3 ms·

I find anyone using Go for networking code bizarre. It's bizarre to me the language ever caught on especially because all their design goals are explicitly the

I find anyone using Go for networking code bizarre. It's bizarre to me the language ever caught on especially because all their design goals are explicitly the wrong goals. Their goal was to make a "simple like C" language which simply disguises the complexity in writing software. Go simply punts complexity to technical debt of any project and assumes you will throw out your code after a year of using it.

AnimalMuppet 8y ago
You are very much wrong, on several fronts. First, you obviously have some measuring stick for what the "right" language design goals should be. But what you don't seem to recognize is that other people can validly have other design goals. It's not "your way or they're wrong". In fact, given the decades of experience the designers of go have (and wide variety of languages that they have experience with), it's almost certain that you know far less than they do. And yet they still made different choices than you would. Instead of wondering how they could be so stupid, that should make you wonder what they knew that you don't. (I've seen some rants from people saying stuff like "they couldn't have made that design decision if they knew anything about Modula 2!" And they miss the talk by Rob Pike where he said (paraphrased) "don't think we're so smart for coming up with that object file format - we stole it from Modula 2". They knew it at a very deep level - almost certainly better than their critic did.) Then there's this: > Go simply punts complexity to technical debt of any project and assumes you will throw out your code after a year of using it. Go was designed for multi-million line code bases that live for decades. Really. Read Rob Pike's notes on Go's design. So, yeah, there's a lot about this rant that is factually off in the weeds...

3 ms·

I find anyone using Go for networking code bizarre. It's bizarre to me the language ever caught on especially because all their design goals are explicitly the

by mlindner 8y ago

AnimalMuppet 8y ago
You are very much wrong, on several fronts. First, you obviously have some measuring stick for what the "right" language design goals should be. But what you don't seem to recognize is that other people can validly have other design goals. It's not "your way or they're wrong". In fact, given the decades of experience the designers of go have (and wide variety of languages that they have experience with), it's almost certain that you know far less than they do. And yet they still made different choices than you would. Instead of wondering how they could be so stupid, that should make you wonder what they knew that you don't. (I've seen some rants from people saying stuff like "they couldn't have made that design decision if they knew anything about Modula 2!" And they miss the talk by Rob Pike where he said (paraphrased) "don't think we're so smart for coming up with that object file format - we stole it from Modula 2". They knew it at a very deep level - almost certainly better than their critic did.) Then there's this: > Go simply punts complexity to technical debt of any project and assumes you will throw out your code after a year of using it. Go was designed for multi-million line code bases that live for decades. Really. Read Rob Pike's notes on Go's design. So, yeah, there's a lot about this rant that is factually off in the weeds...