The Internet We Have vs. The Internet We Built
In a world where web pages routinely weigh in at over a megabyte, often laden with JavaScript and trackers, the concept of a new internet might seem like a monumental, even futile, endeavor. Yet, Thomas Verhave, a developer, took that sarcastic suggestion – "let's make a new internet" – and built one. Not a framework, not a metaphor, but a functional, albeit tiny, alternative internet called weft, all written in pure Python using only the standard library.
The motivation behind weft is a direct reaction to the bloat and pervasive surveillance of the contemporary web. Verhave’s bet is on simplicity, efficiency, and user control. The result is a stark departure from the status quo: a minimal, line-based protocol over raw TCP, featuring its own address scheme, its own markup language, a terminal-based browser, and an application-layer firewall controllable from within the browser itself.
weft: Protocol, Browser, and Firewall in Pure Python
weft’s architecture is designed for extreme minimalism. The protocol itself is line-based, communicating over raw TCP connections. This stripped-down approach eliminates the overhead of established protocols like HTTP, focusing purely on data transfer. The network is formed by just three simple servers that link to each other, creating a self-contained ecosystem.
The address scheme is also custom-built, eschewing the familiar domain name system for something more direct and manageable within the weft network. Similarly, the markup language is designed to be lightweight, allowing for the display of text and basic content without the need for complex rendering engines or extensive styling. This focus on text ensures that pages load almost instantaneously and require minimal bandwidth.
Perhaps one of the most intriguing components is the terminal browser. This browser allows users to navigate the weft network directly from their command line. It’s a throwback to the early days of the internet, prioritizing functionality and speed over graphical embellishments. Crucially, this browser also serves as the control interface for weft’s application-layer firewall. This firewall operates at the application level, providing granular control over network traffic and an additional layer of privacy for users within the weft network.
The Bet on Unscalability
The most striking aspect of weft, and the core of Verhave’s “bet,” is its intentional lack of scalability for mass adoption. The entire project, comprising the protocol, browser, and firewall, amounts to just a few hundred lines of Python code. This is a deliberate choice, a philosophical statement against the relentless pursuit of growth and complexity that characterizes much of modern technology. Verhave acknowledges that weft will never scale to billions of users, and that is precisely the point. It’s an experiment in what the internet could be if its design priorities were flipped: privacy over data collection, simplicity over feature creep, and user control over corporate oversight.
This approach offers a potent counterpoint to the current web landscape. While established browsers and protocols grapple with issues of performance, security vulnerabilities, and privacy concerns stemming from their inherent complexity and reliance on vast ecosystems of third-party scripts and services, weft offers a glimpse into an alternative. It’s a demonstration that even in the age of AI and massive data centers, fundamental building blocks of digital communication can be reimagined with an emphasis on core principles rather than exponential growth.
The implications of such a project extend beyond mere curiosity. For developers, it serves as a reminder that elegant solutions can often be found in simplicity. For users weary of the modern web’s intrusive nature, it presents a conceptual model for a more private and efficient online experience, even if it remains a niche experiment for now. The question isn't whether weft will replace the current internet, but what lessons can be drawn from its creation about the trade-offs we’ve made in building the web we have today.