[44:09] There is a feeling in the community from Python or Ruby world that static typing is bad idea, Because they learn what static typing is from languages that did it badly. Like C, C++, and Java, where it feels like there are some bureaucratic form every time you start writing a program.
I had a bit of epiphany about a year ago when I realize that the last decade testing has become hugely recognized as a critical part of software development. Ironically it was NOT all that big of the deal (there were advocates for it, but didn’t catch on as an important feature) until last decade or so.
And the realization I had was it was triggered by the dynamic language community. using tests to do the type checking for them. They write in this dynamic language, that they write a test to make sure that strings never turn into lists and vice versa. and it’s kind of backwards.
Static checking gives you that at compiles time, You program won’t compile if you do that kind of thing. That means the testing can be functional rather than type checking. That actually makes you write fewer tests and just get the same good coverage.
@sunfmin 其实我还想请教一下，有没有对 Golang 中错误的处理的一些例子。在 Golang 中高频率的 error 返回有的时候让我好烦...
#1 楼 @wppurking 这个那个音频里面也有讲到，还有这里：http://golang.org/doc/articles/error_handling.html
@sunfmin 这里有看过。举个例子，比如碰到 io 方面的 error 时，我真的超级想全部写成
bytes, _ := ioutil.ReadAll(xx) 不过还是得写一个
if err != nil {
return xxx, err
}
因为有时候，只想把 error 弄到上层去，可这不是 panic 所以只能一直这样 return 上去。
想请教下你们在编写 Qortex 时，错误是如何处理的呢？是否可以举一个例子呀？
而且如果你在设计接口，给别人调用，那基本上就要返回 error 吧，你的接口里面出个 panic，一点不明确啊，返回 error 的话，用你借口的人知道调用这个东西需要处理错误，会去想如果返回错误了，我的系统应该怎么处理。
再比如，调用你的方法，是通过 RPC 调用的，error 就更容易返回给 socket 的另外一端。
