actually had a C project where I was using the OpenWatcom 2 fork of OpenWatcom as the primary compiler, due to needing a weird target. While it definitely generally worked, it also gave me a brand new perspective on which to appreciate the stability and reliability of even GCC and MSVC, which I will openly admit rarely run into incorrect compilations.

That said, in my many years of doing C and C++ code, I've run into countless situations where the compiler was, in fact, wrong. Either it refused to compile valid code (MSVC usually, but all of them at times), or it ICE'd during compilation (GCC usually), or it did in fact compile, but did the wrong thing (GCC and LLVM, especially with -O3.) Every time I debugged these down enough to actually report them, someone else was already on top of it, and usually it was even already patched, if not released yet.

I understand that big projects like GCC and LLVM are tricky to maintain, especially since they're dealing with standards as broken and stupid as C. But still, it doesn't change the reality that it's not literally that hard to run into a new compiler bug. The truth is that a lot of us just run into a compiler bug, shrug and make a quick workaround, and then go on our merry way.

There are also some bugs that are kind of bugs and kind of not. For example, the old pointer provenance bug: a comparison of two pointers, which are in fact equal at runtime, can constant fold to being false; it's a valid compilation.

There are many write-ups talking about strange C behaviors, some of which may technically be standards-compliant but are anything but obvious (or boring.)

https://www.ralfj.de/blog/2020/12/14/provenance.html

And sure. It is possible that you never run into any of this fun stuff, of course, but it's extremely easy to experience the degrees to which C is not boring. Just accidentally store a pointer beyond its lifetime (especially a pointer to the stack.) Engage in some undefined behavior of some kind. Suddenly, C becomes... very exciting.

There's plenty of software I work on from time to time that's written in Go which I'm happy are written in Go. Some of those were even projects which I started and chose the language for.

Then there's software I've been involved in where I believe Go was the wrong choice. And software not written in Go where I believe Go would have been the wrong choice.

My point is, Go is just another programming language, with a unique set of strengths and weaknesses, and there are plenty of cases which i have experienced myself where there are genuinely no regrets around picking Go. Not because there aren't shortcomings in Go, but because there are plenty of cases where those shortcomings don't matter.

In my professional life, I have seen significantly more regret associated with choosing JavaScript on the back-end than with choosing Go. But then again, there are services in production written in JavaScript where that just never has been an issue and there are no regrets.

Uhh, maybe.

Where is the tradeoff analysis? Yeah, you might regret using Go when some zero value you forget to fill in somewhere pops up later and ruins your pipeline. But are you considering all the issues that didn't pop up because you chose Go?

Java's boilerplate code? Rust and C++'s lifetime analysis and memory management? C's lack of tooling? Python/Typescript's runtime errors? Functional languages' tiny employee pool? How much did trouble did you save by avoiding these?

Thoughts on C#?

When last I tried it, maybe around 2014? I found it a kinder, cleaner Java with better tooling. Visual Studio (not Code) is still the best IDE I've ever used.

Unfortunately it's not popular in the circles I hang around in and the places I've worked. Now that .NET core is the one true runtime I'd welcome an opportunity to try it again; alas, I doubt I'll have such an opportunity (at least not through work).

I remember the upsides but I'm sure there are downsides I'm not aware of. I'd love to read a critique from someone with real world experience.

Not that you asked me but since Go is my goto language, my thought on C# is that it looks pretty cool. C# with hill-climbing thread pool and async seems rather compelling. I really see only two (major, obvious) downsides with C#:

- It has so much. language. design. This is both a strength and a weakness, of course, but C# really does take this to an extreme. I won't bother making a huge list of examples because I think you get the picture.

- Microsoft needs to stop doing things that harm confidence in .NET. Between the silliness of removing hot reloading from the CLI, the situation around debugging outside of Visual Studio products, and drama around the .NET organization... I feel cautious touching the .NET ecosystem. That's saying something considering the company that backs the Go programming language.

(Note: I have not used C# in production, so I can't speak to what it's like in that case. Seems like it's at least fairly "boring" in a good way for a lot of organizations though.)

Is there a specific aspect of language design that you see as problematic? I agree that it can be perceived as "way too many things to keep track of". I think most of it are small, atomic changes designed to reduce boilerplate that can be intuitively understood (like collection literals, additions to pattern matching and null-coalescing operators). You don't have to spend mental effort on these and if there is a scenario where more idiomatic syntax is available - the analyzer has a good chance of catching it and providing an autofix suggestion.

Go is very boilerplate. It requires at least 3 lines of error checking every 1 line of actual code.

Also it doesn't have packed structs so it's completely incapable of doing low level networking or to handle binary files (you can of course do all the bitwise operations yourself… but is it sensible to use a language that is more error prone than C in 2024?).

Also due to its lack of A LOT of system calls, you will need to use modules written by someone on github, which will happily segfault if you look at them funny.

So now you have hidden memory management! Fun!

Of course if all you do with go is a glorified jq script… sure then it's kinda fine.

I worked on a project with gopacket. It was completely fine.

Try defining a new packet format and ping me.

I’m not sure I understand the packed structs complaint. I have used Go to read binary data and it’s quite easy. You just need to ensure that all of your struct fields have fixed sizes e.g. int32 or [4]int64 or whatever. Unless I’ve misunderstood what you mean?

Yes it works, but you can't state the endianness and you have no control to decide if the compiler will decide to insert padding. It's undefined.

You HOPE it works.

I don’t know about the padding (certainly it never inserted any when I’ve used it) but you can definitely state the byte order upon reading or writing. That would definitely be an oversight. Take a look at the encoding/binary package:

https://pkg.go.dev/encoding/binary

I want a struct, not to having to write the code manually to do a struct every single time.

I know you can do that in go but as I already said: "more error prone than C".

Can you please give me an example of what you don’t like? I’m not sure I understand the “write the code manually to do a struct” bit.

You have to define the struct for sure, but beyond that you just pass it to binary.Read and it comes back with the fields populated. I don’t see how you’d avoid defining the struct.

I believe what he wants, is the usual C trick of defining a struct which represents the wire format (with all the usual caveats). Then cast a char pointer to be an instance of a pointer to that struct. Sort of like this:

    https://github.com/danos/vyatta-dataplane/blob/master/src/ecmp.c#L108-L116

It sort of works on x86 chips, but is not so effective on MIPS, PPC, etc where misaligned access are either unavailable, or slow, or even trap and are slower still.

Once one has to handle that sort of situation, and actually copy the data, the lack of language support for such type-punning becomes immaterial.

https://gcc.gnu.org/onlinedocs/gcc/Common-Type-Attributes.ht...

If the misaligned access is slow… it will be slow whether the compiler reads the variable or if you hand-write even slower and error prone code.

That said, please show me a protocol that generates unaligned accesses on any architecture. People writing protocols aren't noobs :)

Oh I see, thanks for the clarification! Personally, I'm fine without that and with something like

  func Foo(r io.Reader) {
      var m MyStruct
      binary.Read(r, binary.LittleEndian, &m)

  }

but we may be operating in different contexts where the underlying copy is or isn't a problem. That said, depending on the implementation of the reader passed in, the bytes might be being streamed from elsewhere, in which case the copying is minimised.

How low level a networking use do you desire?

I certainly managed to use it to implement a protocol over UDP without any issues, that having byte and bit packed values.

Or do you wish to have something similar to C with structs and (endian dependent) bitfields overlaid on packet buffers?

> Or do you wish to have something similar to C with structs and (endian dependent) bitfields overlaid on packet buffers?

endian dependent until you tell gcc which endianness you want :) Which you can't do in go.

In my personal experience your average go developer that likes go, likes it because he doesn't really know anything else.

As a result most things written in go are, in my own experience, of lower quality compared to things written in other languages.

So using things written in go is usually my last resort, because I expect they won't be high quality before even downloading them.

That is an interesting experience, as I find the opposite true in most of my cases.

And I am a go developer that likes go and use it for 95% of my coding, including MMO servers. I love programming languages and have used and dabbled in many. I still go to go.

