Category: Ancient Web

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

Posted on by

Ancient Web: The Millennium Simulation Put 10 Billion Particles on a 2005 Web Page

In 2005, one of the largest simulations of the universe did not arrive behind an app, a dashboard, or a mandatory account.

It arrived as a web page with links to enormous files.

Visit the Millennium Simulation Project

The Millennium Run used more than 10 billion particles to model how matter evolved inside a cube of simulated universe more than 2 billion light-years on a side. According to the project page, the calculation occupied the principal supercomputer at the Max Planck Society’s Supercomputing Centre in Garching for more than a month.

The output was about 25 terabytes.

That was then used to reconstruct evolutionary histories for roughly 20 million galaxies and the supermassive black holes associated with quasars.

The remarkable part for web archaeology is how directly the researchers exposed the result.

There are DivX movies. There are 1024Ă—768 and 2048Ă—1536 images. There are dark-matter slices at different redshifts. There is an A0 PostScript poster weighing in at 280 MB. There are galaxy catalogues offered as binary structures and plain ASCII, including a roughly 980 MB text version.

No infinite-scroll explainer. No animated marketing layer telling you that science is exciting.

Here is the simulated universe. The giant file is over there. Try not to kill your modem.

The web as a scientific loading dock

The page is a good reminder that the old web was not merely personal homepages and animated GIFs. It was also infrastructure for serious research.

Scientists could publish the paper, then use the web as a warehouse for the surrounding evidence: visualizations, catalogues, movies, references, and raw-ish data products that could never fit into a journal article.

Some of the file formats now look ancient. The DivX downloads especially feel like sediment from another computing era. But the architecture is remarkably durable because the important parts are obvious URLs pointing at understandable files.

That is one reason pages like this survive so well. There is relatively little machinery between the reader and the artifact.

CacheRat’s 1,967 Ancient Web Domains research list exists to surface exactly this kind of thing: not merely old design, but old information architecture that still exposes something valuable.

The Millennium Simulation page is practically a fossilized scientific distribution system.

And somewhere in that 980 MB ASCII catalogue are millions of simulated galaxies waiting for somebody with grep, disk space, and poor impulse control.

Explore the Millennium Simulation Project

Posted on by

Ancient Web: HuecoTanks.com Preserved Climbing History in Windows Media and QuickTime

HuecoTanks.com looks less like an official visitor guide than a long-running argument conducted by climbers who happened to preserve an enormous amount of history along the way. The site dates its presence from 1999 and explicitly says it is independent of Texas Parks and Wildlife.

Visit HuecoTanks.com

Hueco Tanks near El Paso is internationally known for bouldering, but a destination like that accumulates more than routes. It accumulates personalities, competitions, access rules, arguments about bolts, old photographs, local terminology, and stories that can disappear quickly when the people who remember them stop publishing.

This site kept publishing.

Its sections cover climbing history, access information, event results, interviews, essays, safety material, and years of local news. The older pages are particularly valuable because they preserve things in the formats people actually used at the time. The homepage still points to what it calls “Rare Archaeological video,” including footage of Albert Jimenez soloing Wyoming Cowgirl in the spring of 1992. The choices offered are gloriously period-correct: Windows Media and QuickTime.

There is also a roughly 17-minute video from the 1992 Rock Rodeo featuring Don “HardMan” Hardin, along with later Rock Rodeo results and photographs. Those files are more than climbing entertainment. They are visual records of how the place, equipment, routes, and community looked before online video became routine.

The site does not pretend to be neutral institutional documentation. Its access and policy sections contain opinions. That is useful in its own way. A local history made entirely from official notices would miss the conflict experienced by the people actually using the landscape.

For researchers, that mixture should be handled carefully: a climber-run page is evidence of what climbers said and preserved, not automatically the final authority on every policy dispute. But firsthand community archives often contain details no official agency had any reason to record.

