Category: Ancient Web

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

Posted on by

Ancient Web: Carolyn’s Creations Preserves the Petz Modding Underground

Some old websites preserve games. Carolyn’s Creations preserves what players did to the games after they got bored with the factory settings.

Visit Carolyn’s Creations

The page is an enormous archive built around the original P.F. Magic Petz games and their relatives, including Catz, Dogz, Oddballz and Babyz. Carolyn describes it as an archive of everything she hex-edited or converted for those games, plus material for Petz 5. The navigation alone reads like an archaeological inventory: breedz, toyz, clothes, playscenes, hosts, sounds, easy-edit kits, tutorials, fixes and utilities.

This was modding before the word automatically implied a Steam Workshop button.

The troubleshooting is almost as valuable as the files

Carolyn’s notes preserve the operating environment around the software. One section explains that an old Petz player plug-in will not work in modern browsers and points users toward Netscape Communicator 4. Elsewhere she discusses Windows 95, Windows 98, Windows 2000, Windows ME and Windows XP compatibility problems with the casual resignation of somebody who has already lost several weekends to them.

There is also a blunt distinction between the original P.F. Magic games and later Ubisoft titles carrying the Petz name. Carolyn says her tutorials and downloads concern the older games, not Ubisoft releases from 2006 onward.

That detail matters. Old game communities routinely outlive the companies, products and official websites that created them. Carolyn explicitly notes that the original game websites are gone. Her personal archive became part download mirror, part technical manual and part oral history of how the software actually behaved in people’s homes.

The most telling paragraph may be the one about exclusivity. Carolyn says she does not make private “exclusive” creations for individuals and that her ethos has always been to make information and modifications free for everyone. That is not corporate community-management language. It is the value system of a hobbyist web scene that expected people to learn from one another, remix things and pass the knowledge along.

The site even asks other site owners not to hotlink files and to credit her work if they repost it. Those little rules reveal the bandwidth economics and etiquette of the period as clearly as the graphics do.

CacheRat’s 1,967 Ancient Web Domains research list is full of pages where the interesting artifact is not just the file but the ecosystem around it. Carolyn’s Creations is a perfect example. The downloads are useful. The instructions, compatibility notes, community vocabulary and stubbornly maintained knowledge are the real historical payload.

A mod archive this deep should probably have died with three hosting companies and a browser plug-in.

Instead, it is still explaining Petz hex editing to anyone curious enough to click.

Browse Carolyn’s Creations

Posted on by

Ancient Web: A 2000 Pokémon Page Explained the Seizure Incident With Actual Neuroscience

In March 2000, a university neuroscience site looked at one of the strangest collisions between television, animation, and medicine: the 1997 Pokémon broadcast that sent hundreds of children in Japan to hospitals.

The original “Pokémon on the Brain” page is still online.

The page was part of Eric Chudler’s Neuroscience for Kids project at the University of Washington. Instead of treating the event as urban legend or cartoon trivia, it used the incident as a compact lesson in photosensitive seizures and how the visual system responds to flashing light.

The broadcast was Pokémon episode 38. During one sequence, red and blue light flashed rapidly on screen. The page reports that 685 children were taken to hospitals by ambulance after viewers experienced symptoms including blurred vision, dizziness, nausea, headaches, seizures, and loss of consciousness. More than 150 were admitted.

The page then does what good educational websites used to do extremely well: it slows down and explains the mechanism.

It describes photosensitive seizures, discusses the relationship between flicker frequency and visual response, and summarizes experiments comparing ordinary visual responses with those of people sensitive to flashing light. It also records broadcasting guidelines adopted after the incident, including limits on the rate and duration of flashing imagery.

More importantly, the page cites the research. References at the bottom include papers in Epilepsia, Nature Neuroscience, and Acta Paediatrica Japonica. That matters because the Pokémon incident spent years mutating through playground stories, sensational television reports, and retellings that gradually lost the distinction between what happened and why.

This little 2000 HTML page preserved the useful version.

It is also a snapshot of how science outreach looked before every educational topic had to become a video channel, newsletter funnel, interactive app, and social account. The page has small graphics, plain headings, references, a few jokes, and enough explanation to send a curious kid somewhere deeper.

In 2026, that feels almost luxurious.

The surviving Pokémon neuroscience page is worth reading as both science communication and Web history. CacheRat’s 1,967 Ancient Web Domains research list contains more educational pages from the era when somebody with expertise could simply publish the explanation and leave it online for twenty-six years.

