Category: Ancient Web

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

Posted on by

Ancient Web: Somebody Cataloged Every Descent Robot Like a Field Guide

Descent already had a memorable enemy roster. Insectoid’s old DOS-game site decided that was not enough and cataloged the robots as though somebody might need to identify one in the wild.

Browse the Descent robot guide

Each robot gets a miniature technical record: first level encountered, size in meters, weapon type, threat rating, point value, common drops and a short note about how badly it can ruin your day.

The Class 1 Drone is a low-threat nuisance. The Super Hulk is marked “Very Very High.” Boss #2 reaches the more scientifically rigorous category of “Insane.”

This is useful information.

A fan page with actual structure

The page is part of a much larger personal archive covering DOS games, Descent levels, weapons, Jazz Jackrabbit, XPilot and other material. A surviving change log shows years of redesigns, image recaptures, CSS rewrites and painstaking work on the Descent section.

In 2012, the robot images were improved and the author was still adapting old level pages. Earlier entries record the creation of walkthroughs for all 27 main levels and three secret levels, along with custom graphics and robot renders.

That long maintenance history is what makes the page interesting beyond nostalgia.

This was not a disposable “top 10 Descent enemies” article. It was a personally maintained reference system for a game somebody knew extremely well.

The guide even records unused material. The Split Pod, or Red Triangle, exists in the Descent engine but does not appear in the original First Strike campaign; fan-made levels could still use it.

For game historians, details like that are exactly the sort of small technical fact that disappears when fan pages vanish.

For webmasters, the page is another argument for boring HTML tables. They remain readable decades later.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

The Descent robot guide is not pretending to be definitive scholarship. It is better than that in one important way: it is clearly the product of somebody repeatedly playing the game, taking notes, capturing assets and correcting the page over time.

Return to Insectoid’s Descent robot catalog

Posted on by

Ancient Web: Cayzle Spent 27 Years Building a Website About Lion-Centaurs

Some personal sites are broad because their owners lose focus. Cayzle.com is broad because one person has spent more than a quarter century refusing to stop thinking about feline centaurs.

Visit Cayzle’s Wemic Site

The site launched on July 6, 1999. Its central subject is the wemic: essentially a lion-centaur, with a human-like upper body joined to a lion’s body. From there the rabbit hole expands into Dungeons & Dragons rules, mythology, ancient art, heraldry, fiction, character sheets, campaigns, webcomics and original tabletop game design.

There are pages on wemic culture and society, origins, D&D implementations across editions, art, fiction and character rules. There are also free RPG PDFs and a home-grown d20 game called Labyrinths & Liontaurs.

That would already be enough for one strange website.

It kept going.

A living fossil, not an abandoned one

The change log is the real historical document.

In 1999 Cayzle was confirming links, trading banners and expanding campaign pages. In 2000 the site joined webrings and reorganized itself. In 2001 it added prestige classes and third-edition D&D material. GeoCities later died, leaving broken links that had to be repaired.

The site resurfaced with a major redesign in 2013.

It is still being updated in 2026.

The current homepage still offers a guestbook, email list, link roll and multiple webring memberships alongside newer material. That mixture is rare: the old architecture was not replaced by a platform account. It was kept alive and extended.

Cayzle even restored a guestbook in 2026 after the previous hosting service disappeared years earlier.

For webmasters, that is useful evidence that a personally owned website can survive platform cycles if somebody cares enough to maintain it. For RPG researchers, the site preserves decades of homebrew rules and fan culture that would otherwise be scattered across dead forums and vanished hosts.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

There are easier ways to publish online than maintaining a handcrafted lion-centaur archive for 27 years.

Fortunately, Cayzle did not choose one of them.

Read Cayzle’s site history and latest material

Posted on by

Ancient Web: Gaby.de Kept CP/M and ZNODE51 Alive on the Web

Gaby.de looks like what happens when a working computer-history desk is left connected to the public web for a very long time.

Visit Gaby.de

The site centers on CP/M, the operating system closely associated with Digital Research and Gary Kildall, but it spreads outward into Z80 systems, old software, manuals, hardware and BBS history.

The CP/M Center includes downloads for multiple CP/M versions, tools for DOS, CP/M-86 material, FAQs, online manuals, links to original CP/M binaries and sources, the Digital Research Documentation Project, and information about moving CP/M data onto newer PCs.

Then there is ZNODE51.

A BBS survived as a web archive

ZNODE51 was a German Z-System and CP/M BBS. Historical CP/M FAQ material lists it as node 51 in Munich, and the machine behind it was an Altos 580.

