Category: Ancient Web

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

Posted on by

Ancient Web: AmazeWorld Turns Plain Web Pages Into Tiny Clickable Adventure Games

Before browser games meant a JavaScript framework loading a miniature console into a canvas element, a website could make an adventure out of the web’s most basic action: click the next link.

Enter AmazeWorld.

AmazeWorld offers a set of interactive maze/adventure experiences built around that primitive but effective idea. One sends you into a dungeon to find a key, defeat a dragon and escape. Another involves stopping an alien threat. A city scenario asks you to locate and dismantle a bomb. A Cinderella-themed maze sends the player after a princess.

The presentation is less important than the mechanism. A web page is a location. A click is movement. An image can hide choices. Another HTML page becomes the next room.

That is a genuinely web-native game design. Instead of pretending the browser is a different machine, AmazeWorld uses the browser’s own navigation model as the engine. History, Back and links are not chrome around the game; they are part of how the player moves through it.

The approach also made distribution nearly frictionless. No installer. No account. No plugin if the implementation stayed simple enough. If somebody sent you the URL and your browser could render images and links, you had the game.

There were limitations, obviously. Loading a new page for every move could be slow on dial-up. Image-map puzzles could be frustrating. State was difficult to manage compared with modern client-side games. Yet those constraints encouraged designers to think about branching structures and spatial illusion in unusual ways.

AmazeWorld is worth preserving because small browser experiments disappear faster than commercially released games. There may be no cartridge to dump, executable to archive or company catalog proving that the thing ever existed. The website is the object.

It also captures an era when people regularly asked, “What weird thing can HTML do if I push it?” rather than starting from a full game stack.

Try AmazeWorld while the doors still open. It is a reminder that the Web was always a game engine if you were willing to count hyperlinks as doors.

Posted on by

Ancient Web: Kelesis Preserved Compile’s Shooter Obsession One Fan Page at a Time

Game companies disappear. Fan pages sometimes outlive them. Kelesis is a good example of why old enthusiast sites can become more useful after the commercial moment has passed.

Visit the Kelesis home page.

The site revolves heavily around Compile, the Japanese developer associated with shooters such as Zanac and Aleste. Surviving indexes and references show sections for Zanac X Zanac, Compile shooting games, PC shooters, music, art galleries, cosplay, news, links and guestbook material.

A 2001 news entry captures the site while the subject was still alive rather than safely retro. It anticipates the November 29 release of Zanac X Zanac, notes updates to the Zanac material and points toward broader histories of Compile shooters including Zanac, Spriggan and A.E..

That timing makes the archive especially valuable. Retrospectives written decades later know which games became canonical. A contemporary fan site shows what people were excited about before the ranking was settled.

Kelesis also reflects how video-game knowledge circulated before wikis absorbed most of it. A serious fan might maintain a developer chronology, screenshots, artwork, music notes, reviews and links to other specialists. The result was messy and decentralized, but it captured enthusiasm in much higher resolution than a single encyclopedia entry.

The artwork side is another preservation layer. The site has been cited elsewhere for galleries related to artists such as Ayami Kojima. Fan-curated images are vulnerable because they often live on personal hosting, and the contextual links around them disappear when images are copied into anonymous reposts.

There is a technical-history lesson here too. Compile’s catalog spans hardware generations and naming conventions that can confuse newcomers. Old fan sites were often where regional releases, alternate names and series relationships were patiently untangled.

Not every detail on a fan site should be treated as primary evidence. Release dates and credits are worth checking against manuals, publishers and archival databases. The site’s strength is the map it gives you of what to investigate.

Explore Kelesis. It preserves something a polished retrospective cannot recreate: the feeling of a fan documenting Compile while there was still another Compile release to wait for.

Posted on by

Ancient Web: WebCars Turned One Car Guy’s Obsessions Into a Reference Library

WebCars is what happens when a car enthusiast gets a domain and never receives the memo that websites are supposed to stay inside one tidy content category.

Visit WebCars.

The site, maintained by Paul Pollock, ranges across Corvette history, Ferrari and Lamborghini features, Jaguar, Pontiac, Austin-Healey, car humor, concours events, wallpapers and a collection of automotive calculators. There is a detailed Corvette Story project, material on specific model years and engineering documents, and the kind of ownership notes that only appear when somebody has actually spent time fighting an old car.

Pollock has been publishing automotive material online for years. A 2007 Corvette forum post points readers to WebCars wallpaper photographs and identifies him as the site’s webmaster. His Corvette material is still updated and still answers questions that factory statistics cannot cleanly solve, such as how many cars were built with a very specific combination of body style, color and option package.

That distinction matters. A commercial car database wants normalized fields. An enthusiast site can say, in effect, “the available records do not let us know that exactly,” then explain what can and cannot be inferred.

