Two months ago, I was sitting on a bus for 10 hours with nothing but my laptop and what seemed like a stupid idea. Today, Glenn Explore has been played by over 90,000 people, got shouted out by @levelsio and @threejs, and was featured in DigitalOcean's weekly tech talks.

This is the story of how that bus ride changed everything and reminded me why I fell in love with coding in the first place.

Chapter 1: The Bus Ride That Started It All

I'm stuck on this endless bus ride with nothing but time and a laptop. The idea hits me: “What if people could just drive around in a 3D world together? Like, actually explore the real world but in a game?”

Instead of sketching wireframes or planning architecture like a "proper" developer, I just opened VS Code and started hacking. ThreeJS for the 3D magic, Mapbox for real-world terrain data, and whatever else I could slap together to make cars move around a virtual Earth.

By the time I got off that bus, I had something that barely worked but felt... alive. You could drive a little car around, the terrain was real, and it was janky as hell. Perfect.

"The kid who'd build RuneScape bots in a single night was back. I had forgotten how good it felt to just... ship."

Chapter 2: The First Viral Moment

I threw together a quick demo video and posted it on X, expecting my usual 3 likes from my mom and maybe a colleague. Instead, I woke up to absolute chaos.

30,000 impressions. 400 likes. People actually trying the game.

But here's where it got real—people started asking “Where is everyone else? Can we drive together?” I'm watching the player count climb and thinking, “Holy shit, they want multiplayer and they want it NOW.”

So I did what any sane developer would do: I frantically Googled “real-time multiplayer” and found PartyKit. Twenty minutes later—chat system, player dots on the map, basic multiplayer. Done. Shipped. The day ended with 2,300 players and 100k total views.

I was losing my mind in the best possible way.

Chapter 3: Feeding the Beast

The feedback started pouring in. People wanted satellite maps, elevation data, better UI, free camera mode (they called it “car jail” when you could only follow the car). So I built it all.

Three days of non-stop coding later, I dropped another post. Refined UI, satellite maps with actual elevation, slick intro sequence, free camera controls, day/night toggle, time tracking. The response was even bigger.

4,000 players. 300k total impressions.

Then something magical happened. @levelsio—the guy I'd been following for years, watching him build empire after empire—commented on my video. I literally screenshot it and sent it to everyone I knew. This wasn't just viral luck anymore. People really liked what I was building.

When the official ThreeJS account jumped in too? I knew this was real.

Chapter 4: The $4 VPS That Could

Here's the part that still blows my mind. The entire thing was running on a $4 DigitalOcean VPS. Not some fancy cloud architecture with auto-scaling and microservices. Just:

• • .NET backend with SignalR for real-time multiplayer
• • SQLite database (yes, SQLite for 90k users)
• • ThreeJS + Mapbox frontend
• • $4/month VPS

Peak traffic: 90 players online simultaneously. Peak CPU usage: never hit 40%. The whole time I'm thinking about all those meetings I've sat through about “scalable architecture” and “cloud-native this” and “microservices that.”

Sometimes simple just works.

“DigitalOcean featured it in their weekly tech talks. A $4 VPS handling viral traffic while enterprise teams debate auto-scaling strategies.”

Chapter 5: Growing Pains and Real Problems

Going viral isn't all celebration and dopamine hits. When Glenn crossed 50k unique visitors, I had to make a hard choice. Server costs were climbing faster than my salary could handle, so I put up a paywall.

It felt terrible. Here I am preaching “just ship it” and suddenly I'm gatekeeping my own creation. But here's what happened next: 300 amazing people actually paid. They believed in what I was building enough to support it.

And then darker problems emerged. The chat became toxic. Racist comments, abuse, trolling. I tried moderating manually, but I'm not online 24/7. So I did what any pragmatic developer would do: I deployed an AI-powered anti-racist filter using Gemini.

300ms delay per message. Problem solved. The community stayed awesome, the vibes stayed good. Sometimes the solution is just... shipping a solution.

Chapter 6: The Viral Feedback Loop

The most insane moment came when someone posted a video about Glenn on Facebook. I woke up to 2,000 new players and 60 people online simultaneously. The video had over 300k views.

People were sharing stories in chat, getting to know each other, planning virtual road trips together. I realized I hadn't just built a game—I'd accidentally created a community.

So I leaned into it. Added a quest system to onboard new players properly. Future quests to explore the 7 Wonders, different types of road trips. Integrated MeshyAI so players could generate their own 3D models and add them to the world.

I even vibe-coded an email campaign tool in under an hour. Is it perfect? Hell no. Does it work? Absolutely.

Chapter 7: The Return to Free-to-Play

Thanks to those 300 paying legends, I was able to optimize costs and recently made Glenn free-to-play again. Launch day 2.0 was beautiful: 137 new users and 71,128km driven in a single day.

The community came flooding back, stronger than ever. And I realized something: this wasn't about the technology or the architecture or even the viral moments. It was about giving people a space to explore and connect.

What I Rediscovered About Building

Glenn brought back something I had lost in years of enterprise development: the pure joy of just building shit and seeing what happens.

At 9, I was running RuneScape private servers. At 10, coding bots. At 17, managing a bot-net empire with 300 bots across 10 machines. I didn't care if the code was garbage—I'd ship it, break it, fix it, ship it again.

Then corporate life happened. “Best practices.” “Code reviews.” “Scalable architecture.” I became paralyzed by perfection, tweaking endlessly, scared my code wasn't “good enough.”

Glenn reminded me: done is better than perfect. Shipped is better than polished. Community is better than code quality.

The Numbers That Matter (And Don't)

Let me share the final stats, not to brag, but to prove a point about what's actually important:

• • 90,000+ users over 2 months
• • 400k+ impressions across social media
• • Peak: 90 players online simultaneously
• • 300 paying supporters during paywall period
• • Featured by DigitalOcean, @levelsio, @threejs
• • $4/month hosting cost for most of the journey
• • 10 days from first line of code to viral

But here's what actually mattered: People had fun. They made friends. They explored the world together. They told me Glenn brought them joy during tough times.

“90% of the time, you don't need fancy infrastructure. You need to ship features that make people happy.”

The Balancing Act

Full transparency: I've got a full-time job, an amazing kid, and the best girlfriend in the world. When Glenn exploded, my head was spinning with ideas every second, but focus became impossible.

I had to learn to balance the entrepreneurial fire with real life. New strategy: code every other night, completely disconnect on off-nights. No refreshing X every 3 minutes. It's working.

What This Means for You

If you're reading this and have been sitting on an idea, tweaking it endlessly, waiting for it to be “ready”—stop. Ship it. See what happens.

Glenn started as a stupid idea on a bus ride. It became something that brought joy to 90,000 people. Your idea might be next.

• • Simple stacks scale further than you think
• • Community matters more than code quality
• • Viral is unpredictable—just keep shipping
• • Real-time features create magic moments
• • $4 VPS > $400 cloud bill for most projects
• • Done is better than perfect, always