Define open_timeout and read_timeout independently

Hi, is there any reason why we can't define open_timeout and read_timeout independently?

I would be happy to implement it; I think the following code would be enough:

```ruby
    def open_timeout=(new_timeout)
      @http.open_timeout = new_timeout
    end

    def read_timeout=(new_timeout)
      @http.read_timeout = new_timeout
    end
```