Posted on by

Ancient Web: Beastie’s Official Home Explains Why the BSD Daemon Is Not Chuck

The BSD daemon is one of those computer mascots that escaped the operating system and became part of hacker culture. His official home still looks appropriately like something maintained by people who care more about provenance than redesigns.

The History of the BSD Daemon lives at McKusick.com.

The first thing the page clears up is the name. The daemon is not officially named Chuck. Marshall Kirk McKusick says that name came from an advertising myth associated with Walnut Creek. If you insist on calling the character something, the site suggests “beastie,” a play on the letters BSD.

From there the site becomes a surprisingly detailed history of Unix culture expressed through drawings, book covers, T-shirts, conference swag, legal permissions, and inside jokes.

The black-and-white 4.2BSD daemon appeared on the 4.2BSD manuals published by USENIX. Years later, a group at Kansas State University decided they wanted a UNIX shirt. They scanned the old artwork, edited it on a Sun workstation, printed copies, and sat around with colored pencils until they found a color scheme they liked. The initial run was only about forty shirts.

The 4.3BSD daemon has another famous connection: the artwork used on The Design and Implementation of the 4.3BSD UNIX Operating System was drawn by John Lasseter, long before his name became universally associated with Pixar.

The site also preserves the stranger administrative side of mascot history. McKusick has a page explaining how the daemon artwork may be used, why permissions matter, and why he became serious about protecting the figure after nearly losing control of it to a company. There are galleries covering FreeBSD, NetBSD, OpenBSD, USENIX, “!DOS,” Lego versions, cakes, shirts, and other mutations of the character.

This is where an old personal website can beat a modern summary page. It does not merely tell you that Beastie exists. It preserves who made which version, what it appeared on, why particular shirts were printed, and how the image moved through actual Unix communities.

The page is basically oral history with HTML around it.

You can still wander through the BSD daemon archive and follow the individual designs into their stories. CacheRat’s 1,967 Ancient Web Domains research list contains more sites where the people involved in computing history simply documented it themselves instead of waiting for somebody else to summarize it later.

Posted on by

Ancient Web: Beej’s Pirate Image Archive Is a Pre-Google Image Search

There was a time when finding a usable image online did not mean typing three words into an image-search box. Somebody had to collect the pictures, scan them, organize them, write the navigation, and keep the files online. Beej’s Pirate Image Archive is exactly that kind of site.

Visit Beej’s Pirate Image Archive

The archive is refreshingly literal. Its navigation includes pirates, battles, ships, pirate fun, punishments, maps, Howard Pyle artwork, flags, reading material, and source information. If you need an old engraving of a ship, a pirate illustration, or a map with enough skull-and-crossbones energy to make a school project instantly more dangerous, this is where the site expects you to look.

Most of the images were scanned from printed sources. Beej explains that an effort was made to select works whose originals appeared to be in the public domain, while also warning users to verify rights for themselves rather than treating the archive as a legal guarantee.

That disclaimer is another artifact of the older Web. The author is not pretending to be a stock-photo corporation. It is closer to: here is the collection I assembled, here is where it came from, and here is what I think about the rights situation. Check before you do something expensive.

Curation used to be visible labor

Modern search makes collections like this easy to underestimate. Today the indexing layer is mostly hidden. You see a search field and results appear.

On a hand-built archive, the categories expose the curator’s decisions. Somebody decided that Howard Pyle deserved his own section. Somebody separated maps from battles and ships. Somebody maintained a source page because provenance mattered.

The site also has the advantage of being understandable without an account, JavaScript framework, infinite scroll, or recommendation engine. The information architecture is right there in the sidebar. A researcher can see the shape of the collection in seconds.

For webmasters, that is still a useful design lesson: a small archive benefits from obvious categories more than clever interaction. The visitor came for pirate pictures. Give them a button that says Pirates.

For historians, the archive preserves more than imagery. It preserves an older pattern of online research in which enthusiasts built subject-specific collections because general search was not yet good enough to replace them.

CacheRat’s 1,967 Ancient Web Domains research list contains many pages like this: small, specialized indexes assembled before search engines absorbed most discovery into a single interface.

Browse the original Pirate Image Archive

Posted on by

Ancient Web: Freshette’s Graphics Is a Living Fossil of the Decorative Web

Before website themes came as zipped templates and page builders, a lot of the Web was decorated one GIF at a time.

