There’s a guy on YouTube browsing the modern web on a CRT television using a WebTV set-top box from 1997. The page loads are glacial. The rendering is hilariously wrong—CSS3 was not designed for a device with 560 pixels of horizontal resolution. And yet I watched the entire 40-minute video with a stupid grin on my face.

I shouldn’t be surprised. I spent my early career working with the constraints of early web technology. There’s something deeply satisfying about watching people refuse to let old hardware die.

A brief history of WebTV #

For those who missed it (or weren’t born yet—that stings a little), WebTV launched in 1996 as a way to bring the internet to people who didn’t own computers. The idea was straightforward: a set-top box that connected to your TV and gave you web browsing, email, and later, chat. Microsoft acquired the company in 1997 for $425 million, which felt like a fortune at the time and turned out to be a rounding error by Microsoft acquisition standards.

The service was rebranded to MSN TV, limped along for about 15 years, and was officially discontinued in 2013. Microsoft shut down the servers. The set-top boxes became paperweights. End of story.

Except it wasn’t.

The resurrection #

A small but determined community of enthusiasts decided that “the servers are off” was a problem to solve, not a final verdict. What they’ve accomplished over the past few years is genuinely impressive from an engineering standpoint.

The core challenge: WebTV devices were designed to connect to Microsoft’s proprietary servers. The authentication, content delivery, and service architecture were all closed-source and server-dependent. When the servers went down, the devices couldn’t even complete their boot sequence. They were bricked by design—not out of malice, but because nobody in 1996 imagined someone would want to use these things in 2022.

The community reverse-engineered the connection protocols. They built custom proxy servers that intercept the WebTV device’s outbound connections and route them to modern services. The web browsing works through FrogFind, a plain-HTML search proxy that wraps DuckDuckGo results in a format that vintage browsers can actually render. It strips out JavaScript, simplifies CSS, and serves content that a 1996 browser engine can handle.

The results aren’t pretty by modern standards, but they work. You can search the web, read Wikipedia articles, and browse basic sites—all on hardware that predates Google’s founding.

Beyond browsing #

What impresses me most is how far beyond basic browsing the community has gone.

They revived instant messaging. The Escargot project, originally built to resurrect MSN Messenger, was adapted to work with WebTV’s messaging client. So you can now send instant messages from a 25-year-old set-top box to other Escargot users. Is this practical? Absolutely not. Is it wonderful? Yes.

IRC works too. Someone connected WebTV’s terminal to Libera Chat, which means these ancient devices are now on the same IRC network as Linux kernel developers and open-source maintainers. I find this delightful for reasons I can’t fully articulate.

The community also maintains a wiki that documents the hardware, software, and protocols of the original WebTV ecosystem. They’re archiving service content to the Internet Archive—screenshots, page layouts, the original WebTV portal content. It’s preservation work as much as it is a hobby.

And they coordinate all of this on Discord, which is its own kind of irony: using a modern real-time communication platform to organize the revival of a 1990s internet experience.

Why this matters (beyond nostalgia) #

I could frame this as purely nostalgic—grown adults playing with old gadgets. And there’s certainly an element of that. But I think something more interesting is happening.

When Microsoft shut down WebTV’s servers, they made a unilateral decision that rendered millions of devices useless. The hardware was fine. The devices still powered on. But without server access, they were dead. This is the same pattern we see with modern IoT devices, smart home products, and cloud-dependent everything. When the company decides the service isn’t worth maintaining, your hardware becomes e-waste.

The WebTV community is demonstrating that this doesn’t have to be permanent. Proprietary protocols can be reverse-engineered. Server dependencies can be replicated. “Dead” platforms can be resurrected when communities care enough to do the work.

That’s not just nostalgia. That’s a blueprint.

The engineering charm #

What draws me to this story (beyond my obvious weakness for retro tech) is the constraint-driven engineering. These hobbyists aren’t working with modern hardware and unlimited bandwidth. They’re writing proxy servers that need to translate modern web content into something a device with 2MB of RAM and a 112MHz MIPS processor can handle. Every design decision involves trade-offs that most web developers haven’t thought about since the early 2000s.

FrogFind is a great example. It’s essentially a middleware layer that takes the modern web—with its React SPAs, 3MB hero images, and fifteen JavaScript frameworks—and reduces it to plain HTML that loads on WebTV. The technical problem is interesting, but the philosophical statement is what gets me. The web was supposed to be for everyone; FrogFind makes that slightly more true, even for hardware from 1996.

The community factor #

I work in developer relations. My day job involves building and supporting developer communities around modern platforms. The WebTV revival community has none of the corporate support, SDK documentation, or developer evangelism that platforms like Google’s invest in. What they have is passion, patience, and the IRC logs from 1998 that someone saved on a hard drive.

The community is small—maybe a few hundred active members. They find each other through YouTube videos, Reddit threads, and word of mouth. There’s no monetization strategy. No VC pitch deck. No growth metrics. Just people who think it’s cool that a box from 1997 can still connect to the internet, and who are willing to put in the hours to make that happen.

There’s a purity to that which I find refreshing. Not everything needs a business model.

The broader preservation question #

WebTV is one example. There are similar communities keeping alive AOL Instant Messenger (through the Phoenix project), Winamp skins, Flash games, and countless other digital artifacts that their creators abandoned. The Internet Archive does heroic work at scale, but these smaller community efforts fill gaps that institutional preservation misses.

I think we’ll look back at this era and be grateful for the hobbyists who decided that “deprecated” wasn’t the same as “worthless.” The WebTV revival won’t change the world. But it keeps a small piece of tech history alive, and that counts for something.

If you’ve got a WebTV box in your attic, dust it off. There’s a community waiting to help you connect it.