Category: Ancient Web

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

Posted on by

Ancient Web: Crop Circle Center Cataloged Over 7,100 Sightings Without Settling the Argument

You do not have to believe crop circles are mysterious to recognize that somebody has done a ridiculous amount of indexing.

Visit Crop Circle Center

Crop Circle Center says it has been operating for about twenty years and currently lists more than 7,100 crop-circle sightings. Its stated goal is unusually useful: record the sightings regardless of who or what created them.

That distinction matters.

The web has always been good at turning disputed subjects into either evangelism or ridicule. An archive can take a third route and simply preserve the reports.

Crop Circle Center organizes formations by year and date, with location, country, crop information and images where available. Its historical statistics reach back through the 1990s and earlier decades, while current-year pages continue to receive new entries.

Whether a formation was made by artists, pranksters, landowners, enthusiasts, unknown actors or something someone would like to discuss after three hours of UFO documentaries is a separate question.

The archive records that the thing was reported.

Catalog first, argue second

This is a useful preservation principle well beyond crop circles.

Internet history is full of subjects where later researchers need the raw chronology more than somebody else’s conclusion. When did a claim appear? Where? What image was circulating? How was the event described at the time? What patterns emerge across years?

Those questions become harder when original pages vanish and only secondary summaries remain.

The Crop Circle Center currently says it is rebuilding the site into a modern searchable database and is seeking funding to keep the project online through 2030. That is also a familiar independent-web problem: a large specialist archive can represent decades of labor while still depending on one person paying hosting and development bills.

CacheRat’s 1,967 Ancient Web Domains research list exists because obscure specialist sites often contain exactly this kind of concentrated work. They may cover subjects institutions ignore, mock or simply never bother to catalogue.

The interesting artifact here is not proof of aliens.

It is proof that somebody kept a spreadsheet-sized obsession alive on the open web for twenty years and turned it into thousands of browsable records.

That is valuable even if every circle was made by two guys, a board and a rope.

Browse the Crop Circle Center archive

Posted on by

Ancient Web: ToastyTech Never Stopped Making Doom WADs

Some people played Doom in 1993 and moved on. The person behind ToastyTech apparently interpreted Doom’s release as a long-term maintenance obligation.

Visit ToastyTech’s Doom pages

The page opens with a simple explanation: id Software released Doom on December 10, 1993, and “I never stopped playing.” What follows is less a polished portfolio than a fossil bed of Doom enthusiasm—custom levels, screenshots, old versions, experiments, joke material, music files, and work-in-progress archives accumulated over years.

MarsWar is the obvious centerpiece. It is a Doom II megawad with its own story about a Martian terraformer, an invasion from Earth, and space-warp ships built by a company whose initials happen to be “MS.” The site keeps multiple versions available, including older builds because people asked for them. That small detail says a lot about the culture around old game mods: history survives because somebody decides an obsolete ZIP is still worth leaving online.

Elsewhere there is a Doom 1 MegaWad compilation, a Hexen level, Skulltag and jDoom material, Doom Alpha levels modified for deathmatch, original MOD and MIDI files, Ranma 1/2 character skins, unfinished maps, and even a Firefox-themed WAD. There are also screenshots from very early Doom versions and a “Doom Guys Vacation” section whose description is simply: went to hell, took pictures, posted them here.

The important thing is not that every file is a masterpiece. The page preserves how PC game modding actually looked when personal websites were the distribution system. A ZIP file sat next to a TXT file. Screenshots were separate. Compatibility notes mattered. Old versions stayed reachable. The author’s own jokes, experiments, and unfinished work remained mixed in with the serious releases.

Modern mod sites are much better at indexing and downloading. They are often worse at preserving the creator’s original neighborhood around the files. ToastyTech still has that neighborhood.

For researchers studying Doom, homebrew level design, WAD distribution, early fan culture, or the long tail of DOS-era games, these personal pages can be more revealing than a clean database entry. They show what the creator thought belonged together.

That is the same reason CacheRat maintains the 1,967 Ancient Web Domains research list: sometimes the useful historical unit is not the file. It is the whole weird page around it.

Doom is still alive almost everywhere. ToastyTech is interesting because it preserves one person’s uninterrupted relationship with it.