Visit Freshette’s Graphics

Freshette’s Graphics is still sitting there with the evidence.

The homepage opens on an autumn theme and immediately delivers what it promises: leaves, decorative bars, pumpkins, animated graphics, page furniture and links to more holiday pages. The author, signing as Freshette, invites visitors to use material from the site under its terms and asks them to place a logo and link back.

That small exchange was once a major part of how personal websites spread visual culture.

You found a graphic set somebody made. You saved the images. You decorated your own page. Then you linked back to the artist so the next webmaster could find the source too.

No package manager required.

The page still has its old social plumbing attached

The best part of Freshette’s Graphics is not any single pumpkin or gold divider. It is the surviving infrastructure around them.

The page still contains a holiday-graphics RingSurf webring with previous, next, random and join links. It carries an old-style visitor counter reference. The copyright line begins in 1995. Image files have wonderfully literal names such as fg_fallleaves.gif, fallgoldchainbar.gif and fg_pumpkinsoval.gif.

This is the decorative personal Web with the walls opened up so you can see the pipes.

Web graphics sites were important because ordinary homepage builders were not necessarily illustrators. Collections like this gave people a vocabulary of backgrounds, bullets, buttons, dividers, animations and seasonal decorations they could assemble into something personal. The result could be tasteful. It could also look like Christmas had detonated inside Netscape Navigator. Both outcomes were culturally significant.

Today a theme framework tends to enforce consistency. These older graphics libraries encouraged recombination. A page might borrow a border from one person, an animated mailbox from another, a background from a third, then attach five webring controls and a guestbook. Visual identity emerged from collecting pieces.

That is why sites like Freshette’s deserve more than a laugh at their aesthetics. They document a participatory design culture in which website owners visibly made things for one another.

CacheRat’s 1,967 Ancient Web Domains research list includes plenty of pages where the design itself is part of the historical record. Freshette’s Graphics goes one step further: it preserves the raw decorative material other pages were built from.

The Web does not need sixteen animated autumn leaves in a row anymore.

But it is considerably more interesting because somebody left them there.

Browse Freshette’s Graphics

Posted on by

Ancient Web: A.N. Lucas Rescued More Than 700 Tiny Web Buttons

The old Web had a standard size for tiny declarations of allegiance: 88 pixels wide by 31 pixels tall.

A.N. Lucas’s Web Lounge

A.N. Lucas has preserved hundreds of them.

Her Neocities site includes a collection of more than 700 88×31 buttons from the 1990s and 2000s: browser badges, site links, slogans, software logos, causes, jokes and the miniature advertisements that once covered the bottoms and sidebars of personal homepages.

The collection is especially valuable because Lucas says she rescued it from a now-defunct Webs.com page shortly before that site disappeared without warning.

That sentence is basically the entire digital-preservation problem in miniature.

Tiny images carried an entire social layer

An 88×31 button was more than decoration. On personal websites, buttons often worked as recommendations, identity markers and reciprocal links. A webmaster might display badges for Netscape, Linux, favorite software, a web host, a webring, friends’ sites, political views or whatever else deserved thirty-one vertical pixels of enthusiasm.

The famous “Netscape NOW” style is part of that history, but the format grew far beyond browser promotion. Personal-site communities used buttons as handmade social infrastructure. Two sites could trade buttons and become “affiliates” without an algorithm knowing or caring that either site existed.

Lucas’s collection therefore preserves more than graphic design. It preserves fragments of relationships and identities.

Her larger site, A.N. Lucas’s Web Lounge, fits the material perfectly. Established in 2013, it describes itself as her own little slice of the Web, with artwork, writing, collections and the permanent promise that it is “always under construction.” It even encourages visitors to copy and share its buttons.

This is not nostalgia performed from behind glass. It is a continuation of the behavior being archived.

The button collection became popular enough to circulate widely outside Neocities, which is interesting in itself. People still recognize these little rectangles because they represent a Web where linking to another person was an editorial decision made by a human being.

CacheRat’s 1,967 Ancient Web Domains research list exists for the same general reason: useful pieces of Web history vanish unless somebody notices them before the server goes dark.

Lucas noticed.

More than 700 tiny GIFs and images survived because one person copied them before another free-hosted page disappeared. That is not glamorous archival work, but twenty years from now it may be exactly the material somebody needs to reconstruct what a personal homepage actually felt like.

