Category: Ancient Web

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

Posted on by

Ancient Web: Bass Drop’s Club Rebuilt the Fun Part of the Personal Web

Not every Ancient Web site has to be ancient. Sometimes the interesting specimen is a new site deliberately rebuilding what the old web got right.

Visit Bass Drop’s Club

Bass Drop’s Club says development started on January 8, 2024, and the bassdrop.club domain was acquired less than two weeks later. The site is modern enough to have had backend refactoring, mobile-navigation fixes, and performance work. Its priorities, however, are gloriously personal.

There is a guestbook. There are crochet pages. There is a hand-built dress-up game. There is a fishing minigame. The homepage reports what the owner is listening to, playing, reading or watching. Animated stars, tiny graphics, snack images and decorative navigation are allowed to exist without first passing through a brand consultant.

That sounds trivial until you compare it with the modern platform web.

A website as a room, not a feed

Social platforms encourage people to pour everything into the same standardized stream. A drawing, a game, a book recommendation and a personal update all become nearly identical rectangles sorted by an algorithm.

Bass Drop’s Club does the opposite. Different interests get different places. The owner can build a tiny fishing game because it belongs there. Crochet can have its own section because there is no product manager asking whether crochet improves quarterly engagement.

The update log is especially useful. It records the actual construction of the place: navigation changes, performance fixes, the addition of mobile support, new games and new sections. You can watch a personal website become more itself over time.

This is why handmade sites belong beside genuinely old domains in CacheRat’s 1,967 Ancient Web Domains research list. The valuable thing is not merely age. It is the web model: independent pages, direct authorship, strange side rooms and a structure chosen by the person publishing it.

A website used to be allowed to resemble its owner.

Bass Drop’s Club remembers that. It has modern code under the floorboards, but the front room still contains a guestbook, crochet, games and whatever song happens to be playing.

That is not nostalgia. That is a design choice worth keeping.

Wander through Bass Drop’s Club

Posted on by

Ancient Web: Homo Ludens Built a Homepage for Every Hobby One Man Could Collect

Some personal websites choose a subject. Manfred Mornhinweg’s Homo Ludens homepage chooses the person and lets every subject follow him home.

Visit Homo Ludens

The site’s premise is right in the name: homo ludens, the playing human. Mornhinweg introduces the pages as the toys, games and hobbies of one specimen, then splits himself into a collection of mock subspecies.

There is homo ludens radiactivus for amateur radio, radiohistoricus for antique radios, electronicus for electronics projects, volatrix for free flight, nauticus for a homemade folding kayak and sailing, musicus for music, aeromodellisticus for model aircraft, andinensis for climbing, photographicus for photography and mechanicus for tools and car repair. There are more.

The navigation sounds like a joke, but the material behind it is substantial. Mornhinweg has published electronics projects, restorations, technical explanations, photographs and long accounts of things he built or repaired. His pages have circulated for years among electronics hobbyists because they contain the sort of detail that tends to come only from somebody who actually constructed the circuit, opened the inverter or rebuilt the radio.

The site states its design philosophy plainly: content rather than special effects, bells and whistles. That sentence has aged extremely well. The pages remain quick to load and easy to read because most of the machinery is simply text, photographs and links.

There is also something distinctly pre-platform about organizing a website around a whole human being. Modern publishing tends to reward specialization. Pick a niche, stay in it, train the audience. Homo Ludens does the opposite. The unifying subject is curiosity itself.

For webmasters, that makes the structure worth studying. The homepage works as a map of interests, while individual sections can become as deep as they need to be. Nobody has to pretend antique radios and homemade kayaks belong in the same SEO category. They belong together because the same person cared about both.

CacheRat’s 1,967 Ancient Web Domains research list is full of websites built this way: a human being first, taxonomy second.

Homo Ludens is still online and still refreshingly unconcerned with personal-brand consistency.

Explore Manfred Mornhinweg’s Homo Ludens site

Posted on by

Ancient Web: Ark Two Buried 42 School Buses Under Concrete for the End of the World

There are old survivalist websites, and then there is Ark Two: a site built around a 10,000-square-foot Ontario fallout shelter whose concrete structure used forty-two school buses as permanent forms.