Dig through ToastyTech’s Doom stuff

Posted on by

Ancient Web: RailArchive Preserves Railroad History From the Photographer’s Own Negatives

Richard Leonard’s Rail Archive is what happens when somebody who was actually standing beside the tracks in the 1950s later gets a website. It is not just a collection of old train pictures. A substantial part of the archive comes from Leonard’s own photographs and those of his brother.

Visit Richard Leonard’s Rail Archive

Leonard says he began developing rail-related websites in 1998, starting with the Steam Locomotive Archive. The current collection contains more than 1,800 images covering steam, diesel, and electric locomotives along with other pieces of North American railroad history.

The personal material is the obvious draw. There are photographs of steam locomotives operating in the 1950s, including Chicago, Burlington & Quincy, Canadian Pacific, Grand Trunk Western, Illinois Central, Nickel Plate, New York Central, and Union Pacific power. A separate gallery preserves photographs Leonard’s brother David took during trips through Michigan and Ontario in 1958 as regular steam operation was disappearing.

Then the archive keeps going.

There are Union Pacific Big Boys and Challengers photographed in Wyoming and Nebraska in 1957, gas-turbine locomotives, vintage diesels, electric rail material, New York Central collections, and even a “Fantasy Steam” section devoted to locomotives that never existed.

The document scans may be even more useful to researchers. Leonard has digitized period railroad publications including a 1944 training course on locomotive firing, a 1949 Pennsylvania Railroad motive-power booklet, a 1948 Westinghouse publication discussing steam, diesel, electric, steam-turbine, and gas-turbine power, a 1913 American Locomotive Company bulletin, and guidebooks from the Chicago Railroad Fair.

This is the old Web doing what it did particularly well: one person builds an index around material that would otherwise remain scattered through drawers, private collections, old paper publications, and aging negatives. The site’s design does not try to compete with the photographs. It mostly gets out of the way and lets the collection keep growing.

The archive is valuable because it mixes several levels of evidence. You have firsthand photographs, family photographs, contributed images, promotional literature, technical publications, and Leonard’s commentary sitting together. That makes it more useful than a gallery stripped of provenance.

Researchers hunting this kind of material are exactly why CacheRat keeps the 1,967 Ancient Web Domains research list. Sites like RailArchive often contain the useful thing three clicks below a homepage that search engines barely describe.

The Web did not create these photographs. It gave them somewhere to remain findable after the railroad world they documented had changed beyond recognition.

Explore Richard Leonard’s Rail Archive

Posted on by

Ancient Web: Jan Sikusi’s Wonderdream Gives Every Interest Its Own Room

The best personal websites do not answer the question “What is your niche?” They answer “What are you into?”

Visit Jan Sikusi’s Wonderdream

Jan Sikusi’s homepage is a compact directory of interests that would make a social-media strategist nervous.

There is Lightspeed Lattice, Bolognese sword fighting, a Toki Pona font collection, a Museum of Peculiar Artifacts, pages about writing English and Hebrew in cuneiform, a chess adventure through time, a game called Safari Steve, a page about the owner’s Shih Tzu, and a guestbook.

Sikusi says the main interests at the moment are swordfighting, Latin, Toki Pona and chess. By trade, the author is a software engineer, but the page makes a point of saying that software engineering is not supposed to become the entire identity.

That sentence explains the website better than a design manifesto could.

Interests do not need a content strategy

Platform profiles encourage specialization because specialization is easy to categorize. The woodworking account does woodworking. The chess account does chess. The programmer account talks about programming until everyone involved would rather chew drywall.

A personal homepage has no such obligation.

The Jan Sikusi site can place sword fighting next to constructed languages and cuneiform because the organizing principle is simply one person. That was normal on the early web. A homepage was not a publication vertical. It was a map of somebody’s brain.

The structure helps. Instead of one endless chronological feed, the homepage divides material into sections such as games, passions, funny bits and personal pages. Visitors choose a door rather than waiting for an algorithm to decide which fragment deserves visibility today.

That is one reason CacheRat keeps modern handmade sites beside older material in the 1,967 Ancient Web Domains research list. The old web was not important because every page was built in 1997. It was important because people controlled the shape of their own publishing space.

Neocities gave that idea a new place to live, and sites like this actually use it.

There is no obvious conversion funnel here. No carefully constrained brand voice. No attempt to explain why cuneiform and a Shih Tzu belong together.