Gaby.de’s own virtual museum records that Altos system as a circa-1983 multiuser computer with two Z80A processors, 64 to 192 KB of RAM, a 40 MB hard drive and support for CP/M 2.2, MP/M-80 and OASIS. It had been used as Z-Node 51 until 1992.

The archive now exposes roughly 150 MB of CP/M downloads and related tools.

That is not enormous by modern storage standards. It is enormous in historical context.

The site also keeps a virtual computer museum after the physical private collection was dissolved. There are meeting photos, Z-Fest material, Windows 3.1x resources, Z80 links and old Computer Journal articles.

For retrocomputing researchers, Gaby.de is useful because it preserves relationships between machines, software, user groups and BBS culture instead of reducing everything to isolated downloads.

For webmasters, the lesson is simpler: a small information architecture can outlive the hardware it describes.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

The computers moved into history. The BBS shut down. The private collection was dispersed.

The pages stayed up.

Browse Gaby.de’s CP/M Center

Posted on by

Ancient Web: The Cyberpunk Project Was a Library for the Future Before the Future Arrived

Before cyberpunk became a giant entertainment label, people on the web were trying to collect the culture while it was still close enough to touch. The Cyberpunk Project was one of the places doing it.

Visit the original Cyberpunk Project address

The site is no longer reliably reachable, but its footprint is everywhere.

Old descriptions characterize it as a database, library and information project dedicated to cyberpunk. Researchers and later websites cite individual pages for cyberpunk manifestos, cyberculture essays, postcyberpunk writing and historical texts. A 1999 Russian film journal described the project as one of the most substantial cyberpunk sites in the Russian web sphere, despite the project itself publishing in English.

The Internet Archive has preserved versions of the site, and even Wikipedia’s cyberpunk reference material has pointed readers toward an archived copy of the Cyberpunk Project.

That is the strange second life of a serious old website: the domain can fail while its citations keep walking around.

An archive that became source material

One of the project’s best-known preserved texts is Christian As. Kirtchev’s A Cyberpunk Manifesto, dated 1997. Other researchers have cited the site for Lawrence Person’s Notes Toward a Postcyberpunk Manifesto and David Silver’s Introducing Cyberculture.

The archive also carried material reaching beyond fiction into hacker culture, digital identity, futurism, art and the politics of networked life.

That broad scope matters because “cyberpunk” on the early web was not merely an aesthetic of neon cities and expensive jackets. It overlapped with real discussions about information access, privacy, networks, cryptography, identity and technological power.

For researchers, the project is useful partly because later academic work still points back to it.

For web historians, it is a warning. Important reference sites can become unavailable even while their URLs remain embedded in papers, bibliographies and old link lists.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

The original server may be unreliable now, but the Cyberpunk Project succeeded at one thing that matters: it collected enough material that people kept citing it after the web had moved on.

Open an archived copy of the Cyberpunk Project

Posted on by

Ancient Web: Takedown.com Preserves the Web Version of the Kevin Mitnick Hunt

The Kevin Mitnick story has been retold so many times that it is easy to forget some of the case still sits online in the kind of website people built while the events were recent.

Visit Takedown.com

The site is tied to the pursuit of Mitnick by computer-security researcher Tsutomu Shimomura and the reporting around that pursuit. Its surviving sections include biographies of Shimomura, Mitnick and journalist John Markoff, along with timelines, newspaper coverage, photographs, voicemail material and records of online sessions.

That makes it more interesting than a retrospective article written decades later.

It is an artifact from the argument itself.

When network intrusion became public drama

The case accelerated after a break-in at Shimomura’s computer systems. Files later appeared on The WELL, the influential online community, and network traffic was monitored while investigators tried to trace the intruder.

Contemporary coverage framed Shimomura as a high-tech detective following packets and captured sessions through the network. Mitnick was arrested in 1995.

The case became one of the defining hacker stories of the decade, and it was controversial almost immediately. Mitnick’s supporters, journalists, security researchers and law enforcement did not all tell the same story.

That is exactly why Takedown.com is useful.

A historical researcher should not treat one participant’s archive as neutral truth. But primary-source collections do not need to be neutral to be valuable. They need to be identifiable, preserved and read with context.

The site shows how the Shimomura side of the story organized its evidence and narrative for the public web.

For Internet historians, it also captures an era when technical incidents were becoming mass-media events but the source material could still be published directly on a standalone site.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

Whatever conclusion someone reaches about the Mitnick case, the old site is worth keeping in view because it preserves how the story was documented online at the time.