Explore A.N. Lucas’s surviving Web collections

Posted on by

Ancient Web: The Nuclear Weapon Archive Started as a 1994 University Site

Some old websites survived because nobody touched them. The Nuclear Weapon Archive survived because people kept having to move the whole thing somewhere else.

The Nuclear Weapon Archive is still online.

Its own history page traces the project back to November 1994, when Gary Au, then a student at Melbourne University in Australia, founded the High Energy Weapons Archive. That puts the site’s ancestry inside the first commercial-browser era of the Web, when a serious reference project could still begin as a directory on a university machine and become internationally known before anyone had settled on what a “web publication” was supposed to look like.

The archive did not enjoy a quiet childhood. Traffic surged around the fiftieth anniversary of the atomic bombing of Japan in 1995, drawing enough attention that Melbourne University shut the site down under its computing-use rules. Au later took a job with the Australian Department of Defence and handed administration to Carey Sublette in January 1996. From there the material bounced among hosts in Finland, Italy, and the United States, including the Federation of American Scientists and Enviroweb.

That history explains one of the site’s most important design decisions. In 2003 it finally acquired nuclearweaponarchive.org, specifically to stop years of host changes from breaking every incoming link. The old High Energy Weapons Archive name also became The Nuclear Weapon Archive.

What survives is enormous. There are historical sections on the Manhattan Project, Trinity, Hiroshima and Nagasaki, nuclear test programs, national arsenals, weapon-development history, reference material, photographs, timelines, and a long FAQ. The site also records its own update history instead of pretending it appeared fully formed yesterday. Entries jump from the early 2000s through later maintenance work, including updates in 2020 and 2023.

The important thing in 2026 is not that every page is visually ancient. It is that this is an example of a serious specialist reference site carrying its institutional memory with it. Modern sites are often rebuilt until their own history disappears. This one tells you who started it, who inherited it, why it moved, why the domain exists, and what each phase was trying to preserve.

That makes the archive itself part of the historical record, separate from the difficult subject matter it documents.

If you want to see how a Web reference project can survive repeated displacement without turning into a content-platform blob, go through the archive itself. CacheRat’s larger 1,967 Ancient Web Domains research list contains more places built with the same old assumption: if the information matters, put it on a page and keep the page alive.

Posted on by

Ancient Web: The Australian Space Academy Built a Space School in HTML

Some old websites look like somebody’s hobby page. The Australian Space Academy looks like somebody tried to build an entire small educational institution out of HTML.

Visit the Australian Space Academy

The site’s stated mission is to provide space education to a broad audience through the Web, public events, tours, space camps and research opportunities. Its pages describe proposed facilities, observation programs and educational goals, but the most immediately useful part is the material that actually made it online.

There is a lot of it.

The library contains concise notes and guides on subjects including aurorae, cosmic rays, the equation of time, Lagrange points, meteors, planetary temperatures, the ionosphere, radio astronomy, the Roche limit, variable stars, the moons of Mars, space law and the history of solar radio observations.

The classroom section goes even further. It lists in-house courses, self-study material, knowledge tests and an online space course. One “Scientific Observer” course was designed around precise observation, measurement and evaluation rather than merely reading facts off a page.

That is a very old-fashioned kind of science education, in the best sense.

The site expects curiosity to lead somewhere

A lot of modern science content is optimized to answer one question quickly. The Australian Space Academy was built more like a campus directory. Read about an astronomical topic, move into a reference page, test yourself, follow another section, then end up learning something you did not originally search for.

The visual design has not been sanded down into generic educational software either. It still feels like a small organization publishing its own materials directly to the Web. The homepage was still reporting an update in June 2026, which makes this more than a frozen historical curiosity.

Its longevity is important because niche scientific education sites can accumulate value slowly. A brief explainer written years ago may continue answering the same physical question because the Roche limit did not receive a disruptive startup update last Tuesday.

The site’s age also shows in its architecture. Subject indexes, lists of notes, references and manually linked course material do the organizational work that modern sites might assign to a database-backed learning platform. It is less seamless, but unusually transparent. You can see the curriculum simply by looking at the links.

CacheRat’s 1,967 Ancient Web Domains research list includes many pages like this where one institution or specialist quietly built something much larger than its web design initially suggests.