Visit the Ark Two Community

The site’s own description is remarkably direct. Ark Two presents itself as a pluralistic survival community intended to help people endure a nuclear war or other catastrophic event and participate in rebuilding afterward. The shelter is in Ontario, Canada, and the site openly provides photographs, maps and a project history rather than treating the location like a secret bunker in a spy novel.

The construction photographs are the obvious draw. According to the site, forty-two school buses were arranged and then encased in concrete before being covered by roughly five to fourteen feet of earth. Interior pages show rooms, corridors and a map substantial enough that the site’s author jokes, in effect, that occupants need navigation once inside.

But Ark Two is larger than the shelter page. Its navigation branches into survival, technical recovery, reconstruction and what it calls renewal. The worldview assumes that the catastrophe has happened and then asks what a community would need afterward: organization, practical knowledge, infrastructure and social continuity.

Whether a reader shares the site’s assumptions is almost beside the point for Internet archaeology. What survives here is a complete example of pre-social-media preparedness culture publishing its plans directly to the public. The site is not a listicle about emergency kits. It is somebody’s actual decades-long project, documented with photographs of concrete, buses, rooms, maps and physical construction.

It is also a reminder that old personal sites often mixed practical information with very strong personal convictions without separating everything into clean categories. The architecture of the website mirrors the architecture of the project: links branch into whatever the builder believed civilization might require.

For researchers looking at survivalism, Cold War afterlives or early preparedness culture online, that makes the site useful primary material. For everyone else, forty-two buses under concrete is a sufficient reason to click.

CacheRat’s 1,967 Ancient Web Domains research list contains many similarly strange specialist projects whose value is difficult to understand from a search-result snippet alone.

Ark Two is still online, maps and all.

Explore the Ark Two shelter pages

Posted on by

Ancient Web: DataPacRat Was Packratting Information Before CacheRat

Every so often an archaeological dig turns up something annoyingly close to your own business model.

Visit DataPacRat

DataPacRat’s homepage is almost aggressively simple. The owner says the site is a work in progress, explains that content matters more than design, and warns visitors to expect a primitive layout.

Then comes the line that makes the whole thing click.

The author says the data they have “packratted away” will be sorted into four categories: True, False, Opinion, and Art.

That is an excellent information architecture for a rat with a hard drive.

Save first, classify honestly

The categories are useful because they do not pretend every saved thing has the same evidentiary value.

“True” and “False” imply claims that can be evaluated. “Opinion” admits that some material is argument or perspective. “Art” gets its own shelf rather than being forced into the factual bins. It is a crude taxonomy, but it solves a problem enormous modern platforms routinely make worse: context collapse.

DataPacRat also hosts the Wulf Archives, which the homepage says were no longer available at their original location and were being provided as a courtesy until the author found a new host. An update then notes that Anthony Pryor later established a new website and published the Wulf Archives as ebooks.

That small sequence captures what independent preservation often looks like in practice.

A site disappears. Somebody else has a copy. They keep it available. The original creator resurfaces. The preservation copy becomes part of the chain rather than claiming ownership of the material.

No foundation was formed. No digital-preservation grant was announced. Somebody simply had the files and did not throw them away.

The page is also refreshingly indifferent to visual polish. Modern publishing software makes it easy to spend more time arranging cards, hero images and responsive breakpoints than deciding whether the information is worth keeping. DataPacRat explicitly chooses the opposite priority.

CacheRat’s 1,967 Ancient Web Domains research list exists for much the same reason: old information gets valuable when the obvious path to it disappears. The names are close enough that I am forced to acknowledge prior rat art.

There are worse ancestors to discover in the dig.

DataPacRat understood the important part early: collect the weird stuff, label it as honestly as you can, and do not let CSS become the reason the archive never gets built.

Dig through DataPacRat

Posted on by

Ancient Web: Masters of Melody Preserved the BBC Orchestras Radio Forgot

British radio once employed an astonishing number of studio orchestras, dance bands, light-music ensembles and military bands. Most people under a certain age barely know that ecosystem existed. Brian Reynolds built Masters of Melody to stop it disappearing completely.

