Category: Ancient Web

The most ancient times of the internet: 90s, early 2000s.

Posted on by

Ancient Web: Ed Collins Built a Chess-Software Help Site and Kept Using It for Decades

The most useful old websites are often the ones built because somebody got tired of answering the same question twice.

Visit Ed Collins’s website

Ed Collins has been using chess database software in the Scid family since the late 1990s. He says that back then he built a website to help other users understand it, eventually calling the project A Tribute to Scid vs. PC.

The remarkable part is not that he made a help site in 1998.

It is that, nearly three decades later, he was still pointing people toward it.

In a 2025 Chess.com discussion, Collins said he still used Scid vs. PC and linked users back to the old material he had created for newcomers.

That is personal-Web durability in its purest form.

Documentation that outlived several generations of software culture

Chess software has never been short on complexity. Databases, engines, opening trees, game collections, analysis tools, importing, exporting, and search functions can turn a simple board game into a small computing discipline.

A good volunteer help page therefore becomes more useful than its author probably expected.

Collins’s site fits the pattern of the classic specialist homepage: one person learns a tool well enough to explain it, writes down the process, adds examples, and leaves the material available at a stable address.

There is no requirement that it become a company.

There is no need to convert the instructions into a video channel, Discord server, subscription course, or SEO-driven support center.

The page itself is the product of expertise.

That model matters because old software communities are full of missing knowledge. Programs survive while forum threads disappear. Download mirrors remain while installation notes vanish. A tool may still run perfectly, but the one person who knew why a particular setting mattered took down his homepage fifteen years ago.

Collins did the opposite.

He kept the homepage.

CacheRat’s 1,967 Ancient Web Domains research list repeatedly turns up these little islands of durable documentation. They are easy to underestimate because they do not look like archives.

They look like somebody’s hobby page.

Sometimes that is exactly why they survived.

Browse Ed Collins’s personal site and chess material

Posted on by

Ancient Web: LatinAmericanStudies.org Became a Professor’s 35,000-Link Research Machine

There was a period when a professor could decide that an entire field needed a better Internet resource and then just start building one by hand.

Visit LatinAmericanStudies.org

Historian Antonio de la Cova created LatinAmericanStudies.org in 1997 as an online research resource for Latin American and Latino studies.

By 2004, an Indiana University article reported that the site had grown to more than 35,000 links and nearly six gigabytes of material. That was not a university library project with a development team. It was a professor-maintained website.

The difference shows.

The site is organized less like a modern academic portal and more like an enormous research desk that kept expanding whenever another useful source appeared.

The Web as a working archive

The material ranges across Latin American history, Cuba, U.S.-Cuba relations, political movements, immigration, military history, historical figures, scanned documents, newspaper articles, photographs, academic papers, government records, and oral-history material.

One particularly valuable section preserves Cuban oral-history interviews from 1974 through 2004, including audio and transcripts. Other pages assemble U.S.-Cuba documents decade by decade or preserve historical images alongside contemporary records.

The site is not neutral in the sense of being an anonymous institutional database. De la Cova’s own biography, Cuban exile experience, research interests, interviews, and political perspective are plainly part of the project.

That actually makes it a better example of the old academic Web.

Personal scholarly sites often exposed the researcher as well as the research. You could see what one historian considered worth saving, which documents he linked together, what he digitized, and which events he treated as central.

Modern academic platforms tend to separate publications, institutional profiles, archives, course pages, and personal history into different systems. LatinAmericanStudies.org piles them into one giant navigable body of material.

That design can be awkward. It is also extraordinarily revealing.

CacheRat’s 1,967 Ancient Web Domains research list contains plenty of specialist sites where one obsessive maintainer did the work that a later institution might call a digital-humanities initiative.

This one actually became a major reference resource while still looking like somebody’s enormous hand-built website.

In 1997, that was not a contradiction.

It was the model.

Dig through LatinAmericanStudies.org

Posted on by

Ancient Web: Robots and Computers Built a Robot Reference Before Robotics Went Mainstream

Before consumer drones, cheap microcontrollers, and household robots made the subject ordinary, hobby robotics had the atmosphere of a workshop with the door half open.