And many go tools I use tend to be pretty well written. But maybe this is just a sample size of 1 vs a sample size of 1 issue.

From go programs at the very least I can expect they won't respect normal GNU getopt command line options, and work terribly with signals and going background in the terminal and so on.

If it's a server, of course it won't support any way of communicating it's ready. Not the old fork way and certainly not the new sdnotify things.

why would it have to respect GNU getopt? When did that become the golden standard? I never respect getopt because I really don't care about it and it has no bearing on anything I build. As long as they are documented under `--help`. Almost everyone uses Cobra from command-line options. And it is capable of doing getopt if you want, but I don't see why it would be a requirement.

Signals and Backgrounding seem to be just developers that have little experience with that stuff. I can't remember the last time I did any sort of signal handling to do anything specific. And I haven't had issues with backgrounding, but that might be just the tools I use and the infrequency of backgrounding that I do.

Most servers I interact with or work on have some sort of `health` api endpoint to get that status because most servers in go are HTTP. What more are you expecting? I don't even know what you are referring to by the `old fork way` but I will agree most don't use sdnotify, as that makes the assumption you are running Go on something that handles systemd notifications.

I am fairly certain a majority of Go servers run in a container, specifically an Alpine, distroless, or scratch container. So communication would be over stdout or some kind of API endpoint, be it HTTP, GRPC, or any other TCP/UDP endpoint.

> When did that become the golden standard?

30 years ago?

> Signals and Backgrounding seem to be just developers that have little experience with that stuff.

Yes, go developers have little experience. I see we agree.

> What more are you expecting?

Them to tell systemd they started.

> I don't even know what you are referring to by the `old fork way`

As I said, most go developers have little experience. I see you agree once again.

The old fork way is the portable way to do it with whatever init system (systemd included).

> 30 years ago?

I have yet to find a developer, go or otherwise, that cares strongly enough about getopt outside the hardcore FOSS groups that pays attention to this. I know of getopt, and I don't follow it, because no one ever told me to was the golden standard, I was under the impression it was one of many ways, not the only way, especially if you work with systems like Windows at any point, that doesn't follow it either. I also personally don't like the getopt style

> Yes, go developers have little experience. I see we agree.

Or you know, they have no need for it, and don't implement it. You are conflating inexperience, with not caring about the thing you care about.

> Them to tell systemd they started.

Why? Very few services care about your init system these days. Like I said, containers have made it so you don't need to worry about it. I have dealt with many supervisors and init systems, and I have 0 interest in specifically adding code to notify them. At best I will add a network endpoint, and it is up to the unit file to poll for it, if it can.

> As I said, most go developers have little experience. I see you agree once again.

No, I am arguing that in modern day server dev, it is an unnecessary skill to learn, so they don't either learn it, while having plenty of other experience, or they specifically don't care enough about it for it to matter to them.

> The old fork way is the portable way to do it with whatever init system (systemd included).

This assumes you care about an init system at all. Like I mentioned previously, the majority of places Go servers run on are either in Containers, or microvms. The init system does not matter. So few go devs take the time to implement support. It is not lack of experience, it is priorities.

I work in DevOps, ever since we moved primarily to containers, I have stopped caring about adding signal handlers, init system supports, etc. I write far fewer init files, be it unit files for systemd, or old school SysV stuff. It is an unnecessary set of features to add for projects that are internal, small, or run in containers. I am sure if a Go project gets big enough, it gets support for init systems, but those kind of projects are few, and init system support is an after-thought.

If you are so big on FOSS standards, maybe open a PR and add the necessary sd_notify features. Should be only a few lines of code if you import https://github.com/coreos/go-systemd/blob/main/daemon/sdnoti... and call it.

Most Go projects won't add this by default, because the use-case is infrequent for it for most places Go is used. Just like the article said a pro and con of the Go ecosystem is we don't need to worry about the rest of the dev world to gain the benefits of Go. So GNU/linux standards are not a priority. Especially with how easy it is to build cross-platform, using a Linux standard for all platforms is not perfect.

Maybe take the time to be open minded instead of calling the entire Go dev ecosystem "inexperienced", it is simply the case of priorities and needs to the community, and when there is no need for doing the things you mention, they aren't added, even if the devs know about them.

And I disagree that everyone should follow getopt, it is not the only way, and it is not necessarily the best way.

> I have yet to find a developer […]

Users care, because that's how all the other commands work. It's the standard whether you like it or not. Of course you can choose to do the snowflake CLI, but that doesn't really shout "experienced dev" does it?

> windows

oh lol… what % of go programs is used on windows? 1? 2?

> Why?

So systemd can start the other service that depends on that one?

You need to know when the docker daemon is up and running before starting your containers no?

You don't know how to do a thing that anyone writing a server process should know.

Yes you're further and further proving my point that go developers are on average less experienced. And you seem ok with that except when it's pointed out.

> maybe open a PR

I have far more FOSS contributions than you do. Mostly in C/C++ or Python.

> Maybe take the time to be open minded instead of calling the entire Go dev ecosystem "inexperienced"

Maybe take the criticism instead of getting angry.

> Users care, because that's how all the other commands work. It's the standard whether you like it or not. Of course you can choose to do the snowflake CLI, but that doesn't really shout "experienced dev" does it?

I have never had a user complain about this. They just read the readme or the `help` and use it accordingly.

> oh lol… what % of go programs is used on windows? 1? 2?

I am sure it is quite a bit more than that, especially since I tend to see `exe` builds for most go tools I have used.

> You don't know how to do a thing that anyone writing a server process should know.

Oh I know it, I just don't think it is relevant for my day to day, or most of the go-made servers or tools I run.

Another thing that you seem to be assuming is that I am waiting for daemons to start. I used the term Containers instead of Docker specifically because the majority of them are in Kubernetes, or get converted to VMs for services like Fly.io. In those situations the daemon, or equivalent, is already running. I also use Podman, which doesn't have a daemon unless you need one.

And in the few cases where I would have needed to wait, I poll the docker unix socket for the /info path, and get the information that way.

> I have far more FOSS contributions than you do. Mostly in C/C++ or Python.

I never argued that, I have a paltry amount of FOSS contributions compared to most people interested in open source. I just said that if it matters so much to you that you are willing to label an entire dev community as inexperienced, and act all high and mighty about some standards that aren't seen as important these days, then you should teach us how to be better. Open that PR and show us the way. Obviously there is only 1 way to do things, and you are already well versed in it, guide us out of the dark hole we apparently live in.

> Maybe take the criticism instead of getting angry.

I am not in the least bit angry. I am discussing this, because I believe you are stuck in old ways and not adapting to the changes in the industry and development, especially in the areas where Go is used most, which is servers, devops, containers, and the like.

Also, I acknowledge there are industries, environments, etc that can't or won't use containers. But that is a small percentage of use-cases, and the engineers there are probably building things properly.

I will also never say no to implementing such a thing if I get an issue opened requesting it, a PR opened to add it for me, or the like. But I have not seen a need for it in years, both professionally or personally, to be added right away. And I feel like majority of Go devs have nothing against it, and would add it if requested. I know Miniflux for example, after their V2 rewrite into Go, added it 8 months later, due to a request. and have been maintaining/improving it since.

I had no clue how the sd_notify support was in Go in general, until this conversation with you. Not because I wasn't aware of sd_notify, I just never had to go looking for it or needed it. Took me 10 seconds to google it, find a library that saves me time, and I can add it to any server I need it in now. But I will do that if needed, not pre-emptively. Especially since 95% of the servers I make are for work, and dont need systemd. And personal stuff doesn't need it either. But if I opensource anything, and it is requested? I will add it without much fanfare.

> Oh I know it

You just wrote in a previous comment that you did. Make up your mind :)

I'm not wasting even more time with this.

No, because it's an example of someone who chose Go and didn't regret it, and continues to choose Go, because the objections of the article are in practice just not very important or compelling.

I sort of like Go. The explicit error handling is a little obnoxious sometimes, but it's just a way to get things done. Otherwise I see its simplicity as a strength. It is very ergonomic, easy to pick up, and generally performs pretty well. I perhaps wouldn't pick it for every scenario, but there are plenty of scenarios where it would be a good tool.