Visit Masters of Melody

Reynolds describes the site as an antidote to a music culture that pushed light orchestral music out of regular broadcasting. He is not approaching the subject casually. The site says he owns a substantial collection of vintage broadcasts, composed light music himself and researched the subject in the BBC Written Archives.

The result is a dense index of names that search engines are unlikely to surface by accident: Sidney Bowman, Claude Cavalotti, Harry Fryer, Ronnie Munro, Primo Scala, Peter Yorke and dozens more. There are pages on BBC regional orchestras, military bands, old programmes such as Music While You Work, and individual broadcasts with dates and running orders.

Some pages include audio clips or full programmes. A typical entry does more than name the orchestra. It may identify the conductor, broadcast date, signature tune and complete sequence of pieces. That turns a nostalgia page into a research aid.

The site also preserves a vanished production model. Before recorded music completely dominated radio scheduling, agreements limiting recorded music helped keep live musicians working in studios. Reynolds’s pages document the enormous machinery that existed around that arrangement: staff orchestras, regional ensembles, arrangers, conductors and programmes that could fill hours of ordinary broadcasting.

Masters of Melody dates its copyright from 2008, so it is younger than many Ancient Web subjects. It belongs here because it carries forward the older personal-Web model almost perfectly: one knowledgeable person, one narrow subject, years of accumulated research, no obvious concern about whether the topic is commercially fashionable.

That is exactly the kind of site that becomes more valuable as the people who remember the subject become fewer.

The larger 1,967 Ancient Web Domains research list contains many similar specialist archives. They often look modest because the effort went into the material rather than the interface.

Masters of Melody is still being updated, which means the archive has not merely survived. It continues to grow around music that commercial media mostly stopped carrying decades ago.

Explore Masters of Melody

Posted on by

Ancient Web: CroatianHistory.net Is a Sprawling Independent National Reference

CroatianHistory.net looks less like a modern publication than somebody’s research desk expanded until it became a country-sized website.

Visit CroatianHistory.net

The homepage calls itself an overview of Croatian history, culture and science. From there it branches almost immediately into Glagolitic script, music, Dubrovnik, art, science, Latinists, Croats in Bosnia and Herzegovina, sports, the cravat, the Penkala pen, psychology, fingerprinting, Ruđer Bošković and a long list of people, inventions, places and historical arguments.

The result is not tidy.

It is useful precisely because it is not tidy.

This is a personal reference work, not a neutral encyclopedia

The site mixes biographies, scans, photographs, maps, old books, commentary, cultural claims and links to outside institutions. Some pages document specific artifacts or school records. Others advocate particular interpretations of Croatian history and identity.

That distinction matters.

A reader should not treat every assertion on a national-history site as settled scholarship merely because it comes with a map and a footnote. CroatianHistory.net sometimes makes arguments explicitly, including disputed historical or territorial interpretations. Those should be followed to the cited documents and compared with other scholarship.

But that is also what makes the site interesting as an artifact. It does not hide behind an anonymous institutional voice. You can see what its compiler considers important enough to document, defend, correct or reconnect to Croatian history.

The site is also very much alive rather than frozen. Its homepage now includes material from 2025, while older navigation preserves year-based sections reaching back through the 2000s. That makes it a long-running independent web project rather than a time capsule abandoned after one redesign cycle.

There are unusually concrete preservation signals too. The site lists an ISBN and states that, except where noted, its content is available under a Creative Commons Attribution license. It links outward to archives, museums, libraries and other Croatian cultural projects instead of trying to trap the reader inside one domain.

For a 2026 researcher, the best use is exploratory. Find a name, object, manuscript, invention or regional story you did not know existed, then follow the citations outward. That is exactly how old specialist websites remain valuable even when their editorial perspective needs scrutiny.

CacheRat’s 1,967 Ancient Web Domains research list is full of this kind of thing: dense sites built because one person or small group believed a subject deserved more links than a normal publication would ever tolerate.

Modern sites optimize the path through information.

CroatianHistory.net keeps opening more doors.