WebCars also contains wonderfully specific computer-era utilities. Fuel-mileage pages turn long-running driving records into statistics. Other sections provide basic automotive math and reference calculations. It is the old web pattern of building a tool because you wanted it, then leaving it online for anybody equally obsessive.

The Corvette Story material goes deeper than fan photography. It includes period documents such as a multi-page 1963 SAE engineering paper and navigable model-year histories. That gives researchers a trail from enthusiast commentary back toward contemporary technical material.

As always, numbers from hobbyist sites should be checked against factory documentation when the difference between two production figures matters. But that caveat should not obscure what sites like WebCars accomplished: they organized material manufacturers often scattered across brochures, service literature, club knowledge and personal experience.

Browse WebCars. It is less like an automotive magazine than a garage where somebody labeled most of the drawers, then kept adding rooms.

Posted on by

Ancient Web: Michelle Hoyle’s History of Computing Science Became a Citation That Refused to Die

Some old web pages survive because people keep linking to them long after their visual design has become archaeological evidence. Michelle A. Hoyle’s History of Computing Science is one of those stubborn references.

Visit The History of Computing Science.

The project is cited in academic reading lists as dating from roughly 1994 through 2006. Its pages walk through the long prehistory of computing rather than beginning with the personal computer: mechanical calculators, logic, early machines and the inventions that gradually turned calculation into programmable computing.

Individual pages have been cited years later in academic work. The site’s material on Pascal’s calculator, for example, has appeared in scholarly references, while later educational sites credit Hoyle’s computing-history work as a source for their own historical overviews.

That is a useful reminder of what educational publishing looked like before institutions standardized everything into learning-management systems. A subject expert or enthusiast could build a hypertext course-like resource in public, and then other teachers could simply point students at it.

There are drawbacks. Old pages can contain outdated scholarship, vanished images, browser assumptions and links that now lead nowhere. A computing-history page written in the 1990s is itself old enough to need historical context. Claims should be checked against modern scholarship when precision matters.

But the site’s durability is the interesting part. It became part of the citation ecology of the web. Books, university materials and later history pages did not merely consume it; they left evidence that the resource had been useful to generations of readers.

The architecture also fits the subject unusually well. Computing history is not a straight line from abacus to smartphone. It is a network of mathematics, engineering, punched cards, mechanical devices, logic, wartime projects and competing ideas about what a computer should be. Hypertext is a natural way to wander through that mess.

In 2026, the page has a second layer of value. It documents computing history while simultaneously becoming part of web history itself.

Browse Michelle Hoyle’s History of Computing Science. The old web did not wait for a publishing platform to decide whether a lecture series deserved a home. It just made one.

Posted on by

Ancient Web: Ayr.org Gave a Scottish Town Its Own Corner of the Web

Before every place on Earth became a stack of map listings, tourism portals and social-media location tags, towns often acquired websites that felt more like local scrapbooks. Ayr.org is one of those independent place-on-the-web projects.

Visit Ayr.org.

The site presents itself as a home for the town of Ayr in Ayrshire, Scotland, with photographs, history, news, maps and related local material. The domain dates back to the early years of mass public web use and remains reachable over HTTPS in 2026.

The concept sounds ordinary now, but the structure matters. This is not simply a council service page and not a travel-booking funnel. A town-oriented independent site can accumulate material according to local curiosity: an old photograph here, a map there, a historical note that would never justify its own commercial article.

Ayr has plenty of formal history to support that kind of wandering. It is a long-established burgh on Scotland’s southwest coast, associated with Robert Burns country and the wider history of Ayrshire. A local web archive does not need to retell all of that in chronological order. It can work more like somebody pointing things out while walking.

That is also why these sites become useful research leads. Search engines tend to privilege large current sources, while an older local page may preserve terminology, image captions or neighborhood references that lead toward newspapers, maps and archival records.

Of course, an independent site should not be treated as the final authority simply because it is old. Dates and historical claims still deserve checking against primary records. Its value is different: it records how a place was being described and organized on the public web outside institutional publishing systems.

There is a pleasant literalness to owning the town name as a domain and filling it with town material. No app required. No account. No review score deciding whether the harbor or an old street is worth seeing.

Explore Ayr.org as a surviving example of local-web publishing. A town is too complicated to fit neatly into a tourism card, and the older web often understood that by accident.

Posted on by

Ancient Web: Jake Meinershagen’s Homepage Is a Drawer Full of Useful and Useless Stuff

Modern websites are encouraged to explain their purpose in one sentence. Jake Meinershagen’s page answers that demand with something better: “Behold, my stuff.”

Visit Jake Meinershagen’s site.