Visit the Robots and Computers archive

The surviving robot pages at Robots and Computers come from that period. The site organizes robot material as a practical reference: machines, projects, links, components, and examples gathered for people who were interested enough to build or study the things themselves.

That sounds normal now. It was less normal when robotics meant scavenged motors, parallel ports, BASIC programs, surplus sensors, RC hardware, and whatever documentation somebody could find online.

The important part is not that every page is timeless. Quite the opposite. Old robotics pages preserve the assumptions of their moment.

A modern project might begin with a Raspberry Pi, an ESP32, a camera module, and libraries downloaded in seconds. An older project often began with the question of whether the computer could physically talk to the device at all.

The Web was part of the parts bin

Early technical hobby sites did something magazines could not do well: they linked sideways.

A robot page could send you to a motor-control circuit, a manufacturer, another builder’s machine, a university lab, a software download, or a personal page describing a failed experiment. The resource was not one article. It was the web of people trying things.

That structure mattered to robotics because the field crosses mechanical design, electronics, programming, sensing, control systems, and fabrication. A hobbyist rarely had all of those skills at once.

The Web filled gaps.

Sites like Robots and Computers became informal switching stations between disciplines. Their value today is partly historical: they show what information people considered scarce enough to collect and categorize.

The old pages are also a reminder that technical culture used to leave more of its intermediate steps visible. Modern social posts tend to show the finished robot. Old personal sites were more likely to leave behind wiring diagrams, parts lists, half-working prototypes, and blunt notes about what failed.

CacheRat’s 1,967 Ancient Web Domains research list is full of these small technical repositories because they were one of the Web’s best early uses.

Not entertainment.

Not branding.

Just somebody figuring something out and leaving the bench notes online.

Explore the surviving robot pages

Posted on by

Ancient Web: A Candy Rose Built a True-Crime Document Archive Before Social Media Ate the Case

Long before true crime became an industrial supply chain of podcasts, reaction videos, and thumbnails with red arrows, some people were doing the much less glamorous work of collecting the damn documents.

Visit A Candy Rose

A Candy Rose became one of the best-known independent archives surrounding the JonBenet Ramsey case. The site accumulated timelines, interview transcripts, photographs, legal documents, news material, discussion references, and links to primary or near-primary sources.

It also expanded into other criminal cases, with detailed timelines and document collections for subjects beyond Ramsey.

The site is messy by modern design standards. That is almost an advantage.

You are not being funneled toward a conclusion by a recommendation engine. You are being handed piles of material.

The archive became part of the case’s Internet history

Coverage of the early online Ramsey community describes a sprawling late-1990s ecosystem of forums, chat rooms, opinion camps, link pages, and specialist archives. Westword later singled out A Candy Rose as one of the most useful and extensive surviving JonBenet archives.

That reputation persisted because the site did something forums are bad at: it imposed enough structure on years of scattered material that researchers could still find specific dates, statements, photographs, and documents later.

Other writers and podcasters have cited A Candy Rose as a source for case timelines and records. Reddit communities still point newcomers toward it because much of the material is difficult to reconstruct from modern search alone.

The preservation story nearly failed. After the site’s creator died, users worried in 2023 that the domain and its research might disappear. Discussion threads described the collection as effectively irreplaceable and scrambled to determine whether it had been mirrored or preserved elsewhere. The site later returned.

That near-loss is the lesson.

A giant independent research archive can look permanent right up until one renewal notice is missed.

CacheRat’s 1,967 Ancient Web Domains research list repeatedly runs into this problem. The old Web contains enormous amounts of volunteer archival labor that was never backed by an institution with a records department and a preservation budget.

A Candy Rose is controversial subject matter because true-crime communities naturally disagree about interpretation. The historical value of the site is easier to state: it preserved source material and chronology at a scale few personal websites ever attempted.

That is useful whether or not you agree with any particular theory attached to the case.