Explore CroatianHistory.net

Posted on by

Ancient Web: Orville the Mouse Became the Internet’s First Electronaut

On January 8, 2003, a mouse named Orville took two short flights aboard a homemade high-voltage lifter and was promptly awarded one of the better job titles in Internet history: electronaut.

Meet Orville the Electronaut

The page belongs to Jean-Louis Naudin’s sprawling collection of experiments with so-called lifters: lightweight asymmetric-capacitor devices that can produce thrust in air using high voltage and ion movement. The mechanism is generally discussed in terms of electrohydrodynamic or ionic-wind effects, although old pages around the subject often used much more ambitious language such as “electrokinetic” or even “antigravity.”

Orville’s page is memorable because Naudin did not stop at lifting bits of foil and balsa wood. The Maximus lifter was built large enough to carry a mouse as payload. Contemporary copies and references to the experiment describe Orville completing two tethered flights and returning in good health. The power source remained external; this was not a tiny self-contained spaceship. It was, however, a mouse in a handmade flying high-voltage contraption, which is already enough Internet for one day.

The experiment became widely linked. Physics-education writing later referenced Orville when discussing ion-powered flight, and old forum threads kept passing the page around years after the original test. The reason is obvious: a technical phenomenon that might otherwise remain an abstract diagram suddenly has a pilot.

The page is also a perfect artifact of its era. Early experimental Web culture was full of people publishing photographs, test notes, schematics and increasingly strange demonstrations directly from personal pages. There was no need to package the work into a video channel, build a personal brand or wait for a platform to decide whether the subject had engagement potential.

There is a useful skepticism lesson here too. The lifter visibly produces thrust, but that does not establish exotic propulsion or gravity control. The interesting part is the documented home experiment, not the grandest interpretation anybody attached to it afterward.

CacheRat’s 1,967 Ancient Web Domains research list contains plenty of these garage-laboratory pages. They are messy, occasionally overexcited and often far more interesting than polished corporate science communication.

Orville’s flight survives because somebody bothered to make a page for the mouse.

Return to Orville’s flight page

Posted on by

Ancient Web: FlintSource Mapped the Rocks Behind Prehistoric Tools

A prehistoric stone tool does not begin with the tool. It begins with somebody knowing where to find the right stone. FlintSource.NET has spent a quarter-century documenting those source locations with the sort of detail normally produced by somebody who has stood there, picked up the material and taken notes.

Visit FlintSource.NET

The site marked its 25th anniversary in 2024 and dates its beginning to December 30, 1999. Its subject is archaeological raw material: flint, chert, hornstone, quartzite, radiolarite, obsidian and related stone sources.

That sounds narrow until you open an entry.

A typical page may identify the geological formation, give latitude and longitude, describe cortex and color, show field photographs, document sample numbers and photograph actual flakes or nodules. Entries for Nordic flint around Denmark, for example, include GPS coordinates, source photographs and individual samples measured in millimeters. The Calais page contains the kind of field note only an old specialist site would preserve: do not blindly follow one coordinate to the second in fog, darkness or poor visibility because the sample point was on the edge of a cliff and part of it may have fallen into the sea since 2002.

That is much more useful than a generic encyclopedia paragraph about flint.

The website still advertises itself as best viewed in Internet Explorer 5+ or Netscape 6+, which is a fairly effective carbon date for its design. Yet the actual database remains intelligible because the structure is simple: source name, geology, locality, coordinates, photographs, sampling information and references.

For archaeology, material provenance matters because raw stone can connect an artifact to trade, movement or a geological source. FlintSource does not magically answer those questions, but it preserves field observations that can help researchers compare materials and locations.

For webmasters, it is another example of durable specialist publishing. The visual design is old. The data model is obvious. The content survives because pages were built around stable facts instead of a stream of posts that immediately disappeared off the front page.

CacheRat’s 1,967 Ancient Web Domains research list includes plenty of this kind of specialist knowledge: things modern search can technically find but rarely explains this well.

FlintSource is still online, still searchable, and still capable of making a rock on the ground feel like a research problem.