They belong together because Jan put them there.

That used to be enough reason to make a webpage.

Explore Jan Sikusi’s Wonderdream

Posted on by

Ancient Web: KnotPlot Has Been Turning Mathematics Into 3D Knots Since the 1990s

KnotPlot looks like somebody gave topology a graphics workstation and then refused to stop experimenting for thirty years. It is part mathematical reference, part software project, part image gallery, and part personal laboratory.

Visit the KnotPlot Site

Robert Scharein built KnotPlot out of work connected to his PhD at the University of British Columbia’s Imager Computer Graphics Laboratory. The software visualizes and manipulates mathematical knots and links in three and four dimensions, and the site says almost all of its images were generated with KnotPlot itself.

That immediately makes the website more interesting than a normal gallery. The pictures are output from the subject being documented. Decorative knots, Brunnian links, torus knots, knot carpets, stereoscopic images, cubic-lattice knots, hyperbolic knots, Celtic knots, Moebius-strip knots, and stranger constructions sit alongside thesis material and technical references.

The software side is just as substantial. KnotPlot can load a database of more than 3,000 knots and links, let the user sketch knots in 3D, construct them with Conway notation, use a tangle calculator, generate special knot families, and transform existing knots. Current builds still exist for Windows, macOS, and Linux, which is remarkable enough for a program whose copyright line reaches back to 1992.

The site also retains unmistakable old-Web texture. Some galleries are polished and useful. Others are openly filed under a section titled “Pages with little or no interesting content,” because apparently deletion is for cowards. There are PostScript examples, POV-Ray output, thesis figures, stereoscopic experiments, and pages whose main justification seems to be that Scharein once wanted to see what a particular knot would look like.

That accumulation is precisely why the site matters. Modern software documentation tends to explain features. KnotPlot preserves years of intellectual wandering around the features: experiments, pictures, weird side paths, references, obsolete rendering techniques, and the research lineage behind the program.

For somebody studying mathematical visualization, early Web graphics, scientific personal sites, OpenGL software, or just the visual culture around topology, it is an unusually dense primary source.

It also demonstrates one of the Ancient Web project’s recurring lessons: a specialist website becomes valuable by refusing to flatten itself. It does not need to turn every experiment into a clean product page. The mess contains history.

More sites like this turn up in CacheRat’s 1,967 Ancient Web Domains research list, but KnotPlot stands out because the underlying software is still alive while the website continues to carry decades of accumulated research around it.

Explore KnotPlot’s galleries and software

Posted on by

Ancient Web: Ed Thelen Turned Nike Missile Service Into a Public Technical Archive

Some personal homepages accidentally become archives. Ed Thelen’s site appears to have done it on purpose for about three decades.

Visit Ed Thelen’s Nike Missile site

Thelen wrote that he started the website in 1996. He had served in the Nike missile program in the Chicago defense area and later worked as an electronics and computer engineer. Instead of reducing that experience to a short biography, he built outward from it.

The result became a large technical and historical resource on Nike Ajax, Nike Hercules, radar, computers, missile sites, manuals, people, photographs and operating history.

It is exactly the sort of site the old web was unusually good at producing: one knowledgeable person deciding that an entire subject deserved a home.

Firsthand knowledge became infrastructure

The value is not just in the technical pages. The site also preserved correspondence and veteran memory. One old page discusses ways former Nike personnel might find one another online, including guestbooks and service-location lists. That is a reminder that before social networks centralized identity, specialist sites often served as the social graph for their own communities.

Thelen died in 2024, but the site did not simply freeze and rot. Its current stewardship page says Gregory Brown took responsibility for preserving the original material while continuing to add relevant Nike history. It specifically describes ongoing work involving U.S. Army Nike technical manuals held by the National Park Service archives at Golden Gate National Recreation Area.

That succession matters.

Independent websites often fail at exactly this point. The author dies, hosting expires, passwords disappear, and twenty years of accumulated specialist knowledge vanishes because nobody knew who was supposed to keep the lights on.

Ed-Thelen.org now explicitly has a preservation philosophy: keep the original work intact, make corrections conservatively, identify additions, and maintain public access.

CacheRat’s 1,967 Ancient Web Domains research list contains plenty of sites whose value is obvious only after you realize there may be no institutional substitute. This is one of those cases.

