Go: What is the difference between nil and empty slice?

Nil slice — declared without initialization; pointer, length, and capacity are all zero. Empty slice — initialized but contains no elements; may have capacity > 0.

Both behave identically for len, cap, range, and append. The difference shows up in JSON serialization and equality checks.

1var nilS []int
2emptyS := []int{}
3madeS := make([]int, 0, 10)
4
5// All support append
6nilS = append(nilS, 1)    // [1]
7emptyS = append(emptyS, 2) // [2]
8madeS = append(madeS, 3)   // [3]

JSON marshaling:

1n, _ := json.Marshal(nilS)   // "null"
2e, _ := json.Marshal(emptyS) // "[]"

When to use:

• Nil — uninitialized default state; no allocation.
• Empty — API responses requiring [] not null.

Common mistakes:

• Returning nil from APIs — clients see null instead of [].
• Checking only len == 0 when nil vs empty matters.
• append(nilSlice, ...) is safe — no panic.
• slices.Equal(nilS, emptyS) returns false despite both being empty.