Explore FlintSource.NET

Posted on by

Ancient Web: A Columbine Site Preserves an Important Warning About Fake Web Mirrors

A bad archive can preserve misinformation just as efficiently as it preserves files.

That is what makes this page from A Columbine Site worth reading carefully.

Visit the Eric Harris webpages archive page

A Columbine Site describes itself as a long-running research archive dedicated to the injured, the survivors and those who died in the April 20, 1999 Columbine High School shooting. Its Eric Harris webpage section deals with material that existed online before the attack and what happened to that material afterward.

The important part is not the killer’s writing. It is the site’s unusually explicit discussion of provenance.

The page says Harris had used several online services and maintained webpages, but that his actual pages were removed before their addresses became widely public. The site’s author says some directory material and files were preserved, while the original HTML for the webpages was already gone.

That created a problem.

A reconstruction escaped its label

The archivist explains that in 1999 they displayed surviving graphics and links in a layout intended to help readers imagine how some of the missing pages may have looked. Other websites later copied those displays and presented them as mirrors of Harris’s original webpages.

According to A Columbine Site, they were not.

The page now goes out of its way to distinguish actual preserved directories from later reconstruction pages and says there are no surviving copies of Harris’s original websites. Whether a reader is researching Columbine specifically or digital history generally, that distinction is valuable.

Once a screenshot, mirror or copied page loses its chain of custody, repetition can turn a reconstruction into supposed evidence. Search engines then index the copies. New writers cite the search results. Eventually the false provenance becomes easier to find than the correction.

This site is also a reminder that disturbing historical material needs context. A Columbine Site’s homepage frames its collection around victims, survivors, official documents, reporting and the consequences of the attack. That is a far more useful archival purpose than turning the perpetrators into Internet folklore.

For researchers in 2026, the lesson extends well beyond this event: “I found it online” is not provenance. A mirror may be genuine, partial, reconstructed, mislabeled or copied from another reconstruction. The URL alone cannot tell you which.

In this case, the surviving archive does something responsible and surprisingly rare.

It tells you exactly where its evidence stops.

Read the original archival explanation

Posted on by

Ancient Web: Machine Learning in Games Kept the Old AI Reading List Alive

Long before every software company discovered the phrase “artificial intelligence,” Jay Scott was maintaining a website about something much narrower and more useful: how computers learn to play games better.

Visit Machine Learning in Games

The site is aimed at AI researchers and game programmers. It organizes material around heuristic search, neural networks, genetic algorithms, temporal-difference methods and other techniques, then connects those methods to actual game-playing systems.

The subject list is a time capsule by itself. There are neural-network backgammon programs, strong learning Othello programs, pursuit-evasion research, robotic soccer, NeuroChess, Go networks, Metagame and general game playing. Later additions include StarCraft competitions and the BotPrize Turing-test experiments in Unreal Tournament.

More important than the individual examples is the way the site worked as research infrastructure. Scott maintained indexes of games and people, tutorials, links to AI resources, lists of online papers, software, workshops, competitions and a bibliography. Other academic pages from the period cite it as a place to find researchers and work in strategic game learning.

Scott’s current homepage now describes Machine Learning in Games as long out of date but still of historical interest. That is a wonderfully accurate description of why it matters in 2026. The machine-learning field moved on at terrifying speed, but an old hand-curated map of what researchers thought was important at the time tells you something a freshly generated search result cannot.

The page also captures a different AI culture. The problems are concrete. Can the program learn Othello? Can it improve at backgammon? Can evolutionary methods produce better strategies? There is little mystical language and not much pretending the machine has become a coworker with feelings. Mostly there are algorithms, papers and games.

For researchers, that means the dead or aging links are part of the evidence. For webmasters, it is a reminder of what a genuinely useful link directory looked like: a knowledgeable person imposing structure on a technical field because they had actually read the material.

CacheRat’s 1,967 Ancient Web Domains research list contains many of these specialist indexes. They are valuable not because every link still works, but because they preserve the shape of a field at a particular moment.

Machine Learning in Games is still there, cheerfully obsolete and therefore increasingly useful as history.

Explore Jay Scott’s Machine Learning in Games