A veteran with engineering knowledge built a public reference site in 1996. Other veterans and researchers contributed. Decades later, somebody else stepped forward to steward it.

That is how a homepage becomes historical infrastructure.

Explore the Ed Thelen archive

Posted on by

Ancient Web: ConcordeSST Became a Living Technical Museum Before the Jets Retired

Most aircraft-history sites are built after the aircraft becomes history. ConcordeSST is more interesting because much of it was being assembled while Concorde was still operating, then kept growing through the fleet’s final years and retirement.

Visit ConcordeSST

The site describes itself as an attempt to bring the Concorde story together in one place, and it takes that job literally. There are fleet pages, technical specifications, historical timelines, cockpit material, accident documentation, photographs, development history, airline information, and pages explaining where the surviving aircraft eventually ended up.

One of the best parts is the history section. It reaches back before passenger service and into the design problem itself: Britain and France agreeing in 1962 to develop a supersonic airliner, the choice of a slender delta wing, materials testing, wind-tunnel work, flight development, and the decision to stay around Mach 2 rather than chase the much hotter structural problems of Mach 3.

The site also preserves the kind of detail that tends to disappear when history gets compressed into a paragraph. Its early-development material discusses structural heat, aluminum-alloy limits, fatigue testing, hydraulic rigs, engine development, and the enormous ground-test equipment required to prove the aircraft. Concorde was a glamorous airplane, but this site spends plenty of time on the ugly machinery needed to make glamour survive certification.

The change log is a historical document in its own right. Entries from 2002 and 2003 track updates while the real aircraft were still in service: fleet hours, return-to-flight work, the retirement announcement, final tours, museum destinations, and the last flights. The page eventually starts recording aircraft status as “retired.” You can watch an enthusiast website cross the line from current-events reporting into preservation without anybody redesigning the site around that fact.

That is why ConcordeSST is useful in 2026. A modern article can summarize Concorde. This site preserves the layers: what enthusiasts were tracking before retirement, what was added while events were happening, and which technical questions mattered enough to earn their own pages.

For web archaeology, it is also a reminder that specialist sites often become archives accidentally. Somebody starts documenting the thing they care about. Twenty years later, the thing is gone and the website has become part of the evidence.

CacheRat’s larger 1,967 Ancient Web Domains research list contains many sites with the same accidental-museum quality. ConcordeSST is one of the stronger examples because the transition happened in public.

The aircraft stopped carrying passengers in 2003. The old site kept the technical notes, timelines, photos, and contemporary updates online.

Explore the ConcordeSST archive

Posted on by

Ancient Web: Old Bricks Turned Stamped Clay Into an Industrial History Database

A brick with a name stamped into it looks like construction debris until somebody knows how to read the name. Brocross hosts one of the Web’s great examples of turning that kind of ordinary object into a serious historical index.

Visit David Kitching’s Brocross site

David Kitching’s homepage covers several interests: Cheshire dragonflies, canals, industrial archaeology, coal mining, limekilns and local history. The monster hiding inside it is Old Bricks: History at Your Feet, a site devoted to named British bricks and the works that produced them.

Old Bricks recently passed 10,000 brick images. That number is absurd until you understand the problem the collection solves. Brickworks opened, changed names, merged and disappeared. Clay pits were filled. Factories were demolished. Company records could be thin or nonexistent. A stamped brick may be one of the few durable objects still carrying the maker’s name.

The site organizes examples geographically and alphabetically, with separate sections for England, Wales, Scotland, Ireland, foreign bricks, mystery marks, quarry tiles, roof tiles and other related material. Entries often combine photographs with trade-directory dates, works locations, ownership changes and contributor notes.

The mystery section is especially good Internet archaeology. People submit bricks whose initials or marks remain unidentified, and the page accumulates partial clues from finds in canals, excavations, beaches, gardens and demolition sites. One unknown brick pulled from deep water off Spain sits comfortably beside examples recovered from Manchester excavations and Thames mud.

This is not merely collecting for collecting’s sake. Archaeological reports, industrial-history researchers and local-history projects cite Brocross because the database can help connect a stamped object to a manufacturer, place and period. The main Brocross site also preserves extensive material on Poynton’s coal-mining history and other industrial sites.