Then again, I sort of like Java and Python too, two languages I am proficient enough at. All of those are good tools for what they intend to be.

I don't understand why people get so passionate about programming languages. They are tools. You may like soke more than others, but that doesn't invalidate the ones you don't like.

>I don't understand why people get so passionate about programming languages. They are tools.

Because when you're a professional programmer, tools are a huge part of what you do and how you do it, same like a race driver would need to be passionate about cars.

It's just that for an e.g. carpenter, tools are more or less standadized and simple enough to evaluate.

If saws and hammers and routers had as much variety as programming language tooling, and were as multi-faceted to evalute, carpenters would be absolutely obsessed with using the good ones - even more so than they already are.

I think if you tried to tell a professional carpenter that you'd replaced the contents of their toolbox with the equivalent pieces from a discount hardware store you'd be looking for your teeth on the floor.

I certainly wouldn't give up my electronic hardware repair tools without a struggle, it took me years to find ones that I like!

> Because when you're a professional programmer, tools are a huge part of what you do and how you do it, same like a race driver would need to be passionate about cars.

I am a professional programmer. Have been one for more than two decades. And perhaps for professionalism, I think there is no space for passion when it comes to choosing the tools of the trade. Passion would make me pick unsuitable tools because well, I would be passionate. Passionate people don't tend to make rational decisions.

I would expect a professional carpenter to be the same. They may have preferences due to familiarity, positive experiences, etc and so forth. But passion?

Being passionate about using the right tools for the job is still being passionate about tooling.

Passionate with tooling is not synonymous to irrational about tooling. It means invested in the matter of the tools you use.

>I like the explicit error handling everywhere

Then you're doing yourself a disfavor by using Go. In other languages it would be even more explicit, mandatory, and automatically checked whether it's handled!

I like Go, but after writing/reading so much Go code, I get nightmares from `if err != nil` demons punching me in the face. There were so many nice suggestions made to fix this, but there are some extremely conservative and extremely loud-spoken community members who killed all the proposals with vitriolic feedback. Now, the Go team has given up the battle for improving error handling since they are psychologically afraid of these folks.

Every Go developer survey has results where the community overwhelmingly votes to improve error handling, but a few extremists have derailed all progress.

(2022) Discussions at the time:

(130 points, 148 comments) https://news.ycombinator.com/item?id=34188528

(748 points, 544 comments) https://news.ycombinator.com/item?id=31205072

> I've started caring about semantics a lot more than syntax, which is why I also haven't looked at Zig, Nim, Odin, etc: I am no longer interested in "a better C".

Well the post rambles a fair bit, IMHO. The whole bit about Go being “accidental” is BS given that Rust is just as much “accidental” in its origin and design.

One thing stuck out to me is that Nim certainly isn’t a “better C”. It has a GC or you can use reference counting. You can use it as a better C if you really want.

Nim’s type system avoids many of the problems that Go has, though it’s not nearly as pedantic as Rust.

At the end of the day lots of software has been written and shipped in Go which runs fast, has minimal downtime, generally seems effective, and has minimal security issues. I’d argue (much) fewer software projects have been shipped in Rust. Firefox is still 95%+ C++.

> I’d argue (much) fewer software projects have been shipped in Rust. Firefox is still 95%+ C++.

It's funny but this comment reminded me of this tweet[0] from 2022 (!). I don't have a horse in this race as I am happily using Python and C++ at $DAYJOB. I'd argue that even if much less software has been written in Rust (source?), it still qualifies as "lots of software has been written and shipped" with it. Not to mention all the investments by $BIGCORPS in the language.

0: https://x.com/m_ou_se/status/1599173341117435905

Not sure why Go is compared to Rust all the time, whilst most appropriate comparison is Java.

I think this is exactly the right way to understand Go - it's targetted at building servers in environments where having strong consistency of code and a short ramp up time for junior engineers is valuable - i.e. it's perfect for all the big corp scenarios that Java was used for.

I think maybe the more common, but less helpful comparison of go vs rust comes from the fact that they are both part of a new wave of languages and that they both default to producing staticly linked binaries.

> consistency of code

There are many stylecheck tools that should be apart of a good stack. Accepting the creator's style is putting a lot of weight on their opinion. Most organizations have their own preferences for good reason.

> short ramp up for junior engineers

Junior engineers aren't a place you're concerned on being productive. Most of the time at that stage in someone's career they should be learning more, getting up to speed with professional practices, tools, and trying to learn code bases+patterns. Ramp up time for a language is a very minor consideration.

Both of those things have very little to do with server environments.

Bigger corporations struggle with Go's module system and forced reliance on seperate repos for separate modules. (Can you bundle multiple modules in the same repo.. yes but you're going to have a bad time)

> Both of those things have very little to do with server environments

My experience of bigcorp is that they need lots of servers (http is the modern bailer twine) and want developers to act as far as possible as indistinguishable resource. They will have rotating, high churn, globally distributed teams of vendors, contractors, consultants, internal staff and the teams will encompass a vast range of skill levels and abilities.

Some languages amplify skill level disparity, some attenuate it.

This is where I would argue that it's a terrible language choice for the environment. (The need for rest frameworks) There aren't a lot of established or mature options for this, and a lot of them don't give you very much to work off of. It's going to be a bumpy road for exploits in this area.

> act as a indistinguishable resource rotating high churn, globally distributed teams of vendors, contactors, consultants, internal staff...

So this is a huge organizational problem that isn't helped by Go. This is more of a higher level problem and the lack of technical leadership in the company. Any language you go with .. you're not going to do well by trying to utilize everyone there. Picking a language that gives you the impression that you can blend the skill levels will end up with a disastrous mistake with a lot of slippage on the development side. Will you generate a ton of code? Yes, the language encourages it.

What I do agree with you: It does encourage the mass hiring of developers with low to no expectations of performance. (Seems like that's google's focus these days) But that's a bad thing for the developer who wants to achieve a lot of results (not just write lines), and to develop our teams.

Because someone decides what language to write a new thing in is very likely to consider Go and Rust. They are very unlikely to consider Java.

Are Rust and Go sufficiently different that they should each be chosen in different cases? Sure! But that’s literally why someone would consider both and pick one.

They should consider Java though.

People have an irrational hate for it based on the enterprise cruft and horrible third party frameworks you can just completely ignore if you build a new thing

It's not good for commandline stuff but for a long running small service it is pretty great

^ this! Java's far more capable than golang and a better choice for many projects.. just don't use the decades old "Enterprise" stuff.

..and I'm generally a Java-hater as the language itself is objectively inferior to my beloved C#.. but even then there are situations where Java is a better choice.

That's a big if. In practice you're not going to be able to escape needing to interface with Java code others have written. What then? You either waste time putting up shims in front of every single API or, more likely, give up and just go with it, at which point you've lost.

It is much more practical to choose a language that does not have a terrible history of poor API design to begin with.

There are plenty of good libraries that don't suck you into a hellscape of frameworks or enterprise java

What language that's been around for long enough to have a large ecosystem doesn't have a terrible history of poor api design?

But sure, if you pick a really new language, you do get the chance to be that history for future generations

> They are very unlikely to consider Java.

Except they are not unlikely to consider Java.

It's a language with a very robust ecosystem, it is relatively easy to hire Java developers, and a decent job for large, complex projects.

Not considering Java is just as dumb as not considering Go or Rust.

Go is an iteration of C, not of Java.

It's a really bad choice for situations where Java is a good choice as not only is the language limited, the ecosystem around it is also very limited when compared to say Java.

I'm maintaining Go, C# and TypeScript as my main languages as that gives me excellent coverage. I'll add Rust to the mix when I have 6 months where I can accept the productivity drop or have a project where the quality requirements are lower (it only takes a week or two to pick up a language, it's learning how to engineer within the ecosystem which takes the time).

How is Go an iteration of C? You can't use Go to write a kernel, or program a microcontroller, or for high-frequency trading or a web browser or a tensor library or a language run-time. It's either a bad idea or simply impossible, depending.

