go - Why use arrays instead of slices?

Question

I have been reading up on Go, and got stumped thinking about this fundamental question.

In Go, it is quite clear that slices are more flexible, and can generally be used in place of arrays when you need a sequence of data.

Reading most of the documentation, they seem to be encouraging developers to just use slices instead of arrays. The impression I get feels like the creators could have simply designed arrays to be resize-able, and done without the entire slices section. In fact, such a design would have made the language even easier to understand, and perhaps even encouraged more idiomatic code.

So why did the creators allow arrays in the first place? When would arrays ever be used instead of slices? Is there ever a situation where the use of arrays over slices will be compelling?

When I consulted the official documentation (http://golang.org/doc/effective_go.html#arrays), the only useful part I found was:

Arrays are useful when planning the detailed layout of memory and sometimes can help avoid allocation, but primarily they are a building block for slices.

They went on to talk about how arrays are expensive as values, and how to simulate C-style behavior with pointer. Even then, they ended the array section with a clear recommendation:

But even this style isn't idiomatic Go. Use slices instead.

So, what are some real examples of "planning the detailed layout of memory" or "help avoid allocation" that slices would be unsuited for?

Answer 1

Waitting for answers