You know that sinking feeling when you bookmark a page and it’s gone a month later? The web is bleeding content every single day. Design research, documentation, even entire communities vanish without warning. But here’s the truth nobody talks about: the bottleneck isn’t storage – it’s something far more absurd.
For years, web archiving meant one thing: rent a fleet of servers, run crawlers, and pray your storage doesn’t explode. That model is broken. It’s expensive, slow, and designed for a web that no longer exists. Most people assume the challenge is storing billions of pages. They’re wrong. The real bottleneck in web archiving isn’t disk space. It’s orchestrating a browser session on infrastructure designed for a click.
I saw this firsthand. At Kenobi, we needed to capture full MHTML archives of design systems and component libraries for research. Every page was a dynamic, JavaScript-heavy mess. Traditional crawlers choked. We needed something that could spin up a real browser, render the page, and save a faithful snapshot – reliably and at scale. What we found was a gap no one had filled.
So I built Breamer. It’s an open-source tool that runs website capture entirely on Cloudflare Workers and Workers Containers. That means no servers to manage, no infrastructure to babysit. Just a script that launches a headless Chromium instance at the edge, captures the full MHTML (including CSS, images, fonts – everything), and stores it wherever you want. The whole thing is programmable, scalable, and costs pennies.
Here’s the twist: archiving is supposed to preserve a durable snapshot of a constantly changing web. Yet its new engine is the most ephemeral thing in tech – a stateless edge worker. We’re preserving the web using tools that don’t even last a second on their own. That’s the irony, and the genius.
Most developers think web archiving is a storage problem. It’s not. It’s a compute orchestration problem. You need to run a resource-heavy browser session on infrastructure that was built for lightweight, stateless requests. Breamer solves that by treating every capture as a disposable job – spin up, render, save, die. The ephemerality becomes a feature, not a bug.
For researchers and designers, this is a game-changer. You can now automate website capture without owning a single server. Want to archive a competitor’s design trends? Set up a cron job. Need to preserve a documentation site before it goes offline? One API call. The barrier to building archival workflows just dropped from ‘enterprise budget’ to ‘coffee money.’
This isn’t just a tool. It’s a fundamental shift in how we think about preservation. The old model was centralized, fragile, and expensive. The new model is distributed, resilient, and cheap. We’re moving from ‘we need a data center’ to ‘we need a few lines of code.’
Does this mean the Internet Archive is obsolete? No. But it does mean that every developer can now be an archivist. And that’s a future worth fighting for.
FAQ
Q: Why not just use the Internet Archive?
A: Internet Archive is great for public web pages, but it's slow, centralized, and can't handle authenticated content or private research sites. Breamer gives you on-demand, programmable captures for any URL – including internal tools, design systems, or paywalled resources.
Q: What's the practical implication for a developer?
A: You can now automate website capture with a cron job and a few lines of code. No servers, no DevOps. Just a Cloudflare Worker that spins up a browser, takes a snapshot, and saves it to R2 or S3. It's archival-as-a-service for the rest of us.
Q: Isn't edge computing too ephemeral for archiving?
A: That's the counterintuitive beauty. The ephemerality of edge workers actually makes archiving more scalable. Each capture is a disposable job – no long-running state, no memory leaks. The snapshot is the durable artifact; the compute is just a temporary contractor. Works perfectly.