[triodan]

This must've been quite an old article, because lo and behold, nearly every single complaint on that page is being addressed with gusto by the Go community! Proposals on generics and error handling is well underway, and his complaints regarding GOPATH and the go get has been solved with Go modules.

EDIT: As I expected, the linked page is written on Feb 2018[1], things indeed have changed a lot since then.

[1]: https://github.com/grimoire-ca/bliki/commits/master/wiki/dev...

[xg15]

Sorry, but the language existed for how long? I remember hearing about the same complaints about generics and error handling years ago - and they got answers back then with about the same level of arrogant dismissal like the quote about syntax highlighting in the article.

It's good if the community is at least starting to listen now, but I don't think that is sufficient evidence that the language is good.

[tinco]

There is no discussion if the language is good, it's already quite convincingly proven it's good. The question was if generics would make the language even better. Some people think that's a no brainer, others are not so sure, and the latter group included the core maintainers.

People use the language because it is good, I think at this point one can state that as objective fact. Once people use it, they'll complain when something is not as nice as it could be, like the error handling, and like the generics.

I wonder if in 2000 people felt the way about Java 1.3 people feel about Go now. Java had a similar rise to prevalence, and it also lacked some features we now consider crucial, such as generics. I personally would never want to go back to Java 1.3. I feel I'd much rather code in Go instead, even if it lacks generics. But I still think it's worth considering if Go would turn out to become Java, if it now embraces generics.

[edem]

The language is crap. Just because people are using it does not mean it is good. Take a look at javascript. Same deal. You did not state any "fact"s why Go is good. Maybe it is good compared to Cobol or Fortran, but not compared to pretty much everything else.

[tinco]

Oh come on. People are forced to use Javascript, I don't think it would even exist today if it wasn't the only crossbrowser language.

Almost no one is forced to use Go, and yet people flock to it. I don't feel I need to state the things that make Go good because it's already been done a thousand times, and everyone in this thread already knows why it is good.

Have you ever built something in Go? I'm almost as productive in Go as I am in Ruby, and that was within a week of learning it. It's crazy good for a language that lacks so much expression.

[ernst_klim]

>People are forced to use Javascript

People are not forced to use Javascript since there are a dozen of great LangName->Js infrastructures.

People are not forced to use PHP or other badly designed languages as well.

I wouldn't call Go a bad language due to the subjectivity of the term, I would rather call it ad-hoc language, which is basically a DSL sufficient for some domain, but not very suitable for a general engineering.

Go is a compiled PHP, its rather weak type system and oversimplified nature are suitable and sufficient for webdev for the economical reason. Building something complex in Go is possible but painful, Go doesn't help here but rather impedes making you write your own dynamic type system and other facilities to compensate the language's deficiencies.

[skohan]

To be fair, I feel like go's simplicity makes it easy to get small things up and running, but the pain comes when your project starts getting large. That's when additional abstractions are missed.

[tinco]

I don't think Go should be used for larger projects. Apparently though the community disagrees because they're really adding generics.

[cmlndz]

If usage is a measure of ‘goodness’ (whatever that is), then PHP must be the best programming language the world has ever seen. And we know it’s not.

[YeGoblynQueenne]

>> Almost no one is forced to use Go, and yet people flock to it.

Well, at some point people "flocked" to javascript, and they weren't forced to. So it's a good language, then? Or sub javascript for java, or your favourite bashable language.

[hollerith]

I will assume you're referring to Node. The browser explains why many coders know Javascript well; the fact that many coders know Javascript well explains the popularity of Node --without our having to resort to the hypothesis that Javascript is a good language.

[kazinator]

Indeed, you can program the browser in whatever you want, yet they insist on using JS.

[edem]

Plot twist: people never "flocked" to Javascript. People use Node for example because the company they are working for thought that it would be nice, efficient and less costly to have one language for the whole platform. They mostly learned their lesson since then and moved on to real languages which can be used on the backend.

[fierro]

"Go is a crap language, you didn't even state any 'facts' why Go is good!"

>> does not provide 'facts' why Go is crap

[dvfjsdhgfv]

> There is no discussion if the language is good, it's already quite convincingly proven it's good.

I'm sorry, you can't prove a language is good or bad. You can prove it's popular, widely discussed, used by several individuals and organizations, is actively developed, but "good" and "bad" are inherently subjective terms and there's no way to prove it unless there happens to be a universally accepted definition of what constitutes a good or bad programming language.

[spicyusername]

> a universally accepted definition of what constitutes a good or bad programming language.

Implicitly many people, rightfully in my opinion, conflate the utility of a programming language with that language being good. If many people are able to solve real problems with a language, that language is good, the end.

Good does not equal perfect, there is no perfect in the real world of engineering.

[dvfjsdhgfv]

On the other hand, many people conflate the popularity of a programming language with it being backed by a big corporation.

There actually are a few measurable metrics of programming languages, but even when we discuss such a simple factor as execution speed that you'd expect to be universally accepted as positive, there will be people arguing that developer time is more expensive and savings made here are more important than the gains on execution speed. There is simply no way two programmers are going to agree in classifying a number of languages as good or bad.

[spicyusername]

> There is simply no way two programmers are going to agree in classifying a number of languages as good or bad.

If there is one thing humans love doing is finding small reasons to disagree hah.

[joelfolksy]

So in the world you're describing, pretty much all languages are "good," and none are "perfect." Not very interesting so far. What are the adjectives that fall in the middle, the ones that some languages deserve and others don't?

[vonseel]

One might also wonder how long it took Java to respond to user complaints, and I would suspect Go core team has been much faster addressing these issues and responding to developers.

[miaklesp]

It's good only for those who didn't see anything better.

[vonseel]

Are you really arguing that the language is not “good”? Sure, there are things not to like about it, but the widespread adoption of Go as a systems programming language is evidence alone that it is a “good” language.

Good is such an ambiguous term that I’m not sure it’s even worth using. For one person, only esoteric languages like Haskell are “good”, for another person, Python or Javascript are “good” enough.

[62747478182]

> but the widespread adoption of Go as a systems programming language is evidence alone that it is a “good” language.

Years ago PHP was the king of web development, and it was pretty awful in many ways (today is much better, though).

People usually choice languages because of convenience, not because of how good they are.

[eptcyka]

Judging by most responses form the Go team and/or the community around the language, it would seem that none of the issues listed in the article are actually valid. I.e. this is the _correct_ way to handle errors, sum types bring nothing, etc.

Of course, things have changed since, but what does that say about the previously strongly held opinions of the team?

[grumpydba]

> Judging by most responses form the Go team and/or the community around the language, it would seem that none of the issues listed in the article are actually valid

Generics have been considered for years by the core maintainers. See Ian Lance Taylor's proposals. So you're either lying or not informed enough to be so assertive.

[dragonwriter]

> This must've been quite an old article, because lo and behold, nearly every single complaint on that page is being addressed with gusto by the Go community!

“Being addressed” = “not yet effectively addressed” = “still an issue”

[edem]

It it is underway it is not solved. Go is crap.

[akmittal]

There are only two kinds of languages: the ones people complain about and the ones nobody uses. -Bjarne Stroustrup

[joelfolksy]

"There are two kinds of men: the ones complained about by their wives, and the ones who are single" - Me, before sleeping on the couch

[edem]

I don't see how this adds any value to the conversation.

[akmittal]

It says no language is perfect there are some or another compromise in all.

[eeZah7Ux]

and Python.