Browse the surviving A Candy Rose archive

Posted on by

Ancient Web: Russ Nelson’s Homepage Connects Packet Drivers, Open Source, and the First Quaker Site

Some personal homepages are diaries. Russ Nelson’s is more like a wiring diagram for several pieces of Internet history.

Visit Russ Nelson’s homepage

Nelson is a programmer associated with Crynwr Software, DOS packet drivers, free software, and the early Open Source Initiative. He was a founding OSI board member and later served briefly as its president.

Long before “open source” became corporate conference wallpaper, Nelson was already working in the practical layer underneath networked computing: drivers, licensing, interoperability, and the unglamorous machinery that makes computers talk to each other.

He became especially well known for packet drivers used by DOS networking software. Historical FreeDOS documentation even calls him the “Packet Driver King.” In a 1999 mailing-list message, Nelson described spotting one of the Crynwr packet drivers booting on airport departure-display computers in Frankfurt.

That is a very specific kind of nerd satisfaction: recognizing your networking code from across an airport terminal because somebody else’s machine rebooted badly enough to expose it.

The personal homepage as technical memory

Nelson’s old web presence also overlaps with the history of open-source licensing. Mailing-list archives from 2000 and 2001 show him using his site to circulate proposed licenses and ask whether they met open-source standards. His domain appears throughout early license-discussion traffic as both a personal identity and a working technical resource.

There is another strange historical branch.

Nelson created quaker.org in early 1995, described in biographical sources as the first Quaker website. He eventually transferred it to a Quaker publishing nonprofit in 2018.

That mix is exactly why old personal domains are so useful. Modern biographies separate a person’s activities into neat boxes: networking, business, religion, politics, software. A personal homepage often leaves those categories piled together because that is how a real person’s life actually works.

The value is not merely nostalgia for old HTML. The domain serves as a persistent address attached to decades of technical discussion. Search old mailing lists and the same URL keeps appearing underneath messages about packet drivers, qmail, licensing, and free software.

CacheRat’s 1,967 Ancient Web Domains research list is full of sites where the author’s homepage quietly became part of the historical record.

Russ Nelson’s is one of the clearer examples.

The homepage did not just describe his work.

For years, it was one of the places the work happened.

Browse Russ Nelson’s surviving personal site

Posted on by

Ancient Web: The Mermalair Became a SpongeBob Episode-Capsule Time Capsule

Before every television fandom had a giant wiki with thousands of editors, sometimes one person simply decided to document the entire damn show.

Browse The Mermalair’s surviving files

The Mermalair was one of the better examples. The SpongeBob SquarePants fan site built detailed episode capsules, character biographies, appearance lists, show information, and miscellaneous reference pages during the show’s early years.

It became substantial enough that later SpongeBob fan-history pages describe it as one of the largest detailed episode-capsule resources in the fandom. When its earlier hosting situation disappeared, SpongeBuddy Mania began preserving the site in 2006.

That preservation matters because the pages capture a very specific form of fan labor.

A fan site as reference book

The Mermalair did not merely say that SpongeBob was good.

Its episode pages recorded production numbers, original air dates, writers, storyboard artists, animation directors, DVD appearances, VHS appearances, short summaries, longer plot descriptions, and observations about individual episodes.

The surviving episode index reports 188 capsules.

Character pages did the same thing for SpongeBob, Patrick, Squidward, and the rest of the cast: physical descriptions, occupations, interests, family relationships, friends, first appearances, voice actors, and long lists of episode appearances.

This kind of site was effectively a volunteer database implemented as ordinary HTML.

No MediaWiki installation was required. No fandom platform owned the structure. No account system was necessary for the author to publish an update. The webmaster just added another page and another link.

The result is less standardized than a modern wiki and more personal. You can see the editorial choices directly: what deserved a capsule, how much detail mattered, which categories were useful, and which parts of the show the maintainer cared enough to document.

That is why old fan sites remain valuable even when modern reference sites contain more information. The modern site tells you about the television program. The old site also tells you how fans used the Web around the television program.

CacheRat’s 1,967 Ancient Web Domains research list contains many remnants of that decentralized fan-Web period.

The Mermalair survived because another fan community cared enough to host it.

