Go: What is the difference between make and new?

make — allocates and initializes internal data structures for slices, maps, and channels. new — allocates zeroed storage and returns a pointer, but does not initialize complex types.

make is the only way to create a usable slice, map, or channel. It returns the type itself (not a pointer) and sets up the internal bookkeeping — for slices it allocates the backing array and sets length/capacity, for maps it creates the hash table, for channels it allocates the ring buffer.

1s := make([]int, 0, 100)  // len=0, cap=100 — ready to append
2m := make(map[string]int)  // initialized, safe to write to
3ch := make(chan int, 10)   // buffered channel with cap=10

new(T) returns *T pointing to a zero-valued T. For simple types like int or bool this works fine, but for slices/maps/channels the zero value is nil — you still need make to make them usable.

1p := new(int)     // *int pointing to 0
2*p = 42
3
4q := new([]int)   // *[]int — points to nil slice
5*q = make([]int, 0, 10)  // must call make

When to use:

• make — whenever you need a slice, map, or channel.
• new — simple value types when you want a pointer (less common; idiomatic Go prefers literal initialization).

Common mistakes:

• Using new([]int) and then append without make — you get a nil slice panic.
• Confusing make([]int, 10) (len=10, all zeros) with make([]int, 0, 10) (len=0, cap=10).
• Calling make on a non-slice/map/channel type — compile error.