> it feels at times like entire community wants to pretend that there aren’t any flaws or downsides with the language.

Deeper: This will be controversial to some readers here. I wonder if the reason why the "entire community" resists criticism so fiercely is that Rob Pike has a very public profile and "infinite Internet (reputation) points", so no one can beat that final boss. I'm not saying that Rob Pike would use his Internet boss powers to kill off any critics, but I do think it does repress some critics. Also, his style and opinions are very strong.

All said, as a non-Golang writer, but avid reader, I am still very impressed by how carefully that Golang was developed. It could have gone so much worse (C++? JavaScript?), or the community been much more toxic/unwelcoming, e.g., Rust. Overall: It is a huge win for current and future programmers.

What makes you think that? The Go team is clearly working on trying to crack the nut. There is, as of about a month ago, even now an exploratory implementation: https://github.com/golang/go/discussions/71460 It is not an easy nut to crack. Many serious proposals have failed for good reasons. But it is abundantly apparent that the effort is there and ongoing.

The "golang detractors" aren't acting rationally. If you watch carefully you will see that most of the time they don't even understand their detraction points, indicating that they are simply repeating what they saw elsewhere. The unfortunate reality is that they aren't looking to contribute anything meaningful, they are simply trying to be funny and they will continue to believe it is funny until "You idiot, Go already has that" can be said and manages to wear them down.

> three lines of mostly-identical error handling around every function call is optimal

To be fair, while I am sure we could do better, I am not sure we have found it yet. I do my programming in various languages with all these fanciful error handling semantics and it is not good. I regularly wish those languages' idioms had settled on something like if err != nil over what they ended up with. It is not great, but better than the rest. But it is encouraging that smart minds are working hard trying to find better.

The one exception to this trend has been generics. I think it was an inevitable addition to the language, but there are still a lot of holdouts who argue it wasn’t necessary. I think this is probably because it’s more important for library authors than app developers. But it does remain by far the most divisive update to the language.

> The "golang detractors" aren't acting rationally. If you watch carefully you will see that most of the time they don't even understand their detraction points, indicating that they are simply repeating what they saw elsewhere.

I personally don’t think I’ve seen a lot of that. At this point it’s endemic enough that there’s a small and shrinking number of engineers each year who haven’t spent time in it.

> To be fair, while I am sure we could do better, I am not sure we have found it yet.

Everything below here is a more or less reasonable take (even if I have mild disagreements), and I think a lot of people would feel less likely to snipe at the language if more people expressed opinions like this.

One example, if I may: the errors.As function is far from ergonomic. It would have been much better if generics had come to the language before that function was added to the standard library. Modern alternatives exist: https://pkg.go.dev/github.com/jub0bs/errutil

Where is this mantra emanating from, exactly? Go officially became community driven in 2017. If the community didn't seek this, nobody, especially not the core team, would be working on it. Actions certainly do not support your idea.

> there are still a lot of holdouts who argue it wasn’t necessary.

Of course. Obviously it isn't necessary, or even beneficial, for all programming problems. The software industry is quite broad and developers are working on all kinds of different things. One is only going to speak from their own experience. It is all they know. And I expect error handling is contentious in much the same way because of it being something that crosses many different types of programming areas in different ways.

In fact, despite what may seem contrary to what I said earlier, there are classes of of software problems where you do think about errors differently. For example, letting the program close with a message and letting a human deal with whatever went wrong is a perfect solution to many error-related problems and language constructs that make that easy is wonderful. In those cases the fancy error handling semantics are quite appreciated.

Naturally, my assertion that something like "if err != nil" being the best we've got comes from the type of software I normally find myself needing to write. If you work on different types of problems, you're bound to have a different perspective. Expecting to find agreement here is much like expecting basketball players and baseball players to find agreement about what kind of ball they should share. But there is no need to.

It's abundantly clear where the actual Go community stands on this. They have a visible interest in seeing what can be done; to try and find out if there is a better way. That does not guarantee success, but the desire to try is undeniable.