That is about as old-Web as preservation gets.

Dig through The Mermalair’s preserved fan archive

Posted on by

Ancient Web: Oldskool’s Lode Runner Shrine Kept a 1983 Puzzle Game Playable

Some games survive because publishers keep reissuing them. Others survive because a few stubborn players refuse to let the knowledge disappear.

Visit the Lode Runner shrine at Oldskool.org

The Lode Runner material at Oldskool belongs to the second category. Broderbund’s 1983 puzzle-platform game was simple enough to explain in one sentence and deep enough to generate decades of obsession: run, dig, trap enemies, collect gold, escape.

That simplicity made it portable. It also made it a perfect candidate for the kind of fan preservation the early Web was unusually good at.

The Oldskool shrine collected information about versions, levels, editors, ports, and utilities around the game instead of treating it as a dead commercial product. That matters because early computer games rarely had one canonical form. Different machines had different graphics, timings, controls, level sets, and quirks. A page documenting those differences could become more useful than the box manual.

A game became a small research subject

Lode Runner is especially suited to this treatment because level design is the game.

A player does not simply remember the characters or soundtrack. They remember routes, ladders, fake floors, enemy behavior, and the exact moment a dug brick grows back under somebody’s feet.

Fan pages could therefore preserve not only nostalgia but technique.

That is one of the recurring patterns in old game sites: enthusiasts accidentally became archivists. They saved executable files, level data, screenshots, utilities, and practical knowledge because they wanted to keep playing. Years later, those same scraps become evidence for how software actually behaved on historical hardware.

Oldskool.org itself is part of that culture. The name says exactly what the project values: older software, older machines, and the communities that formed around them before retrocomputing became a polished commercial category.

The result is much more useful than a modern page saying, “Lode Runner was an influential 1983 game.”

That sentence tells you almost nothing.

A fan archive that shows the variants, tools, and weird details tells you what the game was like to live with.

CacheRat’s 1,967 Ancient Web Domains research list keeps finding pages like this because the old Web blurred the line between fandom and preservation.

Nobody appointed the fans historians.

They just kept the files.

Browse the surviving Lode Runner shrine

Posted on by

Ancient Web: CCRU Turned a 1999 Art Event Into a Web-Based Time Cult

Some old websites explain themselves. CCRU’s Syzygy pages behave more like they are trying to recruit your nervous system.

Enter CCRU’s Syzygy

The Cybernetic Culture Research Unit emerged around the University of Warwick in the 1990s and mixed philosophy, electronic music, science fiction, cyberculture, occult language, and what it called hyperstition: fictions treated as things capable of helping manufacture the realities they describe.

Syzygy was one of the places where that method escaped the seminar room.

From February 26 through March 28, 1999, CCRU and the art collective 0rphan Drift staged Syzygy: A Digital Unlife Event at Beaconsfield in London. The project responded to millennium anxiety and Y2K with an invented calendar structured around five avatars, audiovisual performances, installations, workshops, electronic music, and a deliberately fictional mythology.

The website is part documentation, part extension of the event, and part trapdoor.

The page refuses to stay in one category

The Syzygy page defines its terms in a vocabulary that moves freely between astronomy, anatomy, mathematics, Gnosticism, network engineering, fiction, and invented “Lemurian Time-Sorcery.” Elsewhere the CCRU archive points toward texts about Y2K panic, strange economies, cyberculture, invented professors, Lovecraftian organizations, and a numerical system called the Numogram.

It is easy to read the whole thing and ask the reasonable question: what the hell is this?

That question is basically part of the design.

Contemporary records make clear that there was a real 1999 art event underneath the fiction. Beaconsfield later described Syzygy as a series of five performance weekends combining projection, dance, audio, installations, and an alternative calendrical system. 0rphan Drift’s surviving archive documents the same collaboration and its Y2K framing.

The Web mattered because it allowed the invented universe to continue beyond the physical gallery. A visitor could encounter the terminology without the event, follow links into CCRU’s archive, and lose the boundary between documentation and mythology.

That is a very 1990s use of hypertext: not merely distributing information, but constructing atmosphere through interconnected fragments.

