The Night Our HOA Voted on the Internet: Inside One Suburb's Mesh Network Experiment

The pitch happened between the debate over fence staining and the vote on the new pool hours, which is maybe the most American setting imaginable for a minor act of infrastructure rebellion.

Ray Delgado, a retired telecom engineer who lives two doors down from the clubhouse, waited for the lull in the agenda, stood up, and asked the fifty-odd homeowners of Willow Creek Estates — a made-up name for a very real subdivision outside Tampa — a simple question: "Why are we all paying $80 a month to six different companies for the same internet?"

What followed was an eighteen-month experiment in suburban self-reliance that worked about as well as you'd expect: part triumph, part cautionary tale, and entirely fascinating if you've ever wondered what happens when regular folks try to build their own network.

The Pitch: Confusion First, Curiosity Later

Ray's proposal was a community mesh network — a web of rooftop antennas and routers that pass signal from house to house, with a handful of shared fiber backhaul connections doing the heavy lifting. Neighbors who joined would pay a fraction of a commercial plan, split across the group.

The first reaction wasn't hostility. It was incomprehension.

"Half the room thought he was proposing a Wi-Fi hotspot you could steal," laughs Marisol, Ray's neighbor and the experiment's most vocal early skeptic. "The other half thought it was illegal."

Ray had come prepared. He'd printed one-page explainers, mapped the neighborhood's sightlines on a drone photo, and — crucially — brought a working demo: two antennas on the clubhouse roof already streaming a YouTube video of a man power-washing his driveway. The room was quietly mesmerized.

The HOA board didn't approve the network that night. They approved a committee, which in HOA terms is roughly the same thing as a blessing with a built-in escape hatch.

The Build: Roof Ladders and Easement Politics

The technical side went smoother than the social side, which surprised everyone.

Ray's design was pragmatic: three shared fiber connections (business-class, about $150/month each) at volunteer "anchor" houses, feeding a mesh of roughly $90 outdoor radios mounted on chimneys and second-story eaves. Line of sight mattered enormously. The house at the end of Magnolia Court, tucked behind two oak trees and a garage, became the project's white whale until a neighbor agreed to host a relay unit.

The real drama was physical and legal-adjacent. One homeowner's roofing company refused to work around "unauthorized equipment." Another neighbor discovered the antennas violated a deed restriction about "exterior apparatus" that had apparently been written with satellite dishes in mind. The board quietly rewrote the language after a lawyer pointed out that FCC rules generally protect antenna installations anyway.

By month six, twenty-three of the subdivision's eighty-four homes were on the mesh.

What Worked (Better Than Anyone Expected)

The economics were undeniable. Monthly costs per household settled around $25 — including equipment amortization — versus $70–$90 for comparable commercial plans. During hurricane season, when the commercial ISPs blinked, the mesh's battery-backed anchor nodes kept several streets online. Ray's group chat of amateur tinkerers debugged problems faster than any call center could schedule a truck roll.

There was also an unexpected social dividend. People who'd lived four doors apart for a decade finally met, usually while standing in each other's yards holding a ladder.

What Collapsed

The failure modes were as predictable as they were instructive.

First: the single point of human failure. Ray got shingles in October. Nobody else fully understood the network's configuration, and a firmware update gone wrong took down half the mesh for a long weekend. The committee's answer — a second volunteer, a written runbook, and regular "network nights" — fixed it, but only after the outage made the group chat genuinely unpleasant.

Second: the freeloaders. A few households wanted access without hosting equipment or paying into the fund. The committee instituted a simple rule — no antenna, no service, no exceptions — and lost exactly zero people who mattered.

Third: scale. Getting from twenty-three homes to forty proved nearly impossible. New joiners required individual engineering: signal tests, mounting permissions, cabling runs. The network plateaued not because it was broken, but because every additional house was a project rather than a signup.

What It Says About Us

The Willow Creek experiment is still running, stable at around thirty homes, and its mixed success might be the most honest outcome possible. It turns out the hard part of community infrastructure was never the radios — it's the meeting minutes, the deed restrictions, the neighbor who won't return calls, and the question of who's responsible when it breaks.

But it also revealed something quietly encouraging. Americans don't actually hate shared infrastructure. We hate feeling helpless inside it. Give people a ladder, a rooftop, and a stake in the thing, and a surprising number will show up on a Saturday morning with a drill and a spool of outdoor-rated cable.

Ray puts it more simply. "The internet was never the hard part," he told me, standing in his garage surrounded by labeled cable runs that would make a datacenter tech weep with envy. "The hard part is that nobody's ever asked the neighbors."

Somebody asked. Thirty houses said yes. That's a start.