I built a complete alternative
to the Internet.
Now I'm giving it away.
FrogNet is a self-forming mesh that runs full applications — web, database, AI, sensors — over WireGuard tunnels, radio, LoRa, or satellite. No cloud. No ISP. No public-Internet attack surface. It has been running across eight nodes on three continents for over a year.
I'm 67, I've spent over $500,000 of my own money building it, and I'm open-sourcing it now because AI-driven attackers are about to make every Internet-connected system a target at machine speed.
I need $10,000 to file defensive patents and form the Foundation that will hold them. Then the whole codebase opens.
Mythos is Anthropic's unreleased AI that finds and exploits software vulnerabilities autonomously, at machine speed. Anthropic deemed it too dangerous for public release and launched Project Glasswing with Amazon, Apple, Google, Microsoft, NVIDIA, and the Linux Foundation to patch vulnerabilities defensively. The threat posed by Mythos is, in our opinion, both immediate and catastrophic. Mythos, or something like it, will escape and will attack everything on the Internet, relentlessly, at machine speeds. FrogNet is the only viable way to prevent a massive loss of data and, potentially, control. FrogNet will be made open source as soon as possible.
Stay Informed
Get notified the moment the code goes public. One email when the repo opens. No list, no marketing, no spam.
The technology A complete Private Internet, running now.
The reference application is the network we have running today — an off-Internet multinational network of networks that operates at speeds sufficient for video, and that includes the transient database, the sensor platform, and the AI host.
Sovereign nodes: Each node runs its own web server, database, DNS, DHCP, and application stack. No cloud. No permission. A single node, completely disconnected, is still a functioning platform.
Any transport: WiFi, Ethernet, ham radio at 4800 baud, LoRa, WireGuard tunnels, satellite — all simultaneously. The network finds its own path.
93.8% compression, up to 40× effective throughput on structured traffic: BLDC-1 semantic compression learns the structure of real traffic and sends only what changed. 10KB becomes 50 bytes. Full web applications run over radio links too constrained for standard Internet traffic.
An invitation It works. Help me make it what it should be.
It works. It has been running across eight nodes on three continents for over a year. It is not polished, it is not documented to the standards the Foundation will eventually require, and there are architectural decisions that will benefit from peer review.
That peer review is the whole point of open-sourcing it. I built this alone. It needs the community to become what it should be. Come look at the code, challenge the assumptions, find the bugs, and help make this the basis for a better way of networking.
The code opens as soon as the Foundation is formed and the patents are filed. If you want to contribute — code, testing, architecture review, deployment — get in touch. I'll get you onboarded now so you can hit the ground running the day the repo goes live.
The threat AI attackers at machine speed change everything.
Claude Mythos finds and chains software vulnerabilities faster than any human team can patch them. Every system with a live connection to the Internet is a target. The exploits aren't new. The speed is.
Glasswing hardens the software. Nobody is fixing the topology. Every patched OS, every hardened browser, every secured endpoint is still sitting on centralized cloud infrastructure that is itself an attack surface.
I would very much like to see all private content removed from the Internet and accessed only through things like mailboxes and dead drops. No more live connections between the unknown and the sensitive.
The decision Why I'm giving away everything I've built (and what it's costing me).
I am 67, writing code since 1975. My wife and I survive on social security. I have spent more than $500,000 building FrogNet — money I did not have, borrowed against a future that was supposed to include taking this to market. Build the technology, prove it works, find investors, build a company, pay back what I owe. I don't know how I'm going to pay the mortgage next month, and I have been in that position more than once during this project.
Then Anthropic announced Claude Mythos.
An AI that finds and chains software vulnerabilities faster than any human team can patch them. Anthropic themselves deemed it too dangerous for public release. They stood up Project Glasswing — Amazon, Apple, Google, Microsoft, NVIDIA, the Linux Foundation — because the threat is that serious. The Federal Reserve convened emergency meetings with bank CEOs.
Mythos, or something like it, will get into the hands of bad actors. That is not a question of if. When it does, every system with a live connection to the Internet becomes a target at machine speed.
I cannot in good conscience keep this proprietary when the technology I've built directly addresses the architectural vulnerability that Mythos exploits. Not the software vulnerabilities — Glasswing handles those. The topology. The fact that private data sits on infrastructure that is reachable from the public Internet. FrogNet removes it.
So here I am. Broke, with something that can help, and asking for $10,000 to file four defensive patents and form the Foundation so I can give it all away.
The plan Open source under a trade foundation.
The FrogNet Foundation (501(c)(6)) holds the code and patents — defensively. Fawcett Innovations, LLC provides consulting, training, and certification. Standard model. Same as the Linux Foundation.
The code belongs to the community. The expertise is available for hire. The Foundation earns membership dues. The commercial company earns service revenue. Neither can close the code.
The barrier $10,000 stands between here and open source.
Four core inventions need defensive patents before the code can open. That costs approximately $7,500. Forming the Foundation that will hold them costs approximately $2,500. That is the barrier. $10 helps. $1,000 helps. Anything helps.
4 Provisional Patent Filings (~$7,500): Defensive filings on the four highest-priority inventions out of nine identified. The Foundation will own them. They exist to protect the community's right to use the technology, not to exclude anyone.
Foundation Formation (~$2,500): 501(c)(6) trade organization. Holds the IP, governs the code, provides the legal home for contributions. Same structure as the Linux Foundation.
The founder John Fawcett — 67, writing code since 1975.
50+ years building real-time networked systems. Boeing (Special Achievement Award), Sierra On-Line/EA/WotC multiplayer game servers, Microsoft SCCM Linux agent. Three issued U.S. patents. Over $500K personal R&D invested. Living on social security. This needs to be in the public domain regardless.
Author of FrogNet: A Living Network — a technical origin story about building systems that refuse to lie. Available as foreword + Chapter One on this site or as the full book on Amazon.