For webmasters, HuecoTanks.com demonstrates another old-Web advantage. A site organized around a real place can accumulate value for decades simply by keeping old pages accessible. A 2000 news page does not have to be overwritten because 2001 arrived.

That sort of continuity is one reason CacheRat maintains the 1,967 Ancient Web Domains research list. Old sites frequently preserve not just information but the disagreements, media formats, and local voices surrounding it.

Hueco Tanks has plenty of newer coverage. This site gives you the sediment underneath it.

Dig through the climbing history and old media at HuecoTanks.com

Posted on by

Ancient Web: The Neitherworld Built a Beetlejuice Fandom Before the Platforms

The Neitherworld began in 1997 because C. “Sparky” Read looked around the Web, noticed a shortage of Beetlejuice sites, and decided that was apparently a problem worth fixing personally. Before fandom was organized around wikis, social networks, and giant media databases, that was how a surprising amount of culture got documented.

Visit The Neitherworld

Read’s own history places the beginning around mid-April 1997 as a directory on a University of Southern California web account. The project covered both the 1988 film and the animated series, but the cartoon side eventually became the larger archive.

Building a fan site then involved considerably more manual labor than opening an account somewhere. Read describes making screen grabs, sound clips, and video files using a brother’s computer during the summer of 1997. Those files then escaped into the wider Web, where copies turned up on other sites. The same thing happened with written material: the surviving episode-guide notes discuss earlier work being copied and circulated without credit.

That irritation produced something useful. The Neitherworld developed its own detailed animated-series episode guide along with fan fiction, fan art, multimedia, character material, and other resources. In April 1999 the project moved onto its own theneitherworld.com domain.

This is a good specimen of what fan infrastructure looked like before the word “platform” swallowed most of the Web. One maintainer decided what needed documenting, learned enough HTML and media handling to publish it, accepted contributions from other fans, and gradually turned a personal interest into a reference site.

It also preserves a less romantic part of early online culture. Images and audio were easy to copy. Attribution was inconsistent. Somebody could spend days creating a guide or capturing media and then watch it reappear elsewhere. The technology changed, but that argument certainly survived.

For researchers in 2026, an old fan site can contain information that is difficult to reconstruct from modern databases: what fans cared about at the time, which episodes or characters generated discussion, how media was exchanged, and what terminology the community actually used. The site’s own history is part of that record.

The CacheRat 1,967 Ancient Web Domains research list contains many sites built this way. They were unofficial, sometimes messy, and often more detailed than anything the rights holder bothered to publish.

The Neitherworld is still worth opening because it shows fandom when fandom had to build its own house.

Explore The Neitherworld and its Beetlejuice archive

Posted on by

Ancient Web: Sweetbee’s Zelda Outlands Page Captures ROM Hacking Before It Was Normal

There was a time when discovering a good ROM hack felt less like browsing a catalog and more like hearing about a bootleg tape from somebody behind a bowling alley.

Visit Sweetbee’s Zelda: Outlands page

Sweetbee’s write-up covers Zelda: Outlands, a substantial fan-made reworking of the original Legend of Zelda for the NES.

The author explains the problem immediately: the original Zelda was still a favorite, but after years of replaying it, every secret was burned into memory. You cannot rediscover a hidden staircase when your hands already know where to bomb before your brain does.

Then Sweetbee found Outlands.

A fan sequel hiding inside the original game

The page describes Outlands as far more than a sprite swap. The overworld is rebuilt. Dungeons are redesigned. Enemy graphics and behavior are rearranged. Item locations and secrets change. There is a new first quest and a punishing second quest.

Sweetbee’s verdict is unusually strong: the hack feels polished enough to function almost like a legitimate straight sequel to the original game.

That reaction is useful historically because ROM hacking was still culturally different from what it is now.

Today there are organized patch databases, Discord servers, GitHub repositories, video showcases, speedruns and entire communities dedicated to game modification. The old page describes a messier ecosystem: obscure websites offering unreleased and hacked games, ROM downloads surrounded by awful ads, homemade cartridges, IPS patches and fan pages maintained by whoever cared enough to document the thing.

