Whether you use ws or wss should be orthogonal to whether you're using IP or localhost

If you're using IP then make sure that the IP is actually reachable from your client – no NAT in the way, for example, and not using the “outside” IP address when the client is “inside” your firewall. Also, make sure that network interface is allowed through whatever firewall software you're using.

When it comes to wss this is what you're going to have to use on the greater internet (I don't know if browsers outright deny ws connections over the internet, but they might …) To make that work, you have to have a server certificate configured on your wss server, signed by a root certificate, that is trusted by your wss client (e g, browser.) And the subject (name) of the certificate must be the domain name that the browser connects to. For testing, using a self-generated root certificate, and installing that in your browser certificate store, and then using that to sign a generated certificate for your “localhost” service is fine. For live deployments, consider something like Let's Encrypt, which will require you to do some additional machination to make sure their servers see a web service that can verify domain ownership.

Other than that, make sure to use Wireshark to figure out what's going on, and if that's not enough, you have the source – crack open the library, add some logging/printfs (or breakpoints in the debugger) and figure out what's going on!