The links underneath are exactly what that sentence promises. There are personal ramblings, astronomy beginner information, instructions on finding things with a telescope, material about the Large Compact Telescope, a page about standing in line at Powell, a countdown timer, a discussion timer and material about artificial intelligence.

There is no obvious content strategy tying this together because a human being is the organizing principle.

That used to be an ordinary architecture for the web. Your homepage was where you put the things you knew, made or wanted to remember. If you wrote a useful telescope guide, it lived next to your joke page. If you built a timer because you needed one, other people could use it too. A site did not have to choose between being a blog, software product, knowledge base or portfolio.

The astronomy material is a good example of why these pages remain useful. Beginner advice written by an experienced person often contains the small practical details formal references skip: how to start looking, what expectations to have and how to translate an abstract sky chart into something you can actually find through glass in the dark.

Then, one click away, the subject can change completely.

This kind of structure is terrible if your goal is to make every visitor travel through a conversion funnel. It is excellent if your goal is to leave behind a map of what interested one person over time.

In 2026, personal homepages like this are also increasingly valuable to search. The web is saturated with pages designed around queries. A miscellaneous personal page can contain an answer nobody commissioned because somebody solved the problem for themselves and bothered to publish it.

That is the old web at its most practical: not a museum aesthetic, just people leaving tools and notes where strangers can reach them.

Browse Jake Meinershagen’s collection of stuff. The lesson is almost embarrassingly simple: a website does not need a niche when the niche is the person who made it.

Posted on by

Ancient Web: Evergreen Argues That a Website Should Feel Like a Place You Made

Not every Ancient Web artifact has to be ancient. Evergreen is interesting precisely because it treats the old personal-web instinct as something worth practicing now instead of something to mourn behind glass.

Read the Evergreen manifesto.

The surrounding Evergreen material encourages people to make websites as personal spaces: home pages, writing pages, art galleries, about pages, manifestos, link pages, guestbooks, blogs, dream journals, shrines and whatever other category happens to matter to the person building it. Tutorials connected with the project point toward tools such as Hotglue, which lets users assemble pages visually while still leaving room for custom HTML and JavaScript.

The important word is not “retro.” It is ownership.

A social profile gives you a rectangle inside somebody else’s product. A personal site can contain a page devoted to one band, an unfinished essay, a scanner dump, a list of friends, a collection of tiny GIFs and a deeply unnecessary shrine to a fictional character. There is no product manager asking whether these objects improve engagement.

Evergreen deliberately reaches for pieces of web culture that platforms mostly erased: webrings for wandering between related sites, guestbooks for leaving a mark, links pages that reveal what a person actually reads, and hand-built navigation that reflects the site’s subject rather than a universal template.

That does not mean old technology was magically better. The early web had broken links, unreadable color combinations, terrible accessibility and pages that only worked in one browser at one screen size. The useful lesson is not to reproduce every mistake. It is to remember that a website can be an authored object rather than merely a delivery surface.

The manifesto also belongs to a broader preservation problem. Handmade sites vanish constantly because their value is difficult to measure with traffic statistics. A tiny page may matter intensely to fifty people and look worthless to an advertising system.

In 2026, that makes projects like Evergreen less nostalgic than they first appear. They are asking whether the web still has room for small things that do not need to scale.

Visit Evergreen and follow the trail outward. The best argument for the personal web is still the same one the old web made accidentally: people are more interesting when they are allowed to build something stranger than a profile.

Posted on by

Ancient Web: The Cronan Memorial Shows How Real a Person Could Become Without Ever Meeting Them

One of the easiest lies to tell about the early Internet is that online relationships were somehow less real because people had never met. The Cronan Memorial is a small, stubborn counterexample.

Visit the Cronan Memorial.

The person remembered online as Plain and Simple Cronan was a sharp, funny participant in science-fiction and fandom discussions. Later recollections from people in those communities describe looking forward to his posts alongside conversations involving shows such as Babylon 5, Deep Space Nine and other genre television.

Then, in 1999, the community learned that Cronan had died from Hodgkin’s disease. He was only 19.

What survived was not merely an obituary. It was a web memorial built for someone many participants knew primarily through text. Years later, people still described being surprised by the force of the loss. One remembered realizing that although they had never met Cronan in person, they had effectively known him for years.

That is historically useful because the web has always been better at documenting famous deaths than ordinary online grief. News archives retain celebrities. Social platforms retain whatever their retention policy happens to allow. Small memorial sites preserve a different category: the person whose importance existed inside a particular group of usernames, threads and recurring conversations.

The page also comes from a period when identity online was often deliberately incomplete. A handle could be the name everybody knew. A personality emerged through thousands of messages instead of a profile assembled from legal names, employment histories and verified photographs.

None of that made the relationship imaginary. In some ways it made the writing itself carry more weight. Humor, timing, opinions and the habit of showing up were the person as experienced by the group.