CacheRat’s 1,967 Ancient Web Domains research list includes a lot of pages where the Web itself was still treated as an experimental medium. CCRU went considerably further than most.

They did not build a website about their weird theory-fiction project.

They made the website part of the weird theory-fiction project.

Descend into the surviving Syzygy pages

Posted on by

Ancient Web: Fluffy’s World Was Cool Site of the Day Before That Meant Anything

Before the Web had influencers, feeds, or recommendation engines, it had people wandering around and telling each other, you need to see this weird page.

Visit Fluffy’s World

Fluffy’s World is a surviving example from that period. The site dates to 1994, was updated through 1995, and mixed illustrations, stories, essays, and odd little creative experiments under the banner of Wayward Fluffy Publications.

On January 27, 1995, it was selected as Cool Site of the Day.

That phrase sounds almost meaningless now because the modern Web manufactures “featured content” by the million. In 1995, being pointed to from one of the Web’s early discovery lists could be an event. The network was small enough that somebody could genuinely stumble across a strange personal creation and send a noticeable chunk of the online population toward it.

A website could simply be a place

Fluffy’s World is useful because it does not behave like a modern publication.

There is no obvious funnel. No newsletter pop-up arrives to ask whether you would like ten percent off a PDF. No algorithmic feed attempts to determine which illustration will maximize session time.

It is closer to a handmade room.

The site’s surviving reputation describes it as a collection of art and stories, and that description is refreshingly complete. Somebody made things, put them on a server, arranged links between them, and let visitors roam around.

That was enough.

The technical limitations helped define the experience. Pages were light. Images were small because they had to be. Navigation was explicit rather than predictive. If you followed a link, it was because somebody deliberately placed it there.

What makes Fluffy’s World worth preserving is not that every drawing or story became historically important. Most personal-web material never needed to become important. Its historical value is that it shows what ordinary creativity looked like when publishing on the Web was still novel.

CacheRat’s 1,967 Ancient Web Domains research list contains a lot of sites with that same quality: not optimized, not scalable, and therefore far more revealing about the person who made them.

The old Web did not require a brand strategy.

Sometimes you just needed a domain, a few HTML files, and enough nerve to call the place Fluffy’s World.

Explore the surviving Fluffy’s World pages

Posted on by

Ancient Web: LGR Kept the Personal Homepage Alive Beside YouTube

Not every site in the Ancient Web pile was born in 1997. Some are interesting because they refused to stop looking like personal websites after the rest of the Internet moved on.

LazyGameReviews.com is one of those.

Clint Basinger’s LGR work began in the YouTube era and became known for detailed videos about DOS games, old PCs, weird peripherals, The Sims, thrift-store finds, and forgotten hardware. The video audience grew enormously.

The website, meanwhile, remained closer to an “awesomepage.”

That contrast is the story.

The platform was distribution; the domain was home

By 2015, people examining the site’s source were already describing its design as unusually classic-looking. One contemporary discussion noted that the navigation used an image-map style layout and that even the blinking cursor effect appeared baked into the graphic.

That is not an accident of neglect so much as a recognizable personal-Web language.

LGR’s subject matter also fits the design perfectly. A channel obsessed with beige computers, obsolete storage, forgotten sound cards, giant PC game boxes, and hardware nobody sane saved does not require a glossy startup landing page.

The simple domain works more like a calling card and stable pointer toward the larger body of work.

There is a broader preservation lesson here. Creators who exist entirely inside a platform inherit that platform’s URLs, policies, recommendation system, interface changes, and eventual mortality. Maintaining an independent domain—even a very simple one—creates a small piece of identity outside the feed.

LGR’s website is therefore not “ancient” in the strict GeoCities sense. It is useful because it demonstrates that the personal homepage model survived well into the social-platform era when somebody actually wanted it to.

CacheRat’s 1,967 Ancient Web Domains research list includes plenty of genuine fossils. It is also worth noticing living species that never adopted the new ecosystem completely.

YouTube can host the videos.

The domain can still say: this is my place.

Visit the LGR homepage