Browse Takedown.com’s biographies and case material

Posted on by

Ancient Web: Jim Storer Built an Online Field Guide to Mechanical Puzzles

A mechanical puzzle collection becomes much more useful when somebody measures it, photographs it, records the designer, explains the goal, compares known solutions and then keeps doing that hundreds of times.

Visit Jim Storer’s Puzzles Home Page

James A. Storer’s site is enormous.

Its index runs through burr puzzles, sliding-block puzzles, packing problems, interlocking assemblies, Rubik-style twisty puzzles, dexterity puzzles, disentanglements and objects that are difficult to classify without first taking them apart.

Individual pages are not just photographs.

A typical entry may identify the designer and approximate date, describe materials and dimensions, explain the objective, cite books or other collectors, and then document known solutions in detail.

For the puzzle called Trap, for example, the site records a 1986 Minoru Abe design, a later Storer-built version, the piece layout, movement rules and multiple solution counts. Another page compares Quzzle and Quzzle Killer down to minimal move counts.

This is collector behavior upgraded into documentation.

The useful kind of obsessive

Storer is a computer scientist, and the site reflects that temperament. Mechanical puzzles are physical objects, but many of them are also search problems: states, legal moves, shortest solutions and proofs that a better route does or does not exist.

That makes the collection useful to several different audiences at once.

Puzzle collectors can identify objects and designers. Mathematically inclined readers can study solution structure. Historians get dates and references. Makers can inspect mechanisms. Casual visitors can click “six fun challenges” and lose an afternoon.

The design is plain, but the information density is exceptional.

For webmasters, it is another example of a specialist site winning by organizing material that general search cannot synthesize cleanly.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

There are prettier puzzle sites.

Few feel as much like somebody opened a private research notebook and left the index public.

Browse the full mechanical-puzzle collection

Posted on by

Ancient Web: Ginjirotchi.ca Keeps Tamagotchi Care Logs on the Open Web

A Tamagotchi is a tiny device built around routines: feed it, clean it, watch it grow, screw something up, remember the result. Ginjirotchi.ca turned those routines into a public record.

Visit Ginjirotchi.ca

The site is part of the old Tamagotchi fan-web ecosystem: individual collectors documenting shells, characters, growth paths, care experiments and day-to-day runs long before every hobby was organized inside a social app.

Other Tamagotchi preservation sites still point people toward Ginjirotchi.ca specifically for its logs, photographs, journal entries and information on vintage models.

That personal documentation has become more valuable as the hardware ages.

Care logs are historical data

Virtual pets are easy to dismiss as toys, but old Tamagotchi models can be surprisingly opaque.

Behavior depends on feeding, discipline, mistakes, sleep schedules, age and model-specific rules. Fans learned by running the devices repeatedly and comparing results.

Ginjirotchi.ca belongs to that tradition.

The site’s owner, Dennis, is also visible in the broader Tamagotchi community. Other long-running collectors refer to group hatchings with him and to care methods he developed while trying to reach particular characters. A 2025 log on Mimitchi.com even mentions his experience with older Japanese Tamagotchis unexpectedly resetting.

Those cross-links matter. They show an old-fashioned fan network where knowledge moved between individually owned sites rather than one central platform.

For collectors today, pages like these can help identify models, understand character evolution and see how people actually used the devices when they were less distant from their original era.

For web historians, the more interesting artifact may be the format itself: a hobby journal that accumulated practical knowledge because somebody kept publishing what happened.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

A virtual pet eventually dies, resets or gets put in a drawer.

A good care log can outlive all three.

Return to Ginjirotchi.ca’s Tamagotchi pages

Posted on by

Ancient Web: The Merkur Encyclopedia Keeps Ford’s Strangest American Import Repairable

The Merkur brand barely survived the 1980s. The Merkur Encyclopedia has spent much longer helping the cars survive the decades afterward.

Visit the Merkur Encyclopedia

Ford sold two Merkur models in North America: the XR4Ti and Scorpio. Both were European-derived cars routed through Lincoln-Mercury dealers, and both disappeared by the end of the 1989 model year.

That short production run created exactly the sort of ownership problem specialist websites are good at solving.

Parts knowledge becomes tribal. Wiring details get weird. Interchange information matters. Factory assumptions stop matching what is available thirty years later.

The Merkur Encyclopedia collects that accumulated owner knowledge.

The web as a repair manual supplement

Surviving pages include technical material on Ford’s EEC-IV engine-control systems, ECU calibrations, injectors, vane-air meters, horsepower figures, lighting and model-specific parts.