Someone please explain to me what's C-like about Go other than vaguely the syntax and that it compiles to machine code.

To be pedantic for a moment...

> You can't use Go to write a kernel ...

Not a production kernel, but MIT did use Go to "study the performance trade-offs of using a high-level language with garbage collection to implement a kernel" [1]

There is also gVisor [2] which implements, as best as I can describe, a kernel in user space. It's intent is to intercept syscalls made in containers and to redirect its execution in a sandbox.

> ... program a microcontroller ...

I'm not sure if one would classify this as a microcontroller, but USB Armory did write a, iirc, Go compliant runtime for bare metal ARM and RISC-V [3].

There is also TinyGo [4] with the following listed microcontroller support [5]

[1] https://github.com/mit-pdos/biscuit

[2] https://gvisor.dev/

[3] https://github.com/usbarmory/tamago

[4] https://tinygo.org/

[5] https://tinygo.org/docs/reference/microcontrollers/

The refusal to use anything language designers have learnt since 1970 is what makes it C. No sum types, poor error handling, no map/filter funcs, all the other stuff written in the post. Java has all of that, any language after the 90s has most of that, Go doesn't. Actually even C has sum types lol.

The fact that it can't fill many of the usecases that C can and yet has ergonomics as bad as C makes it even worse.

C -> Alef -> Limbo -> Go ?

With influences from other languages along the way.

The Go language was actually created while waiting for C++ to compile. Their goal was to create something that was better than C++ for network services and they succeeded- but that's a pretty low bar! Most companies that don't have major performance concerns don't use C++ for networked services. If Rust was already mature it would have been a good option- except for the compile times which was one of their main original issues could well be worse in a Rust code base! I have seen this myself- Go being used instead of Rust with the biggest reason being it compiles so much more quickly.

It's weird. At a time I was looking for a "better" python. Something simpler and safer than C/C++ but faster than python and more importantly that can produce a single binary. So I looked at everything from Rust to obscure language like Hare.

Go should have been the obvious choice but for a reason I don't understand I dislike its syntax. For Rust I understand: it uses a lot of special characters that aren't easy to remember and to type with a non qwerty keyboards (plus other unrelated pain points). For the different lisp, it was the parenthesis and the reverse polish notation. But for Go, I'm unable to rationalize why I don't like it.

For the anecdote, I settled on compiling my python code with nuitka. No speed gain that I'm aware of but I can now provides a binary. I'm also looking more and more at C# because of its progressing AOT compilation and although I dislike the verbosity of its base mode and the fact it's so tied to windows.

I liked a lot nim and crystal but the small community was a barrier, although I'm really impressed by what nim is managing to do with such a small community and it may me think it's an excellent language

(I will try to motivate myself to pick up one of the language I mentionned above also)

>and the fact it's so tied to windows

Dotnet Core is not tied to windows except for certain frameworks like wpf (and there are alternatives for it that work everywhere), credential store.

And it's actually really good to use these days.

Yup, common misconception for folks that haven't used it in a decade.

Our team uses C#. We dev on Apple silicon Macs. Some use Rider, others just use VS Code. We build on Linux via GitHub Actions. Ship to prod running AWS t4g Arm64 instances.

C# to me is like TypeScript++. The language, syntax, and core constructs are close enough that anyone with a good handle on JS and TS can pick it up easily and be productive.

Sorry for the segue, but how your team's experience with C# on VSCode? Any recommendations for plugins? I've heard of a lot of people recommend Rider but not much, aside from neonsunset, talk about VSCode.

Great.

C# DevKit is really all you need (and the same plugins you would normally have like GitLens, etc.).

Refactoring experience isn't as good as Rider (JetBrains are kings of refactoring tooling). But for all other cases VS Code is fast and ergonomic.

Rider does have some nice things for supporting working with SQL databases that I do envy once in a while.

Just in case: DevKit is optional and requires an account, you can just use the base C# extension which is what provides the language server and the debugger, if you prefer VSCodium there's a fork of it which packs Samsung-authored netcoredbg instead of vsdbg so that is covered too.

For F# - Ionide works great, I like it a lot, integrates seamlessly with existing C# projects.

I'm 100% sure your code integrates with the linux ecosystem as an american tourist in europe :D

Nuitka is compatible, but comes with the shortcomings of python C-API

There are other approaches which give up on C-API and build a bridge to languages such as Go and Rust.

I've spent a few years working on py2many. Would appreciate feedback on the approach.

One of my biggest beefs with Go and dozens of other languages is: static typing without a sufficiently developed type system is masochism.

If I can't get at least ADTs and parametric typing, give me gradual like CL.

I wish I could take the Rust pill like everyone else and be happy.

I am not a stranger to the programming world. I am fluent in Python, Java, Scala, Erlang. I can write acceptable Haskell and C++ and Common Lisp. I am well versed in functional programming (though I don’t hate OOP — it is still a good thing when done correctly, and I am a big fan of Smalltalk).

However, for some reason, I am completely unproductive in Rust. It always fights me. There’s always something I want to do (graph structures? spliced lists?), and it’s always “you don’t need that” and “it’s not the Rust way”. There’s always explicit lifetimes and you miss something and you need to rewrite huge swaths of your code. And nothing is refactorable. It’s not even a good functional language, there are no persistent data structures, and it’s a pain to make them, and no ways to express yourself functionally like I used to do with Scala.

But apparently, everyone and their dog like Rust and are insanely productive in it. What am I missing?

I do miss several of the things you list, too, especially (safe) self-referential data structures. But Rust is a fantastic compromise solution if your number one priority is to match the performance (or rather, performance possibilities) of C and C++, and have memory safety while doing it. If your number one goal is to be able to write memory safe ideomatic code that is entirely comparable to those languages, and then add some functional programming and lots of quality of life stuff to that, then it's pretty sweet.

In other words: Don't think of it as rivalling the entry cost of Python or Java, or the expressiveness of Haskell or Scala. Try to think of it as a replacement for C++, with memory safety and piles of quality of life improvements. And then with a sprinkle of the other things on top.

I am fine with C++ memory safety if I am mostly working with RAII and explicit move semantics when needed (it is not perfect, but it will get us through the night). However, fighting the borrow checker drastically _decreases_ my quality of life, and I couldn’t find the way to get myself to free flowing code as I had in other languages.

Interestingly, I have the same feeling when trying to write something with Vulkan — the amount of ceremony and boilerplate before you even get to the first triangle is a huge barrier for my productivity. I want something more immediate.

I love Go, so I am biased. However, the beautiful thing about Go is that it doesn't even attempt to prevent classes of bugs by making them impossible. It's a car with ABS but no lane assist, with power steering but no collision detection.

Out of all the bugs which Go permits I have yet to see one which could survive in production for some time without being discovered. Almost all of them would cause a failure the first time a segment of code is run.

Instead it is a plethora of footguns which are attempted to be managed through convention.

This post is always a scary read:

https://www.uber.com/blog/data-race-patterns-in-go/

> Evidently, the Go team didn't want to design a language.

This excerpt should be enough for you to know if you will find the article insightful or useless.

And this is why I normally start by going to the HN comments before reading an article.

While the article raises some valid critiques, it often overlooks the fundamental tradeoffs that make Go successful in its niche. Go’s simplicity is not a flaw but a deliberate choice, enabling rapid onboarding and maintainable code for large teams. Its explicit error handling may seem verbose but ensures clarity and avoids hidden surprises common in exception-heavy languages. The complaints about ecosystem isolation and tooling ignore the fact that Go provides powerful built-in tools like pprof and dlv, which are better suited for Go’s runtime than general-purpose alternatives. Go isn’t trying to be Rust or Python—it’s a pragmatic tool for scalable, performant systems.

This. And there is one technique that would help in this.

Too often I see code like "xyz := pkg1.SomeFunc(a, b, c)" that makes xyz's type non-evident, especially when interfaces are involved.

Please write Go code like

  var xyz pkg2.SomeType

  xyz = pkg1.SomeFunc(a, b, c)

My 0.02€, YMMV.

I don't get it. Why not use a code editor that shows the inferred type on hover?