There is a useful Web lesson here. Community contributions did not require inventing a social network. Kitching built a specialist reference, gave people a way to submit evidence, credited contributors and let the database become more valuable with every new object.

CacheRat’s 1,967 Ancient Web Domains research list contains many such projects: databases that grew because somebody cared about a category the mainstream Web barely noticed.

Brocross is still active, and Old Bricks is still adding names. Apparently ten thousand bricks was not enough bricks.

Explore Brocross and Old Bricks

Posted on by

Ancient Web: APOD Kept a Plain HTML Window on the Universe Since 1995

One of the oldest continuously useful things on the web is also one of the least complicated.

Open the Astronomy Picture of the Day archive

Astronomy Picture of the Day launched on June 16, 1995. Robert Nemiroff and Jerry Bonnell built the project around a brutally durable format: pick an astronomical image, explain it clearly, publish a page, repeat.

Three decades later, the full archive is still basically a gigantic chronological list of links.

That is beautiful engineering.

The archive reaches back to Neutron Star Earth on June 16, 1995. Scroll forward and the history of web-era astronomy starts accumulating one ordinary page at a time: planets, galaxies, spacecraft, eclipses, supernova remnants, observatories, simulations and whatever else the sky happened to provide.

NASA describes APOD as the largest collection of annotated astronomical images on the internet. The important word there may be annotated. The project did not merely collect pictures. It attached human explanation to them.

The archive is the interface

Modern media archives often hide old material behind search boxes, JavaScript applications, recommendation systems or APIs. APOD’s full archive does something almost offensively simpler.

It lists the dates and titles.

Want July 1, 1995? Click it. Want something from 2006? Scroll. Want yesterday? It is near the top. The page is huge, but it is legible to humans, crawlers, scripts and future archaeologists without requiring much interpretation.

There is even a warning that some versions of Chrome may freeze while pre-fetching all those links. That is a very funny failure mode for a page whose fundamental architecture is otherwise nearly indestructible.

The archive also demonstrates why boring URLs and boring HTML age so well. CacheRat’s 1,967 Ancient Web Domains research list is full of ambitious sites whose navigation died before their content did. APOD largely avoided that trap by keeping the relationship between date, title and page obvious.

There are newer interfaces, mirrors and apps around APOD now. Fine.

But the old archive page still works because it never tried to become an experience platform.

It became a list.

And the list became thirty-one years of astronomy history.

Dig through the full APOD archive

Posted on by

Ancient Web: The Pleiadian Hoax Preserved an Old UFO Argument in Place

One of the more useful things an old website can preserve is not a conclusion but an argument. Mercury Rapids still hosts a 2006 article attacking the famous Billy Meier UFO photographs, followed on the same page by a lengthy response from one of Meier’s advocates.

Read The Pleiadian Hoax

Meier’s case became famous because the photographs looked unusually clear compared with the distant lights and blurry shapes common to UFO imagery. The page summarizes his claims of contact with beings he called Plejarans and then walks through evidence that critics said showed models, borrowed imagery and staged photographs.

The article is not coy about its position. It argues that famous images were faked and points to examples it says came from existing astronomical pictures, paintings, television footage, catalog imagery and small physical models. It also discusses testimony attributed to Meier’s former wife and older photo-analysis disputes.

Then the page does something that makes it more interesting as a Web artifact: it keeps the reply.

Michael Horn, identified there as an authorized American media representative for Meier, responds at length and contests the article point by point. He challenges the critics, cites other investigations and argues that the evidence has been misrepresented. The result is messy, partisan and considerably more useful than a page that quietly replaced the original article after receiving objections.

For a researcher, that preserved disagreement matters. Early Web pages often acted like public filing cabinets for controversies. Claims, rebuttals, links and personal accusations could remain sitting beside one another for years. You can see not only what people believed, but how they argued and what evidence they thought would persuade readers at the time.

Nothing about surviving on an old website makes either side correct. That distinction is especially important with extraordinary claims. The historical value is in the record of the debate, not in treating the page as scientific confirmation of aliens or definitive proof of every skeptical allegation.

The larger 1,967 Ancient Web Domains research list contains many pages whose value lies in this kind of context: old disputes frozen before social platforms compressed them into screenshots and disappearing threads.

The Pleiadian Hoax is still worth reading because the disagreement was left intact.

Read the article and response together