The Forgotten Fork: On the Deliberate Divergence of a Shared Path
In the early, sprawling days of the web, the idea of a single, canonical source for information was more of a philosophical ideal than a technical mandate. Before search engines could effectively consolidate signals, the web’s structure was a tapestry woven from countless individual decisions. One of the most fascinating, and now largely forgotten, examples of this is the deliberate divergence of the original NCSA Mosaic and Netscape Navigator documentation.
Both browsers, born from the same University of Illinois project, shared a vast common ancestry. Their early documentation existed on servers at `ftp.ncsa.uiuc.edu` and later `home.netscape.com`, covering nearly identical technical ground. For a brief period, a webmaster or curious user could find the same technical specification, the same explanation of the `img` tag, at two different addresses. They were parallel paths leading to the same destination, a living archive of their shared codebase.
But this symmetry was not to last. As the ‘browser wars’ intensified, the paths began to fork. The maintainers at Netscape, building a commercial product, made a deliberate and strategic decision. They began to prune their documentation tree, not just adding new, proprietary features, but subtly altering the language and examples on existing pages to reflect their specific implementation. The URL paths, which once pointed to identical content, now began to host subtly different texts. It was a quiet, structural declaration of independence.
This was an early, manual form of what we’d now handle with a `rel="canonical"` link or a 301 redirect. Yet, they chose the opposite. They did not consolidate their signal to a single source of truth. Instead, they cultivated divergence. They allowed the duplicate paths to exist, knowing they would slowly drift apart in content and purpose. The URLs themselves became semantic markers, signifying not just a topic, but an allegiance.
Today, we are conditioned to see duplicate content as a problem to be solved, a technical debt to be reconciled. The story of these early documentation trees offers a different perspective. Their forking was a purposeful architectural choice, a way to use site structure and URL space to narrate a schism. It was a declaration that the same information, when viewed through a different lens—or from a different domain—was no longer the same. The path you chose didn’t just lead to the information; it became part of the information itself, a permanent record of a deliberate and historic divergence.
Notes & further reading
A few pages I came back to while writing this: