Buffered vs unbuffered channels: semantics and use cases

Choosing the wrong channel buffer size causes unexpected blocking (unbuffered where you needed async) or hidden backpressure bugs (buffered with wrong capacity).

Understand the guarantees each provides:

// Unbuffered: send blocks until receiver is ready
// Guarantees synchronization — sender KNOWS receiver got the value
ch := make(chan int)
go func() { ch <- 42 }() // blocks until main receives
fmt.Println(<-ch)

// Buffered: send only blocks when buffer is full
// Decouples producer and consumer; provides a queue
ch := make(chan Job, 100)
for _, j := range jobs {
    ch <- j // blocks only when buffer full
}
close(ch)

for j := range ch {
    process(j)
}

Rule of thumb:

• Use unbuffered for synchronization/handoff
• Use buffered when producer and consumer run at different rates and you have a known bound

Unbuffered channels synchronize goroutines — the send completes only when a receiver is waiting. Buffered channels act as queues, decoupling timing but not eliminating backpressure when the buffer fills.