The write-up even refuses to become a full walkthrough. Sweetbee gives tips but argues that excessive guidance would defeat the purpose of a hack designed to make Zelda mysterious again.

That is a lovely little time capsule of how these projects were experienced.

You did not necessarily arrive through a centralized mod platform. You heard about a strange cartridge. You found a page. You tracked down a patch. You figured out how to run it. Then you compared notes with somebody else who had fallen into the same hole.

The page also links back to the creator’s Outlands material, including the IPS patch and an HTML manual. Even where those surrounding links have aged, the review preserves the social context around the hack: this was not merely modified code, but a small fan-made release with documentation, reputation and a discovery trail.

CacheRat’s 1,967 Ancient Web Domains research list exists for exactly these fragments of game culture—the pages that preserve how people encountered software, not just the software itself.

A ROM file can tell you what the hack was.

A page like Sweetbee’s tells you what it felt like to find it.

Read Sweetbee’s Zelda: Outlands review

Posted on by

Ancient Web: OpusGames Turned a 1997 Personal Page Into a Game-Development Archive

Doug Lanford started what became OpusGames in September 1997, during the final push on Robotech: Crystal Dreams. The game never received a normal commercial release, but Lanford’s personal website survived—and that makes the site considerably more useful now than anyone could have planned at the time.

Visit OpusGames

Lanford’s own update history says the first version was thrown together over a four-day weekend. He had bought a cheap scanner and needed an excuse to use it. That is about as pure a 1997 personal-homepage origin story as one could request.

The site gradually accumulated material from Lanford’s career as a programmer and software engineer. There are sections on WarJetz, BattleTanx: Global Assault, Jurassic Park for Sega CD, Warriors of the Eternal Sun, homebrew Game Boy Advance experiments, code samples, hardware work, and other projects. But Robotech: Crystal Dreams is the historical center of gravity.

That Nintendo 64 game was being developed by GameTek and disappeared after the company’s collapse. A version was demonstrated publicly, but the finished product never reached players in the normal way. OpusGames preserves unusually close-to-the-source material about the project: development details, images, magazine material, dialogue and audio, and other remnants from somebody who actually worked on it.

The site’s update log is almost as interesting as the content. It records little additions over years: maps added in 1998, Robotech material and large sound files in 2001, a homebrew GBA game in 2003, and later revisions continuing long after the Web had moved through several generations of design fashion. The front page was still receiving updates decades after the first version.

This is one reason personal developer sites deserve preservation. A corporate product page normally tells you what was supposed to ship. A developer’s own archive can tell you what was attempted, what failed, what got canceled, and what the work actually looked like from inside.

For game historians, canceled titles are particularly vulnerable. Their documentation tends to scatter into magazine scans, prototype ROMs, old forum discussions, private collections, and the memories of former developers. A surviving firsthand site can connect those fragments.

The CacheRat 1,967 Ancient Web Domains research list contains a lot of personal pages, but OpusGames shows why “personal” should never be mistaken for “unimportant.” Sometimes the person maintaining the homepage was standing directly inside the history being documented.

Explore Doug Lanford’s games and development archive

Posted on by

Ancient Web: The Sleipner A Disaster Page Is a 29-Line Warning About Bad Numerics

Some engineering failures need books. Some can scare the hell out of you in 29 lines of HTML.

Read the Sleipner A disaster page

Douglas N. Arnold’s page summarizes the sinking of the first concrete base structure for the Sleipner A offshore platform during a controlled ballasting operation near Stavanger, Norway, on August 23, 1991.

The structure developed a leak, failed and sank.

Nobody needs a dramatic soundtrack after that sentence.

The numerical error was not academic

The post-accident investigation traced the failure to a wall in a critical triangular section called a tricell. According to the page, the design problem came from a serious error in the finite-element analysis combined with insufficient reinforcement anchorage.

