Why P2P file transfer is the only sane way to share files

Picture this: you and a colleague are sitting at the same desk. You need to hand them a 2 GB video file. The logical thing — the human thing — would be to hand it over directly. Instead, the modern workflow says: upload it to Dropbox, wait for the upload, send a link, wait for them to download it. Two round trips to a data center across the country, just to move bits three feet.

We've been trained to accept this as normal. It isn't.

The cloud default is a workaround, not a solution

Cloud file sharing solved a real problem in the early 2010s: devices couldn't talk directly to each other. Firewalls, NAT, different networks — the internet wasn't built for direct connections. So companies built servers in the middle. You upload to them, the other person downloads from them.

It worked. It still works. But it comes with baggage we've learned to ignore:

• Your files sit on someone else's servers. Who has access? Who can be compelled to hand them over? How long are they retained?
• Speed is bottlenecked by the slowest server. Your 1 Gbps connection doesn't matter if the upload server is throttled at 20 MB/s.
• File size limits are arbitrary business decisions. There's no technical reason a free Dropbox account can't hold 100 GB. The limit exists to sell you a subscription.
• You need an account, a browser, often a credit card. Sharing a file has become a commerce event.

What peer-to-peer actually means

P2P file transfer eliminates the middleman entirely. Your device talks directly to their device — the way computers were always capable of doing. There's no upload. There's no server holding your data. There's no account required on either end beyond the one you use to find your contacts.

When both devices are on the same network, the file moves at the full speed of that network. We've seen transfers hit 900 MB/s on a gigabit LAN. That's not a marketing number — that's just physics. No server is in the way.

When devices are on different networks, modern NAT traversal techniques (STUN, TURN, hole punching) can establish a direct connection in most cases. File Yeet uses the iroh protocol under the hood, which handles all of this automatically. When a direct connection truly isn't possible, an encrypted relay picks up the slack — but even then, the relay can never see your files.

Privacy isn't a feature — it's a structural property

With cloud file sharing, privacy is a policy. "We don't look at your files." "Your data is encrypted at rest." These are promises made by people who have your files. You're trusting their word.

With P2P, privacy is structural. The server never has your files, so it can't leak them, can't be compelled to hand them over, and can't have a policy that silently changes. There's nothing to trust because there's nothing to betray.

File Yeet encrypts every connection end-to-end. If you're on a relay, the relay sees only encrypted bytes — it has no key, so even a hostile actor with full access to the relay infrastructure gets nothing useful.

The case for going back to basics

We're not saying cloud storage has no place. It's great for backup, for accessing files from multiple devices over time, for collaboration on living documents. But sending a file to a specific person is a point-to-point operation. It always was. The cloud made it feel like something more complicated.

File Yeet is our answer to that complexity. Install it, find your friend, pick a file, send it. Your device talks directly to theirs, encrypted the whole way. No upload bar. No storage cap. No subscription required.

That's not a novel idea. It's just the right one.