That should not be an issue, unless the code is written in notepad. As for interfaces, a better approach is to return struct and accept interface in your functions wherever it is possible.

Sometimes it's better to to return interfaces, like contract style programming.

I’m successfully using Go and agree with the author at the same time. For me it’s not a choice between Go and Rust- Rust is not as productive for high level code because every API maximizes performance.

Since this article was written Go introduced generics which does solve some of the complaints. The rest I mostly solve with linters, libraries, and conventions.

Go has a syntax that is well designed (arguably one of the best) if you’re aiming for familiarity with C style syntax in an imperative language. Other than that it’s a poorly designed language with by far the best engineering work having gone into things that aren’t actually the language per say (cross compilation, asynchronous runtime, GC, module system, super fast builds, etc). They wrote an ACM Queue article essentially stating that: https://cacm.acm.org/research/the-go-programming-language-an...

Go is just an evolution of C, it works at a higher level but shares many ideas. If you can’t appreciate the simplicity of C, then you probably won’t appreciate the works of Bell Labs, and I have lost all interest in the debate.

CTRL+F "rust", 24 matches, I had a feeling that would be the case. What does Golang, for the most part, have to do with Rust? I also find the following bit somewhat funny:

>The success of Rust is due in large part to it being easy to adopt piecemeal and playing nice with others.

And as if Rust itself didn't suffer from the same kind of imperfections one can find in Golang. So much for that "nothing new under the sun" back in 2012! But then it starts talking about the "rust shills" boogeyman, and one has to wonder if it's not trying to justify one's choices. (which is fine, anyway) And I agree wholeheartedly to each and every single one of the "lies" listed in the article, that you could very easily rewrite half of which to fit Rust, and the other half requires no changes to apply.

> And as if Rust itself didn't suffer from the same kind of imperfections one can find in Golang.

Which imperfections are these?

Even async in Go isn't that good, ultimately. You can't monitor channels, you can't properly react to errors, you can't shutdown and restart them. A panic in a channel will kill your program. Etc.

It's "we learned about green threads and didn't bother to learn anything else" approach (also prevalent in many other languages)

It's more that it's CSP from early to mid 1990s, over which Go just added syntax sugar

I remember it took a me a bit of investigation to make Ruby call Go functions, but it wasn't really hard: https://dario.cat/en/posts/portal-between-ruby-and-go/

The argument about FFI and CGO was the most unappealing to me. If you really need to do that, it feels that at some point some decisions were made that weren't the right fit for the project.

Also, I feel that many arguments in the article are potential pitfalls for unintended side effects I've experimented in other languages, like Ruby itself. Keyword arguments are nice, but I know that they've bitten me several times.

Go forces the developer to be explicit about their intentions, that's why I also don't find something bad that your struct functions need to have the pointer receiver. It can be annoying when you forget, but that's what linters are for.

> I've started caring about semantics a lot more than syntax, which is why I also haven't looked at Zig, Nim, Odin, etc: I am no longer interested in "a better C".

A strange take. Zig, Nim and Odin are about fixing semantics not to bring syntax sugar.

The title lacks a (2022) Otherwise great article :)

His criticisms include:

- No sum types

- Default zero values

- "Go is an island": Fully integrated toolstack makes it easy to use the same tool top-to-bottom, but makes it difficult to use with any other project

None of those are lies; apart from the first, they all have advantages and disadvantages. Sure, there's an advantage to forcing you to specify every value, but there are disadvantages too. And sure, there are disadvantages from Go being an "island", but he lists some of the advantages right there in his post.

If those are deal-breakers for him, then sure, go use something else. But lots of us like the advantages more than we dislike the disadvantages.

I liked the idea of a language with minimal syntax that is easy to learn and easy to understand because the code is forced to be straightforward. It didn't work out that way to me in practice.

It's a bit too low level for many use cases in my opinion, and that does get in the way somewhat. It also works against the "easy to learn" part unless you start with developers familiar with low level programming.

I also found some types of library code surprisingly difficult to read, especially when empty interfaces where used.

The standard library was great though, it covered stuff that others don't and it just worked out of the box.

The article is great. The bigger picture, of course, is that it’s always “pick your poison”.

On one hand, say, I love operator overloading, I love how Python does it (once you satisfy an interface, its operators Just Work).

On the other hand, I can appreciate the choice not to do it at all because half of the ecosystem will do it, and another half won’t. Also, it would require implementing function overloading, and it is a can of worms.

Or generics and rich type systems, which all come with their own tradeoffs. I hear that Rust cajoles you into tinkering with the type system, and wach tweak requires refactoring of more codebase than anyone would like (don’t take my word for it, it’s just what I heard from a few different sources). I know that Nim is so expressive that it can be annoyingly trivial to be too clever and run into a valid edge case that will make the compiler barf and die. Go sidesteps the issue by not wading into that territory, and that may be perfectly okay, albeit verbose.

It’s always picking your poison, so I guess check your tolerances and allergies so it doesn’t kill you before you get the job done…

> I love operator overloading, I love how Python does it (once you satisfy an interface, its operators Just Work).

Is this different from other styles of operator overloading? Why does it matter whether, when I want to overload the + sign, I need to define a function called `__add__` or `operator+`?

For me it's easier to introspect, fewer unexpected corners. Contrast this with operator overloads which you can put literally anywhere.

I want to build things and could care less about intellectual masturbation. It was built for proletarians like myself. I like Go because it's neither Rust nor Python. But I do understand that there are times when having some nice abstractions over common patterns doesn't hurt. At the same time, Go has come a long way over the years.

Should be called "Lies we tell ourselves to keep Go-ing"

> Because they're Go experts, they know the cost of using Go upfront, and they're equipped to make the decision whether or not it's worth it

Well maybe not.

If I’m an expert in only Java 6, I might not be aware of all sorts of useful features that other languages have, such as sum types, traits, type inference… I only know one side of the trade off.

I might be vaguely aware of those ideas and dismiss them because I can’t imagine how they would fit in my Java 6 workflow.

And yet when some of them arrive in later Javas, I begrudgingly use them. Years later I can’t imagine how I ever lived without them!

We’ve seen this play out so many times.

Remember when lambdas were just for weird functional languages?

The article is arguing that Rust is a better choice which is laughable. Go is a small GC-ed language. Rust is an obnoxiously complex language with strong emphasis on memory management. I don’t see any projects where you could legitimately hesitate between the two.

Truth be told I quite like Go. It’s small and simple. It does the job. You can onboard people on it quickly and in practice with the toolchain you don’t run into the issue mentioned.

Would I rather use Ocaml most of time? Of course but it doesn’t have the same ecosystem. Would I rather use Rust? Heck no, I’m not going to deal with the borrow checker and life cycles if a GC is fine.

I have recently started to port a large codebase from Go to Rust. I thought it was going to take more code in Rust, but I have been surprised at how much less code it takes. As of right now, the Rust version is 60% smaller in terms of LOC.

I think a large part of this is due to error handling in Rust and the sugary `?` syntax which is so incredibly useful.. I can't count how many times I have done the following in go

x, err := Foo()

if err != nil { return nil, err } // or something like this

I'm pretty sure a large chunk of the LOC reduction was in that alone. Though, a large chunk is also some crates that make a lot of boilerplate go away (serde, serde_json, etc).

Seems like Zig is actually meeting much of the writer's expectations for good language design. Although I'm still not wholly convinced by it's move from LLVM and hell-bent desire for incremental compilation.

>After spending years doing those FFI dances in both directions, I've reached the conclusion that the only good boundary with Go is a network boundary.

It works perfectly well with stdin/stdout as well, as seen in many LSPs

Currently onboarding to go. For me personally, it's too opinionated e.g. for: error handling, mutability, unit testing, naming conventions, lack of features like sum types.

As a new joiner, some things have a "religious" feeling, like: the go gods have given us THE language and we must use it their way, without questioning too much.

I have the feeling other languages are less opinionated and this allows for more ergonomics and comfort when developing.

I suspect it's designed that way intentionally.

It's a language built to make it hard for new joiners to mindlessly add complexity, as the language itself will fight you. It makes it hard to add dependencies that aren't very visible and vendored in, it makes it hard to change how the language behaves with overloading, preprocessor magic or macros, and so on.

