2012.1129 permalink

An apology

About six months ago I wrote a now-infamous post about why I left Google. I still stand behind most of what I wrote, but after thinking about it I need to address one point in particular: "Go failed to solve any significant problems."

I'm sorry I wrote this and I think it shows a significant lack of perspective. First, I don't know Go as a language, and I didn't then when I bashed it. When I wrote the post I was still frustrated with the unimaginativeness of Dart, and Go, from what I had seen, looked like another Java-ification of an otherwise interesting language (Erlang).

Let's suppose for a moment that I was right about that much. Even in that case, I don't think my point had much merit because Erlang has a number of problems that make it difficult to use in practice. The two that I've heard about most commonly are obscure syntax and suboptimal performance. If Google managed to solve just these problems with Go, then they've changed the landscape of programming languages enough to make writing concurrent programs much more approachable to normal people (i.e. people who don't get excited about phrases like "process calculus").

Anyway, I shouldn't have written this about Go. Google has started a number of interesting compiler/language projects and the only one I actually know enough about to criticize is Dart (which I still think is a disaster, but that's another rant). All of the others I've looked at have been meritorious in some significant way, whether as a matter of design or, more commonly, of exceptionally thoughtful implementation.