The numerical model underestimated shear stresses by about 47 percent.

That is the entire moral of computational engineering in one number.

A later, more careful analysis predicted failure at a depth of roughly 62 meters. The actual failure occurred at about 65 meters.

The first structure was lost completely. Arnold’s summary puts the economic loss at roughly $700 million and notes that the collapse produced a seismic event measuring about 3.0 on the Richter scale.

The platform deck involved tens of thousands of tons of structure and equipment. The failed concrete base ended up as debris hundreds of meters below the surface.

This was not a rounding error in a homework problem.

Why this page still works

Arnold’s page is part of a broader collection on disasters attributable to bad numerics. The Sleipner example is especially effective because the chain is so clean:

A numerical model produced stresses that were too low. The design followed the model. The concrete wall was not adequate. Reality ran the test suite.

The page links the summary back to investigation material and structural-engineering articles, and it uses a few diagrams and photographs instead of burying the event under interface chrome.

It was last modified in 2009 and still reads perfectly well because the information architecture is basically indestructible: title, facts, explanation, references.

There is a lesson there for both engineers and web developers.

Models are not reality, and presentation layers are not knowledge.

CacheRat’s 1,967 Ancient Web Domains research list exists to surface pages like this—small technical documents that have survived because their usefulness is much larger than their visual footprint.

You can read the Sleipner page in a couple of minutes.

You may remember the 47 percent for the rest of your career.

Read Douglas Arnold’s Sleipner A summary

Posted on by

Ancient Web: This Boating Guide Was Tested in Netscape 2.0 for Unix

Geoff Kuenning’s boating page has one of the better compatibility statements left on the old Web: it was tested with Netscape Navigator 2.0 for Unix. Nearly thirty years later, the page is still doing what it was built to do—helping somebody remember a pile of rules that are much easier to recall when they rhyme.

Visit Definitions and Mnemonics for Sailors and Powerboaters

Kuenning explains that he assembled the material as a beginner rather than presenting himself as a maritime authority. That distinction matters. The page is essentially a public set of study notes covering navigation lights, sound signals, buoyage, collision regulations, compass work, weather, boat terminology, and the other information that starts accumulating when somebody learns to operate a boat safely.

The memorable parts are the mnemonics.

A vessel showing red over red is remembered as “Red over Red / This boat is dead.” A trawler becomes “Green over White / Trawling tonight.” Five short horn blasts become “Blast quick five / To stay alive.” And the familiar American buoyage reminder survives as “Red Right Returning.”

These are not offered as magical replacements for knowing the rules. Kuenning explicitly warns that memory aids are no substitute for the official regulations and proper seamanship. That makes the page more useful, not less: it knows exactly what kind of document it is.

The copyright dates are 1996 and 1997, and the page permits nonprofit and educational reproduction with attribution. Its construction is equally period-specific. There is no application shell, no cookie banner, no account, and no script required to reveal the next definition. It is a long, organized reference page designed to be read, searched, printed, and linked.

That old information architecture aged unusually well. A modern webmaster looking at it can see why. The headings describe the material. The text carries the information. Links lead somewhere useful. Almost nothing stands between the reader and the reason the page exists.

For historians of the Web, the Netscape note dates the culture as neatly as a browser screenshot. For boaters, some of the material remains recognizable because navigation rules and the human need for dumb little rhymes have both proved more durable than browser wars.

Pages like this are why the CacheRat 1,967 Ancient Web Domains research list is interesting beyond nostalgia. An ordinary personal study aid can survive long enough to become both a usable reference and a specimen of early Web publishing.

Read the complete boating mnemonics reference

Posted on by

Ancient Web: DPGraph Drew Eight-Dimensional Math in Assembly Language

DPGraph is a mathematical graphing program from the late 1990s with a specification that sounds increasingly strange the longer you read it. It draws dynamic and interactive mathematical visualizations extending from ordinary two-dimensional plots through what its creator describes as eight-dimensional graphs, and the whole program was written in assembly language.