It's built to fit Google specifically, with their army of new grads, and deep investment into reinventing all the dependencies themselves, and having a huge monorepo.

If you're not google, you're gonna have a bad time, which is also good for google.

Then why am I still messing around with Axum and friends and people using Go are shipping microservices like there's no tomorrow?

Go is great for a variety of projects. Smaller codebases, little cli tools, simple servers, etc. It has its quirks, as any language does, but it is foolish to write off an entire language when some of the very properties you dislike may actually be a good fit for the problem at hand.

When engineers focus on language features they focus on the wrong things. Yes, languages like Java and Rust are fantastic for large codebases because generics and strong encapsulation become especially important as the number of engineers working on a project scales. Contrarily, Go would probably be a million times better for a small project of one to three staff that primarily revolves around a few well-designed data structures and basic array manipulation. In this context, the more sophisticated features of a Rust or Java are actually hindrances and end up making things more elaborate than needed.

Languages are a resource and a tool, not an identity marker. Use the language that makes sense for your constraints. Programming is all about modeling problems. No language will do this for you and each one has features that help or hinder and some are better or worse form problems of different shapes. Learn about how computers and compilers actually work so that you are empowered to use a wealth of languages and can choose the best one for your project on rational grounds.

Or, you can embrace Haskell and use the only perfect programming language to have been thus far invented :)

I find it fascinating the extent to which language choice makes a programmer emotional looking around at these comments.

Well, the post title implies people using Go are delusional, so that's expected. Starting a conversation by calling someone stupid isn't likely to end up productive.

The argument that a language is inconsistent if it allows all structure members to be defaulted to zero, but does not default function arguments to be omitted such that they default zero, seems to me like someone climbing out on a limb, out of the window of lunatic asylum.

> and most importantly, you adopted a language that happened by accident.

Is the author implying that Go was created by accident?

"Accident" is hyperbole, but IIRC Go evolved out of Google's internal high-performance HTTP server tooling, so they didn't initially set out to create a general-purpose programming language.

The author's last post they referenced is a bit bizarre. I don't think that some overly simplified and error prone std library APIs is a particularly compelling reason to dislike a language. I didn't read the entire thing though because it was extremely long

There _are_ two problems with Golang that I _would_ like to wave a magic wand and fix if that was a power I had.

1) Sum types (E.G. C/C++ union types) - Yeah, something similar to that should exist... it's syntax sugar over #2

2) 'casting' / reshaping perspective on data ; as long as something has the same struct size a programmer should be able to tell the compiler 'but this is actually that'. Which also implies a way of fixing the layout of a given data structure. I figure that's why Golang doesn't allow this already.

Yeah, 24 bytes (len, cap, pointer) per slice (dynamic array) has a cost, but if that really gets your goat use a fixed size array, or pointer/reference to such [n]thing.

3) Seriously, for a slice, make it like a map, a pointer/reference to a len,cap,blob - so that when I pass a slice by value I pass the WHOLE slice, not a COPY of len and cap and a reference to the slab. Current golang has the worst of all worlds with passing slices around, in that changes mutate, UNTIL it's resized or manually copied. The current design has the behavior it does to support slices of slices, which requires the pointer to the blob. A more complex scheme of a container behind the slice, and references to that could also work, but would be an entirely different datatype.

I started learning Go and when I got to the chapter on slices is when I dropped it.

I can't put my finger on why it was so off-putting, but it just left a bad taste.

It's exactly what I described in point 3 as 'the problem'.

Maps 'work' because as a programmer you're aware of how expensive it is to copy a whole hash table, and it makes sense that a pointer to the thing is the whole thing. So it's an exception, but one _clean_ exception that just behaves consistently.

Slices are horrid in that they SOMETIMES work the way someone who passes by value expects, and that's the same for both NOT updating AND Updating. This is because the management stage is passed by copy, not reference, but the backing data IS a reference, which initially is copied. That leads to in-place overwrites when no change in size has happened. It leads to appends that do get written but not reflected in other copies of the slice, and then if there's a need to re-allocate it leads to even the updates not updating the other copies of the slice.

Slices are the most beginner un-friendly part of Golang.

The superpower of go is goroutines and channels. The kryptonite of go is its limited libraries. Go is a great choice for many concurrent applications. I couldn’t finish reading the article because it lacked focus.

I just want to say that I started working with Go a few months ago at my new work, and I've been disappointed. The type system doesn't even have union types, and the implicit implementation of interfaces has caused a couple headaches that I would've never had with explicit languages such as Typescript (strict).

At this point, I prefer strict Typescript with very restrictive ESLint rules than Go. I don't understand how this language was pushed for web servers, it lacks common constructions to represent very frequent data types in API design.

Also, the whole go func(){} and channels is nice if you are coming from a language with worse ergonomics (C?), but when doing multiple web requests to other places and wanting them to happen in parallel, it gets messy fast. It lacks a pretty abstraction -- I very much miss an await-like construction which uses wait groups under the hood.

So, I very much agree with the post. I like errors as values, though.

It really does not matter. There are so many critical services running off of Go. You don't wanna use it? Go to hell.

Okay so I’m wondering, if you’re not in the Golang universe yet, what language is better to start with learning?

I'm curious as well. Are there alternatives to Go in some fields completely dominated by it? (like Kubernetes controllers/operators)

Depends on what you want to do.

Go's channels are probably awesome, if you need parallel computing.

If you want to build in k8s land, you can't avoid Go.

But besides that?

Maybe, use Gleam or Elixir.

I see this false promise all the time: Simple tools are simple to use.

NO. Not necessarily. Use a handsaw to fell a forest if you like, or a hoe to till 1,000 acres. But I will not. Tool simplicity will come back to bite you if the task you're trying to accomplish is complex or high-scale. And simplicity of design in no way correlates to simplicity of outcome, or of your workflow using the tool.

Golang has, to be fair, more going for it than just its relative simplicity. But in many ways the OG criticism rings solidly true. The "we'll do just enough language features and tooling" becomes an abdication and an excuse in a complex world.

So what’s the alternative then to go?

When the article started the part "but Tailscale still uses it" I almost felt watched

One thing what i observed is any language you right ask AI to optimize in Rust..why its so much of a debate when scale comes. In 6months AI will make any rewrite possible with full context aware. So don't spend your energy on things that can be done by machines and focus on things which u can get deep on for distributing the software.

There's a lot of valid critique of Go, but I've never found anything like it that lets me build lasting, high quality, bug free software.

Explicit error handling means that I actually think about and handle errors. No surprises in production. No random exceptions throwing deep inside some dependency. I've been running some Go services for years with no crashes.

The brain-dead simplicity means I am not tempted to waste time being clever.

The tooling means my code is trivial to build, even years later on a new machine.

Well, i feel like this is the 100th article i read about why golang is bad "mkay".

For my personal background, i started with golang about 6 years ago and im using it mainly for private and open source projects.

Yes golang for sure isn't perfect, but what language is tho? I think the major point is - the language you use should match your use case. If it doesn't it will always feel "bad" or you will more likely tend to find points why the language isn't perfect (for your needs).

Sure you could write a website builder in ASM or you can write an Operation System in Javascript - but why should you?

Just look at your use case - check your needs and decide what language fits the best.

If its golang? Fine use it. If its not golang, than don't. But don't take a language and try to match it on "everything" and than complain that it doesn't do the job - because no language will.

Thats my 5 cent's on this topic....

So I’m completely new to the Go but I have very specific need and I couldn’t fit it anywhere else:

I want to orchestrate asynchronous, highly fallible tasks. I could use Rust but I feel I would drown in correctness of handling various edge cases over which I’d rather panic and wind down everything I can.

Elixir/Erlang would also be an option, but getting it installed in specific version isn’t simple (and ASDF was guilty for long time of installing Elixir built on different OTP than defined one) and also problem lies in interacting with IOs, which also bring consequences.

And so, yeah, Go is simple enough so that I feel confident that if aj get hit by a bus there won’t be enough magic so that it won’t be picked up by someone and it… works.

