Idea to improve the monitor command documentation

[v3sk0o0]

Hey,

I was working on a local setup, testing the monitor command with Redis.

While I could successfully capture all changes via the monitor command in redis-cli, I encountered an issue where no changes were captured after a certain period using the monitor command in the Go client.

After referring to the documentation, I found that the client needs to be blocked during the monitor command, and for prolonged monitor operations, the socket timeouts in Go client. After debugging, I discovered that this problem stemmed from the default configuration when initializing a new client. I addressed it by adding the ReadTimeout option with a value of -1.

From the doc:

// Timeout for socket reads. If reached, commands will fail
// with a timeout instead of blocking. Supported values:
//   - `0` - default timeout (3 seconds).
//   - `-1` - no timeout (block indefinitely).
//   - `-2` - disables SetReadDeadline calls completely.
ReadTimeout time.Duration

Before:

rdb := redis.NewClient(&redis.Options{
		Addr:        "localhost:6379",
	})

After:

rdb := redis.NewClient(&redis.Options{
		Addr:        "localhost:6379",
		ReadTimeout: -1,
	})

Proposal:
I suggest adding an additional help message to the monitor command, the monitor_test or stating explicitly how to make the connection blocking

Setup to reproduce:
go: 1.22
arch: arm64

Kind regards,