Mark Pesce’s old homepage feels less like a personal profile and more like somebody left a door open into the moment when the Web still thought it might become a navigable three-dimensional universe.
Visit Mark Pesce’s original homepage
The page identifies itself plainly as Pesce’s original homepage, created in 1994. That date matters because several of the documents gathered there were not retrospectives about early cyberspace. They were written while people were still trying to decide what cyberspace was going to be.
Pesce was deeply involved in the early development and promotion of VRML, the Virtual Reality Modeling Language. His page preserves conference talks, essays, technical proposals, Wired pieces, experimental media, and a remarkable amount of period vocabulary: noosphere, cyberspace, virtual worlds, distributed simulations, visualization, collective space.
Some of it reads like engineering. Some reads like philosophy. Some reads like a person who had just discovered that the Internet existed and immediately decided the next obvious step was to rebuild reality inside it.
That was not an unusual mood in 1994.
Before the Web settled down
One of the most useful things on the page is Pesce’s paper from the First International Conference on the World Wide Web in Geneva in May 1994. It describes a visualization system called Labyrinth and introduces an early proposal for a language intended to describe shared three-dimensional spaces on the Web.
The site also preserves talks from SIGGRAPH, VRML conferences, Stanford, MIT, and other venues, alongside writings about cryptography, games, mapping the Earth, networked audio, and virtual communities.
What makes the page valuable in 2026 is not that every prediction came true. Quite a few did not.
The value is seeing what technically serious people thought might happen before browsers, social networks, search engines, smartphones, and app stores narrowed the public imagination of what “online” meant.
Pesce’s page belongs to an Internet where HTML was still small enough that a person could reasonably imagine adding another language and turning the whole network into a shared 3D environment.
CacheRat’s 1,967 Ancient Web Domains research list includes pages like this because old predictions become more useful after enough time has passed to see what reality kept, discarded, or quietly reinvented under another name.
The Web did not become the world Pesce’s homepage imagined.
But three decades later, people are still trying to build that world.