Amqp: Somehow Channel.QueueDeclare "exclusive" and "durable" flags are mutually infuenced

Here is a minimal example:

package main

import (
    "fmt"

    "github.com/streadway/amqp"
)

func main() {
    conn, _ := amqp.Dial("amqp://guest:guest@localhost/")
    defer conn.Close()
    ch, _ := conn.Channel()
    defer ch.Close()
    q1, err := ch.QueueDeclare("test_qqq", /* durable: */ true, false, /* exclusive: */ false, false, nil)
    handle(err)
    fmt.Println("%+v", q1)
    q2, err := ch.QueueDeclare("test_qqq", /* durable: */ true, false, /* exclusive: */ true, false, nil)
    handle(err)
    fmt.Println("%+v", q2)
}

func handle(err error) {
    if err != nil {
        panic(fmt.Sprintf("%v", err))
    }
}

Using:

• amqp version v0.0.0-20181107104731-27835f1a64e9
• RabbitMQ 3.7.3 Erlang 20.2

As you can see, in this example I am trying to recreate same queue twice, first with exclusive:true, then with exclusive:false. It fails as expected, but produces unexpected error message:

Exception (406) Reason: "PRECONDITION_FAILED - inequivalent arg 'durable' for queue 'test_qqq' in vhost '/': received 'false' but current is 'true'"

I spent half an hour looking blankly at my durable: true statement, but actually should be looking into exclusive argument. Is there a way to keep future users from such a loss of time? It should be like args 'durable' and 'exclusive' are in conflict with each other. Or is it RabbitMQ problem?