After being in pull rope competitions with bash, sh, Perl and Python I have fun working with it.

My biggest problem with Go is readability of some projects. In many cases, the code is split across files in an unintuitive way, and you can only navigate with the help of an IDE.

> I've mentioned "leaving struct fields uninitialized". This happens easily when you make a code change from something like this:

Really? Are you not using gopls? It absolutely will warn you about this. And the mutex case. And a lot of the other similar criticisms.

> Go not letting you do operator overloading, harkening back to the Java days where a == b isn't the same as a.equals(b)

In languages that have it I've never used operator overloading to produce anything good or that was obviously better than just using methods.

> The Go toolchain does not use the assembly language everyone else knows about.

Honestly I've never noticed or had to care. And you can build plenty of things that don't require CGO.

The whole "gnostic" tone in language articles on HN is becoming overbearing. There are parts of the language to criticize but couching it in this "lies _we_ tell ourselves" and somewhat kitchy writing that presents minor matters of taste as fully existential issues makes this an unsatisfying read.

> In languages that have it I've never used operator overloading to produce anything good or that was obviously better than just using methods.

I used a Scala library for S3 once that overloaded + to mean upload.

    var bucket; var file;
    bucket + file;

Which is obviously bad and unnecessary.

It's really a feature that should be used very rarely by those who make basic libraries, but there it can make a lot of sense - data structures, numeric types, etc.

That says more about the library design more than it does about the library.

There are much better examples of the operator overloading for example cons:

item :: item2

Later you can break that down in pattern matching.

The worst thing - default values as a solution for absent values. What can go wrong with implicit value assignments?!

Would you rather have undefined initial values?

I would. Not because it's something pleasant.

There is a huge difference between no value and default value. I had plenty cases where I needed to distinguish them because absent value is a violation of a contract. Defaults are not a solution.

And more. Defaults bring their own problem. Chosen values are neutral only for some operations. The most simplest case - int default value is 0, but it kinda "works" only for sums. With nulls you'll get an error in production. With defaults discovery of the error is postponed even more. It's literally the same as making up values ignoring clients will (and this decision even not ours but designers of the lang).

Defaults are implicit. It's simplifies mistakes due to human factor.

Oh well, here we go again... (2022)

This is a silly article, though I can only speak for myself. First of all, language design is pretty much the last reason anyone should consider when choosing a language. For me as a solo developer who needs to get results in time, the main criteria for choosing Go were compilation speed, development speed, general tooling, and third-party support. Especially the latter is extremely important; I cannot develop all kinds of libraries like e.g. an Excel or Word document parser in-house, I have to use whatever MIT-licensed libraries that are there.

I've developed in all kinds of languages in the past: Pearl, Java, C++, REALBasic, PowerMOPS (a Forth derivate in the 90s), Ada, Rust, Python, Java, Racket and various scheme dialects, CommonLisp, ObjectPascal/Lazarus, just to name a few. Out of these, REALBasic was by far the one in which I was most productive. Alas, it became unaffordable and the vendor lock-in really burned me. No more commercial languages or IDEs it is for me.

If Ada had a garbage collector, I would probably use it even though the development time is a bit longer. Likewise, I'd love to use CommonLisp and think it's the best dynamic language. But it simply doesn't have enough 3rd-party support, and I'd also be wary about how it runs on mobile platforms.

I've got to say I'm pretty happy with Go so far. Is it the ideal language for me? No, because full OOP with multiple inheritance would make my life much easier. But it works, is easy and fast to develop in, and the results speak for themselves. I have no problems with bugs. The explicit error handling may be a bit cumbersome but it forces you to think about and deal with every error, unlike people in languages with exception system who often ignore errors until so far up the call chain that they don't even know what actually happened any longer. I don't see the point of that. If you have an illegal state of an object it doesn't matter if you call the object nil or DefunctInstanceOfBla, you're going to have to deal explicitly with the illegal state.

Notably, C# was also in my final selection of languages. For my principal use case - distributed client/server applications with GUI - Go's GUI options were not so stellar and I was thinking about using C# instead. AFAIK, C# is very suitable and a great language, too. I decided against it because of the C#'s bizarrely complex runtime library situation and naming schemes (WTF?) and simply because I would have had to learn it first and already knew Go fairly well.

Beware the language aficionados and purists. I used to be one of them, too, advocating Scheme & CL. However, in the end purely practical considerations are always more important. Programming languages are tools for getting things done.

“An idiot admires complexity, a genius admires simplicity, a physicist tries to make it simple, for an idiot anything the more complicated it is the more he will admire it, if you make something so clusterfucked he can't understand it he's gonna think you're a god cause you made it so complicated nobody can understand it. That's how they write journals in Academics, they try to make it so complicated people think you're a genius” ― Terry Davis

Please note, no offense intended, I just like this quote to describe Go success.

I feel like the author is fed up with being disagreed with, yet makes very opinionated statements about how others should (or will, at some point in the future) feel. They make good points about problems in Go, but I gotta say they also make a lot of points that might never become a problem at all.

I've written code that runs in production for 20 years now and every language I ever shipped software with, from ASP, PHP, and C# to Swift, TypeScript and Rust, one thing is always true: there's a pretty good reason the language was picked, and yet a long list of grievances. The reason a language is picked is most often productivity, some belief that software will ship more safely and often if we pick this language. Of course that belief can be uninformed and proven wrong later. But it's really hard to say, for any language, that "you will regret picking this language".

Sure maybe you shouldn't pick PHP for your new startup backend because you'll have a harder time finding developers. Maybe you shouldn't pick Python because despite asyncio there's still that one call you accidentally did that will lock up your highly concurrent service. Maybe you shouldn't pick Java because it's got verbose factories and something about billion dollar mistakes. Possibly picking Rust is a mistake because your developer team will spend 80% of their time fighting the borrow checker. Picking Swift to do your backend seems rather odd, and C++ is too risky with all its undefined behaviors and segmentation faults. Beware of Node.js unless you're ready to dedicate beefy RAM instances to it.

That is of course all hyperbole, I've seen productive code in all of the above languages. And I should add that many of them were not designed any more than Go was, to address one of the author's concerns. I've picked Go many times since it hit 1.0 and I don't regret it, even with production code that is over 10 years old. Go does concurrent processes and I/O really well. The GC works very well for long-running programs with lots and lots of individual goroutines. I also saw developers that had never touched Go enter the code base and make productive edits, which I consider a benefit of Go's simple design (or non-design, as per the author). I've also done similar systems in other languages, especially Node.js, Python, and Rust, but they haven't been as productive in my experience for this use case.

Ironically code written in Rust did wake me up in the middle of the night more than Go did, possibly because it was harder to express the mental model the way originally intended, and instead the code was written to satisfy the computer's memory model. I would still use Rust every day for things where that matters more, but the author seemed to imply somehow avoiding Go avoids unexpected errors, but they happen in every language.

TLDR: Programming languages live on because they make people feel productive. Maybe this feeling is aligned with the business needs, maybe not. But don't go saying that everyone should avoid a language entirely because of your personal experiences.

Took some time to read the linked Tailscale blog posts as examples of horrible hacking around Go. And I would agree it's crazy for a product oriented startup to do this, but this team is also made up of people that have contributed a lot to several core parts of the Go runtime.

Since these blog posts were written, Go has made official the interning approach that the netaddr.IP type depended on, now without the horrible pointer hacks, and the IP type itself has been migrated into an official package with even more improvements, now benefitting everyone. The iOS network extension memory problem was a very nice deep dive into what happens behind the scenes (something that is assuredly not perfect in any language) and resulted in a better Go experience for everyone using the language.

So a team pushed the limits of a language and made the language better for everyone out of the box. This happens in most languages, but rarely do you see such approachable blog posts about it, where the solutions are readable in the language itself, and with enough commentary even understandable.

I recommend reading these articles as examples of good language evolution, not as examples of a broken language.

Context for those unaware: This article was a response to a 2022 HN thread [1] about the original article of the author's. Then the following discussion happened: [2].

