When you are seeding, you broadcast to other peers that you have pieces available. The most efficient way to exchange data is for them to open a connection to you. Without an open port (from port forwarding) they have no way to make this direct connection.

Sharing Linux ISOs over BitTorrent requires connecting to other people’s computers to download the file from them.

Ports are what they sound like, ports. Like USB ports on a computer. They are places where data can travel through. But they’re also specific and unique, which can be useful.

I wrote up a big thing explaining it further but it got complicated quickly.

The basic point is, ports are a quick and easy way to determine WHERE a piece of data or request for data should go.

Your computer is on a network, and has an internal address on that network. Let’s say it’s 192.168.1.1. The Internet is a whole separate network, with its own addresses.

Your modem is assigned an address on this Internet network. Let’s say it’s 56.56.56.56.

Ok, simple, right? You have a Linux ISO I want, so my computer sends a message to yours, at your Internet address of 56.56.56.56. Theoretically that should work fine (it probably wouldn’t for complicated reasons but I digress). Your modem gets the message, forwards it to your computer which is the only computer connected to it at 192.168.1.1. Bob’s your mother’s brother.

But wait, you don’t have one computer, you’ve got an Xbox too. So you install a router, a special computer that lets you share your modem with other computers. Your Xbox has been assigned an address of 192.168.1.2. So who gets the request for the file now? The request was sent to 56.56.56.56, but the router has two options to deliver it to, and no way to know, so the request gets dropped.

To solve this, we set up port forwarding. You tell your router that all requests sent to port 5000 should be directed to 192.168.1.1. And then when you broadcast the availability of the Linux ISO, along with your Internet address you also say to use port 5000.

So now when I want a piece of your file, I send the request to 56.56.56.56:5000, and your router sees that and says hey I know where that goes. Then your computer gets the request, and happily sends the file to me.

Now let’s say you use a VPN. You have the same problem. The VPN has one Internet address, but hundreds or thousands of people using it simultaneously. If you don’t have a port assigned to you, then requests for your file will get dropped by the VPN provider, because it’s impossible to know it was meant for you.

In all these examples, you can still download FROM people, even without a pre established port. I’m honestly not entirely clear on how this works. But I’m pretty sure it has something to do with your router remembering that you just sent data (in the form of a request) to the other persons address, so when data returns FROM that address, it knows where to send it. I think. There could be, and probably is, some other trickery going on there. But I’m working from memory here 🤷‍♂️

Now there are other ways around ports, and other complicated details that are probably important. And things like CGNATs that basically break this fundamental function of the Internet. And I might’ve gotten some of the details I’ve told you wrong. I’m happy to be corrected.

But what I’ve said is essentially true, and should help make a clearer picture for you.