Here's an ultra-simple explanation of internet addresses for systems connected via an ISP.
This won't directly help anyone hook up an OT server properly, but it might make it easier to understand the instructions.
NB: there are some minor inaccuracies to make this easier to understand. Unless I've made a stupid typo, I already know what they are.
1. There are more private computers than there are IP addresses available for them
2. ISPs deal with this by giving their customers addresses which are reused
3. If you have address 10.xx.xx.xx or 168.192.xx.xx, this is what's been done. Those networks are "unroutable" - i.e. if they're used in the real internet (not your ISP's private network) the router infrastructure rejects the connections
4. ISPs usually only give each customer one address
5. Network 127.0.0.1 is another special case - every system can use that to address itself. Note. It doesn't behave exactly like the other networks, even an unroutable one like 192.168. It's very useful, but shouldn't be used to test real external connectivity. Notably, 127.0.0.1 does not go through your local router (see below for what a router does).
6. "localhost" is an IP host name available on every system, and it resolves to 127.0.0.1.
7. It's correct to call those xx.xx.xx.xx numbers IP addresses, but those are the address of the system. You can have multiplm programs using the networks at the same time, so you need more address granularity than one per system. This is achieved using a fifth number, called the port (those xx's can't go over 255, but ports go up to 64K).
So - if you have only one address, but 5 computers, and you've just clicked on a link that will return an artistic image to your browser ... how does the sending system find your computer if it can't use 192.168.0.5 or whatever?
The usual approach is called "port mapping" or something similar.
There's a component between your physical location and your external connection to your ISP. These are often called routers, but be careful - they're not the only things in the Internet called routers, and they often do more than just IP routing . It knows the addresses to use to talk to your ISP's infrastructure, and it knows the addresses of all the systems in your location, including (e.g.) your 192.168.0.5.
So, you click on a link on an external web page ... your request has a "return address" of 192.168.0.5:80 (port 80 is the default for HTTP) ... and that address is useless to the server that will send the data.
Your router has one address on the real network ("the internet"), and the web server with your page on it can address that - let's say it's 85.118.124.9
The router modifies your return address to 85.118.124.9 - but it can't just use port 80, because the other people using browsers at your location might also want port 80. Instead it picks a different port on 85.118.129.9 - lets say 15832. It "remembers" that anything it gets back on 85.118.129.9:15832 should actually be sent to your port 80 (i.e. 192.168.0.5:80), then sends the modified request to the web server. When the response arrives, it's modified (so the packet addresses are 192.168.0.5:80, as your system expects) and sent to you. The browser grabs the packets, assembles the picture, and displays it.
The other thing that can complicate this a bit is when you're running a server, where people want to connect using a specific port, rather than negotiate a port number. In that case you don't want the router to pick a random port number - you want it to pick the specific port(s) the external clients expect to use. The tutorial linked above explains how to make the router do this.
As i said above, I wrote this in the hope that it makes it a bit easier for at least a few OT users to read the tutorial on port forwarding.
If it makes sense, it's time to look at the excellent tutorial that was linked above.
If not, it might be time to talk to a friend who's good at configuring home networks