The old web is full of abandoned fan pages and dead guestbooks, but memorial pages deserve a different kind of attention. They record the moment an online community discovered that somebody on the other end of the modem had been woven into their real lives.

Return to the Cronan Memorial. It is a quiet artifact, but an important one: proof that decades before people argued about whether Internet friends counted as “real friends,” users had already learned the answer the hard way.

Posted on by

Ancient Web: Book-A-Minute Was Speedrunning Literature Before Speedrunning Had a Name

Long before every platform rewarded people for boiling culture into fifteen seconds, Book-A-Minute was already taking famous books, stripping them to a handful of lines and seeing whether the skeleton was funny.

Visit Book-A-Minute.

The site is part of RinkWorks, one of those old entertainment domains that accumulated jokes, puzzles, games, movie material and strange side projects instead of deciding what its “content vertical” was supposed to be. Book-A-Minute’s specific gimmick is brutally simple: condense fiction into the smallest irreverent version possible.

Its major sections include classics, science fiction and bedtime stories. The range is part of the joke. Jane Eyre, Ender’s Game and children’s books can all be fed through the same compression machine. The result is not study material and does not pretend to be. A 2005 newspaper write-up made that distinction explicit: these were tiny comic reductions, not replacements for reading the book.

That format turns out to be a useful little artifact of web humor. There is no video, reaction face, autoplay, recommendation engine or elaborate setup. The page assumes you already know enough about the source material for three or four exaggerated plot beats to land.

It also preserves a kind of amateur criticism disguised as stupidity. To reduce a novel effectively, somebody has to decide which obsession, plot mechanism or character behavior defines it. The better entries work because they identify what readers remember and then hit it with a hammer.

RinkWorks itself dates to the late 1990s, and Book-A-Minute feels native to that period: a single premise expanded into a hand-built archive because somebody thought it was funny enough to keep doing. Modern social media could reproduce the joke, but it would scatter the individual bits into a feed. The old site keeps them together as a browsable collection.

That matters in 2026 because early web comedy is easy to lose. A joke page may have no commercial value, no database somebody wants to migrate and no obvious reason for a company to preserve it. Yet collections like this explain how people actually entertained each other online before every joke became a post inside somebody else’s platform.

Browse Book-A-Minute and pick a book you know well. The compression is ridiculous, but that is the entire machine: remove ninety-nine percent of a story and see whether the remaining one percent still tells on it.

Posted on by

Ancient Web: Petra Cortright’s Website Treats Bad HTML Taste as Part of the Artwork

Most artist websites eventually become tasteful grids: white background, giant thumbnails, biography, exhibitions, contact. Petra Cortright’s old personal site goes in the opposite direction. It looks like somebody handed early web culture a glitter cannon and told it to organize a serious art archive.

Visit Petra Cortright’s hello.html.

The page is a dense index of work organized by year and medium, but calling it an index makes it sound calmer than it is. The navigation is full of arrows, asterisks, misspellings, aggressive capitalization and titles such as SICKWOOF, 666 SMIELYZ, TROLLY XPLOSION, CATS SPIRT SPSIT SPIT and SSSSSSSSSSSSWWWWRRRRRLLLLLL.

Underneath the visual noise is a surprisingly useful chronology. Video pieces run through 2007–2011. Still-image work reaches back to pixel paintings and selected LiveJournal entries from 2004. There are ASCII drawings from 2006, GIF pieces, collaborations, landscapes, drawings, web video and later links to catalogs and studio contact information.

Cortright is important to Internet art partly because the materials of ordinary online life are not hidden behind a gallery-neutral presentation. Webcam video, cheap effects, compression, browser behavior, animated GIFs and amateur-looking graphics are allowed to remain visibly digital.

The famous 2007 work VVEBCAM, for example, came from the period when webcams, YouTube aesthetics and desktop effects were becoming ordinary cultural material rather than specialized computer tricks. Elsewhere the site preserves ASCII drawings, pixel paintings and animated pieces without pretending those forms need to be translated into something more respectable first.

That makes the website itself part of the record. A clean redesign could preserve the files while destroying some of the context: the feeling that web pages were spaces artists could manipulate directly instead of neutral frames supplied by a platform.

There is an old-web confidence to it. Links do not apologize for being links. File names and page names remain visible. Navigation can be eccentric. The website assumes the visitor can poke around and figure things out.

This matters because net art is unusually vulnerable to preservation problems. Browsers change. Plugins disappear. Embedded services die. Video formats become awkward. Domains lapse. Even when individual works survive in museum collections, the original personal site can retain relationships between pieces that institutional descriptions smooth away.

Explore Petra Cortright’s web-art index. It is both portfolio and artifact: evidence from a period when making art for the Internet could still mean making the Internet itself look stranger.