First off, let's appreciate the fact the author has managed to write this humongous article in a day. That's some writing skill.

And indeed, leveraging writing skills to unleash a flood of words in order to overwhelm anyone trying to respond is just the first tactic of many used by the author to sound more convincing, which is going to be the point of my post. I'm going to analyze each rhetorical device, fallacy, and bias present in the article.

> The author is a platypus

The point of this paragraph is not to present any arguments, it is devoid of them, but to elicit sympathy from the reader for the poor author who is clearly getting ganged up upon. Comments in the referenced HN thread did not in fact attack the credentials of the author in any way. See Appeal to pity [3].

> Mom smokes, so it's probably okay

Lots of words here, but the gist of it is:

> Or you can be horrified, as you realize that those complex problems only exist because Go is being used. Those complex problems would not exist in other languages, not even in C, which I can definitely not be accused of shilling for (and would not recommend as a Go replacement).

Setting aside the validity of the technical arguments, the tactic used here is to imply that since certain other languages don't have the specific mentioned problems, they are strictly superior. In reality, had another language been used, the Tailscale team would simply be dealing with different problems. To riff off the C example, if Tailscale used C, they'd be dealing with memory safety issues instead.

> The good parts

A clever attempt at a lie of omission. First we list a tiny subset of the pros of Go, then we claim that this tiny subset is not enough to make up for the numerous cons. See Cherry picking [4] and Straw man [5].

> And since, just like C and Java, you do not get to decide what is mutable and what is immutable (the const keyword in C is essentially advisory, kinda), passing a reference to something (to avoid a costly copy, for example) is fraught with risk, like it getting mutated from under you, or it being held somewhere forever, preventing it from being freed (a lesser, but very real, problem).

Here we see a thinly veiled reference to the main bias behind the article. Can you guess what language the author is implicitly talking about here?

> Go is an island

Nothing really interesting here, but it's worth noting that, just like the majority of arguments presented so far, this largely applies to most languages out there (but of course not to the language we're really talking about here).

> All or nothing (so let's do nothing)

The thrust is another strawman, quite simply nobody argues this: "It's not like you can prevent all problems anyway.".

This section is only a lead-up to the next one anyway...

> "Rust is perfect and you're all idiots"

Huh, why do we suddenly have a tirade about Rust in what's supposedly an article on Go? If you correctly guessed earlier that this entire article is actually about Rust, congratulations.

> Folks who develop an allergic reaction to "big balls of mutable state without sum types" tend to gravitate towards languages that gives them control over mutability, lifetimes, and lets them build abstractions. That those languages happen to often be Go and Rust is immaterial.

Wait a second, language*s* with sum types, control over mutability, and lifetimes? So, Rust, Rust, and Rust?

> Because function signatures don't tell you much of anything (does this mutate data? does it hold onto it? [...]

Only Rust is good enough, we get it already.

---

Overall, we have quite a few strong technical points here that could have made for a compelling article, but they're marred by copious use of fallacious arguments and blind fanboyism for a certain language.

---

[1]: https://news.ycombinator.com/item?id=31191700

[2]: https://news.ycombinator.com/item?id=31205072

[3]: https://en.wikipedia.org/wiki/Appeal_to_pity

[4]: https://en.wikipedia.org/wiki/Cherry_picking

[5]: https://en.wikipedia.org/wiki/Straw_man

Yeah I mean if you care about correctness at scale Rust is basically the only imperative language worth using, due to its separation between mutable and immutable state. I believe having that separation is a fundamental requirement for correctness at scale. I would love to see a simpler language than Rust have that kind of separation, but that doesn't exist at the moment.

Not everyone cares about correctness at scale! Python for example is a perfectly fine language to use when getting something out is more important than being correct at scale. A lot of game dev can also live without full correctness because things like player positions can just be fixed up afterwards if there's a bug. But if correctness at scale is a priority, Rust is the only serious imperative option.

Erm..

> As it happens, I am not a junior developer, far from it. Some way or another, over the past 12 years,

I'd say that's isn't too far from it to be fair.

[meta] Does anyone else color schema? I found the blue links hard to read with the dark background.

Or maybe the links are left to the default for the browser to decide.

Either way I would prefer an easier to read color.

Golang it's a huge career mistake. You can comeback in 5 years to this message to confirm the obvious.

Sorry to spoil the party, but the future is mobile and none of your favorite programming language ( Rust, go, Python, Nim, Lisp, Zig, Smalltalk ...) is ready for this. The future will be Kotlin, Swift and Javascript and I already hate it.

The present is already mobile but the backends serving those mobile devices run on servers. There's also no reason mobiles can't run Rust/Go/Python - Android already runs Java.

You are right there is no reason mobiles can't run Rust/Go/Python. But actually they don't, because there is no infrastructure and no tooling to build mobile apps using these languages. And this is the big oversight of the creators of these languages. All these languages we love so much as developers will die. Not because they are missing generics or error handling or coroutines or a static type system or other nifty language features. They are simply missing tools to create mobile apps.

I think you're vastly overstating the case saying these languages will die..

The majority of code we write runs on servers, even if its ultimately in service of a mobile/web client. Even if mobiles didn't support them, that doesn't mean these languages will die. The servers still have to run.

Creating the tools to run them on mobile is a solvable problem. Rust is already supported in the Android platform [1]. It takes more work to write an application in a non-standard language but its doable. Just saying don't worry about every other language dying out because of mobile, that won't happen :P

[1] https://source.android.com/docs/setup/build/rust/building-ru...

> I think you're vastly overstating the case saying these languages will die..

Sometimes you need to wake up people. Because I don't want to live in a world where we have Javascript in the frontend and in the backend. But this is going to happen if the creators of other programming languages are continuing to ignore development for mobile apps.

Absolutely. New programming languages keep coming in. But what survives for long term are those which help solve business problems. Love them or hate them, but programming languages like Java, JavaScript, Kotlin, Swift do them more than any other languages for the Servers, front end and mobiles. Then there is Python ecosystem for the huge AI job market. Wearable devices take the challenge to next level along with all those advancements in mobile space. All these cool languages have a long way to go in building the eco system to be effective at these landscapes. By the time they try to get there, there will be new other languages that would be termed 'cool' at that point in time. And, the problem landscape will be elsewhere altogether. Then these cool languages would no longer be cool by then. Nor would they be addressing the growing landscape with their just growing eco system. they slowly fall into identity crisis and die down. I'm sorry, what a grim picture I'm portraying :-)

The author clearly does not weigh in on the "whys" some things work in Go like they do.

So, adding an extra struct field results in the base value, and by design this should be used as a base default. The go proverb make base values useful goes hand in hand.

Obviously functions are not the same. And calling function with the wrong args is a compile time error. There is also CI check that check for this use case.

Overall this seems like a comparison to Rust, and its obvious they are not same same. Apples to Oranges.

Rust is also not the same as C, and therefore shall not be compared to it. Crabs to Cats.

Of course they are not same; that's why we want to compare them in the first place: were they the same, there'd be no need to compare them at all. Since, you know, they would be same and have no differences.

I really wanted to like go, and I tried to write a discord bot using it, but the very opinionated brace style (which isn't the one I prefer to use), and the fact that I struggled to much to try and split my code across two files kinda turned me off it. In the end I just went back to python

Python has the most opinionated brace style though

You can abuse list comprehensions to enclose your code in [ ] instead of having to deal with all that whitespace.

oh that's evil :)

Genuinely curious, why am I being downvoted? Did I say something wrong?

Probably just your opinion being shallow.

I mean, come on, coming to Go and of all things complaining about brace style? Coming to Python and of all things complain about syntactically significant whitespace? Coming to Lisp and complaining about parentheses? This is so old.

Of all the things you can complain about the language, how could you choose the thing that Go got right? It's K&R style is probably the best.

AI can rewrite any dream code which u write and putting ur brain on that is pure seless way of pointing start with some island as you pointed get to scale and in 6months every code can be rewritten how many ever time u want..tokens will be so affordable to run full context code base shift..don't worry about it just focus on distribution for ur product with the code you right..thats a good thing to worry about.

Reply on news.ycombinator.com