The tone is often wonderfully direct. An ECU page warns readers not to combine certain XR computers with a large vane meter because the engine can run lean and damage pistons or head gaskets. Another page is simply a printable XR4Ti bulb chart.

Old Merkur Club discussions repeatedly point owners back toward the Encyclopedia for technical tips and model information.

Even Motor Trend listed the site in a Merkur XR4Ti buyer’s guide and credited Allan Slocum’s Merkur Encyclopedia to 2002.

That is the kind of recognition specialist sites earn slowly: not by having the biggest audience, but by being where owners find the weird answer.

For automotive historians, the site preserves knowledge around an unusually short-lived Ford experiment. For current owners, its value is more practical: obscure cars remain usable only while documentation, parts knowledge and repair experience remain accessible.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

The XR4Ti may be forgotten by most drivers.

Its bulb chart is still online.

That is preservation at the level where machines actually keep running.

Browse a Merkur Encyclopedia technical page

Posted on by

Ancient Web: Rechenmaschinen-Illustrated Put the History of Calculating Machines Back Online

A book about mechanical calculators becomes far more dangerous to your free time when somebody puts it online and starts adding pictures.

Visit Rechenmaschinen-Illustrated

The core of Rechenmaschinen-Illustrated is Ernst Martin’s The Calculating Machines, a major historical reference on mechanical calculation. The site presents an online version and extends it with illustrations, articles and additional information.

That extension matters because many machines described in old technical literature are difficult to understand from names and prose alone.

Mechanical calculators are physical mechanisms. Cranks, pinwheels, registers, stepped drums, carry systems and carriage movements make much more sense when the reader can see the hardware being discussed.

A reference used by other references

The Smithsonian’s National Museum of American History points researchers toward Rechenmaschinen-Illustrated as an online version of Martin’s book with added illustrations.

Other calculator collectors do the same.

That gives the site an unusual place in the old web: a private or specialist resource that became useful enough to be cited by institutional collections.

The material reaches into companies and machines that rarely appear in general computer-history writing: Brunsviga, Odhner, Facit, Triumphator, Lipsia and many more.

For researchers, serial numbers, model variations and manufacturing timelines can matter as much as the famous inventions. Collector sites often preserve exactly those details because collectors need them to identify what is physically sitting on a table.

The design itself also qualifies as an artifact. The homepage still announces that it uses frames. A modern browser can complain; the information underneath remains worth reaching.

For webmasters, that is a useful contrast. Presentation ages quickly. Structured specialist knowledge ages much more slowly.

This site was found in CacheRat’s broader Ancient Web research corpus.

Explore the 1,967 Ancient Web Domains research list

Mechanical calculators were eventually replaced by electronics.

Fortunately, the people obsessed with the old machines did not stop documenting them.

Open Rechenmaschinen-Illustrated’s calculator history

Posted on by

Ancient Web: An Atlas of the Universe Zooms from Nearby Stars to the Edge of Visibility

Most astronomy pages give you an object. Richard Powell’s Atlas of the Universe gives you the problem of scale.

Visit An Atlas of the Universe

The site was created around 2000 and is organized as a sequence of maps that keep pulling farther away. One starts with the nearest stars. Then the field widens to the local stellar neighborhood, the Milky Way, nearby galaxies, galaxy groups, superclusters, and eventually the visible universe.

That sounds simple until you try to build it.

The useful trick is the sequence

Space is hard to understand because the numbers get stupid very quickly. A light-year is already inconvenient. A million light-years is worse. Then somebody says “a billion” and most brains quietly replace the whole subject with decorative stars.

Powell’s approach is to make each scale a handoff to the next one.

The maps are not trying to be cinematic. They are diagrams. Labels, distances, positions, explanatory text, and related maps do the work. The result feels closer to a good technical atlas than a modern astronomy portal.

A catalog record from the American Association of Physics Teachers describes nine main maps ranging from the nearest stars to the limits of the visible universe, with printable images, a glossary, and related links. The project has also circulated in several languages.

For a webmaster, the site is a good reminder that information architecture can be the feature. The underlying subject is enormous, but the navigation answers one question over and over: what is the next larger thing that contains this?

That is enough to turn cosmology into something you can browse.

There is also something pleasantly Web-native about it. No account. No scrolling animation pretending you are flying through space. No forced video. You choose a scale, read it, then move outward or inward.

This site came out of CacheRat’s larger Ancient Web research corpus, which is full of old specialist sites built around exactly this kind of focused usefulness.

Explore the 1,967 Ancient Web Domains research list

The universe did not get any smaller. Powell just found a way to put useful handles on it.

Return to An Atlas of the Universe