Rob Pike recently wrote a great blog entry describing the zen of Go and how it compares to a few other languages, particularly C/C++:

http://commandcenter.blogspot.com/2012/06/less-is-exponentia...

Here's the YC thread on that. http://news.ycombinator.com/item?id=4158865

Rob Pike is a great communicator.

I'd particularly recommend "Meet the Go team". Some really interesting questions, and most importantly answers. In the video :)

http://www.youtube.com/watch?v=sln-gJaURzk

I liked the sense that Go 1.0 was a genuinely meaningful milestone, and feature talk is taking a back seat to actually using it in anger for a while.

If it only had generics... that would make it a really viable option.

Have you actually built any real projects in Go? Because pretty much everyone I know who has tried (and who has learned enough Go to write minimally idiomatic code) has found this to be a non-issue.

Not all languages are the same, and not every "feature" makes as much sense in every language.

Before using it in a real project, it is obvious to do some research, isn't it? So some questions and concerns arise, naturally. Regarding generics, I really don't like what Go offers instead [1].

By the way, creators themselves haven't eliminated the chance of adding generics in later versions of the language [2].

Also, I don't understand why not have destructors (deterministic, not the finalizer...) in many cases where currently `defer` is used.

I also don't like the trick you have to do for checking the interface integrity [3].

All in all, I have a sympathy for Go, lets see what will happen in 2.0.

[1] http://stackoverflow.com/questions/6255720/making-generic-al...

[2] http://golang.org/doc/go_faq.html#generics

[3] http://golang.org/src/pkg/encoding/json/stream.go#L187

Go advocates and sympathizers would do well to drop this fully general riposte.

Does anyone know more about the virus written in Go that was mentioned in this talk?

ken knows more ;)

Performance is a legitimate complaint [1]. To be fair, Java generics suffer the same problem.

[1] http://shootout.alioth.debian.org/u32q/benchmark.php?test=al... - much slower than Java (memory figures may not mean much except that Java was tuned to use a large heap for better performance)

64-bit benchmarks are more fair, since Go's 32-bit x86 target hasn't received much attention (it's irrelevant to most users).

http://shootout.alioth.debian.org/u64q/benchmark.php?test=al...

Most of the performance difference is because Go's GC isn't very good, which is something that they're working on.

I would have loved to have scene what Rob would have said in the next hour. I feel like every time I see a talk or slides about Go, I get giddy that I might get to see more patterns or more "good ways to use channels effectively". I enjoy Go but I still find myself writing things several times with channels before it's as clean an readable as I like.

Sounds like you have enough experience now to write some really good blog posts about it!

"Here is how to write a monad using Go channels".