The Australian Space Academy is worth exploring because it treats the Web as what early educational publishers hoped it could be: a place where a curious person could arrive knowing almost nothing and keep following links until they knew considerably more.

Explore the Australian Space Academy

Posted on by

Ancient Web: A 1997 Klik Game Site Is Still Pointing at Executables

Before indie games meant storefront algorithms, launch trailers, Discord roadmaps, and patches measured in gigabytes, a game could be a one-megabyte .exe sitting on somebody’s personal site.

Visit the old K-Zero site

The K-Zero pages at 1997.nagash.net are a surviving pointer into that world. The site itself can be temperamental today, but game-preservation projects still identify it as the original home of small Windows games created with the Klik family of game-making tools.

Kliktopia, which catalogs old Klik & Play, The Games Factory, and Click & Create titles, traces several K-Zero releases back to the site. One is TAMAgirl SHE, estimated to date from 1997 and distributed as tamag.exe, about one megabyte in size. Another is an Aurora Gameplay Preview from roughly 1998, made with The Games Factory. Other K-Zero titles listed by preservation projects include DigKid, FunnyLand, and an Aurora intro.

That is a useful little fossil because Klik software occupied an important middle ground in 1990s game creation. You did not need to write a renderer, sound engine, input loop, and sprite system from scratch. Tools such as Klik & Play let ordinary users assemble playable Windows games visually and then hand the resulting executable to other people.

The Web was the distribution network.

A game release could be a filename and a hyperlink

There was no requirement for an account system, installer service, analytics SDK, launcher, or marketplace page. A creator made a game, uploaded it, wrote a few lines of HTML, and hoped somebody found it.

That simplicity created a preservation problem. When the personal site vanished, the game often vanished with it. There was no centralized library guaranteeing another copy existed.

The reason K-Zero can still be reconstructed in 2026 is that later collectors kept the files, filenames, tool versions, screenshots, estimated release dates, and original source URL together. Even when the original page itself becomes flaky, the chain of provenance remains visible.

That is exactly the sort of information digital preservation needs. A naked executable is interesting. An executable tied to its creator, original website, development tool, approximate release date, and neighboring projects is history.

CacheRat’s 1,967 Ancient Web Domains research list contains many personal software pages from this period. They are worth checking while the links still resolve, because a surprising amount of software history still survives one forgotten directory at a time.

Try the original K-Zero site

Posted on by

Ancient Web: PoopNames.com Gave an Entire Domain to One Stupid List

There was a period when somebody could think of one dumb joke, buy the exact domain for it, and apparently decide that was enough Internet for the day.

Visit PoopNames.com

PoopNames.com is exactly what it sounds like. Its page title is The Poop List, and its description promises a funny list of names for various types of poop. The domain dates to 2003, which puts it squarely in the era when novelty websites did not need an account system, content strategy, recommendation engine or venture capital pitch.

They needed a joke and HTML.

This is not an important reference work. That is why it belongs in Internet history.

The Web used to have room for tiny ideas

One of the easiest things to forget about the older Web is how many sites existed for no defensible reason beyond “somebody thought this would be funny.”

A person registered a domain for a single gag, made a list, emailed it to friends, got linked from message boards, and left it sitting there. The whole experience might take thirty seconds. That was fine. Not every page needed to become a business, community, app or personal brand.

The result was an Internet full of odd little landmarks.

PoopNames.com represents that culture almost perfectly because the premise does not become more sophisticated when you explain it. There is no hidden technical breakthrough. There is no grand archive buried three directories down. The joke is the site.

That simplicity also made novelty pages unusually memorable. You could tell somebody the domain verbally and they already understood why they should visit. Search engines were useful, but word of mouth, email chains, IRC, forums and instant messages were often enough to move a ridiculous URL around the Internet.

Modern social networks still produce endless disposable jokes, but the unit has changed. A post belongs to a platform and disappears into a feed. A standalone novelty site occupies its own address. Twenty years later, you can still point to the place where the joke lived.

CacheRat’s 1,967 Ancient Web Domains research list includes serious technical archives, museums, personal pages and a healthy number of sites that contribute absolutely nothing except evidence that the Web once tolerated glorious nonsense.

That last category deserves preservation too.

PoopNames.com is not profound. It does not need to be. Its historical lesson is simpler: once upon a time, one stupid list could justify an entire website.

Honestly, that was healthy.

See The Poop List at PoopNames.com