Status Updates From Ultimate Go Notebook

知道data semantics，便能明了data behavior

receiver semantics && builtin types receiver should be value receiver

chapter 2 看完了

padding and alignment

重新看一遍

begin study profile code

begin study the benchmark

study testing

channel pattern: retry timeout

study the pattern of channel: fan out/in、cancel、drop

study mutex and channel, especially the pattern of channel

开启concurrent编程

interface pollution，我们要牢记：开发者并非创造了interface，而是发现interface。

使用composition interface重构之。

Don’t design with interfaces, discover them". - Rob Pike

value pointer receiver.
作者一直强调 value/pointer semantic，包括之前在 range 时，以及现在的 method receiver。

开头就讲了CPU Cache 与 Cache Line。

做了三个benchmark：分别是遍历linked-list、行方向遍历矩阵、列方向遍历矩阵。

让我惊奇的是，遍历linked-list比列向遍历矩阵还快，更准确的说，它只比行向遍历矩阵慢一些

> The linked list is twice as slow as the row traversal mainly because there are cache line misses but fewer TLB (Translation Lookaside Buffer) misses. A bulk of the nodes connected in the list exist inside the same OS pages.

看完第二章，大概的讲解了一些知识点，比如gc、iota、pass by value、escape analysis、assign、padding & alignment

第二章：零值；字长；内存对齐

看完第一章，需要重复查看。