Visit the DPGraph website

Dave Parker released DPGraph in 1999, and the surviving update history still traces changes from that first release forward. The program was built for interactive mathematics and physics visualization: equations, animated surfaces, transformations, implicit plots, and other objects that are easier to understand when they can be manipulated instead of frozen into a textbook image.

Then there is the implementation detail. DPGraph’s site says the program was written entirely in assembly for speed. That was not unheard of in the 1990s, but an entire high-level visualization application written that way is still a pretty aggressive engineering decision. Modern software stacks often require several layers of frameworks before a window appears. DPGraph went in the opposite direction.

The site’s own usage claim is equally ambitious: it says more than two million mathematicians, physicists, teachers, and students at more than a thousand schools were licensed to use the software. That number is the site’s claim rather than an independently audited statistic, but it indicates the scale Parker believed the program had reached.

The old compatibility notes are now part of the attraction. DPGraph and its viewer were distributed for generations of Windows stretching from the Windows 95/NT era into Windows 10, with references to running under Linux and emulation as well. The viewer could also be used to present graphs through Web pages, a reminder of the period when downloadable helper applications and browser-linked viewers were normal parts of publishing interactive material online.

The latest prominent news on the site dates to 2019, and Parker now describes DPGraph as freeware with no foreseeable updates. That is not necessarily a sad ending. A piece of specialist software can be finished without becoming useless.

For software historians, the site preserves release notes, documentation, examples, downloads, and the author’s own description of what he built. For teachers and mathematically curious visitors, it remains a doorway into a visualization tool that came from a very different software culture.

It also illustrates one of the recurring patterns in CacheRat’s 1,967 Ancient Web Domains research list: small software sites can outlive entire commercial distribution ecosystems because the author kept the files and documentation in one stable place.

The dimensions may be difficult to picture. The old website explaining them is pleasantly straightforward.

Explore DPGraph, its history, documentation, and downloads

Posted on by

Ancient Web: Chris Covell Turned Retro Console Curiosity Into a Technical Archive

A lot of retro-gaming sites tell you what a game was like. Chris Covell’s site keeps asking what the hardware was actually doing.

Visit Chris Covell’s site

Covell is a Canadian retro programmer and researcher who has lived in Japan since 2002. In an interview preserved on his own site, he says he became interested in assembly programming for systems such as the NES and Commodore 64 in the late 1990s and has been doing retro programming since around 1998.

That background explains the archive.

This is not merely a collection of nostalgia pages. Covell’s site contains technical FAQs, hardware experiments, homebrew development, scans and translations of Japanese material, documentation of obscure peripherals, and research into how old games were actually made.

The useful kind of obsession

One of the site’s best-known documents is the NES Technical / Emulation / Development FAQ, updated to version 1.7 in 2008. Covell describes it as a launching point for people trying to understand NES/Famicom hardware, emulation and development. It covers memory mappers, audio, regional hardware differences, expansion features and other practical details.

The tone is technical without pretending to be bloodless. Covell openly says the document includes his opinions and humor rather than presenting itself as a sterile specification sheet.

Elsewhere, the site preserves material that would be easy to lose entirely.

Covell scanned and translated Japanese children’s books showing how games such as Super Mario Bros. 3 and Dragon Quest VI were developed. His pages preserve interviews and behind-the-scenes material from Japanese game development. He documents odd hardware such as NEC’s unreleased Tsushin Booster, a proposed PC Engine modem/BASIC peripheral from around 1989–1990.

He also builds things.

In interviews, Covell describes using homebrew demos to test claims about old hardware. If magazines or websites repeated a questionable technical limitation, his response was often to write software and see whether the machine actually behaved that way.

That is a very old-school hacker method: stop arguing abstractly and make the hardware confess.

The site’s front page itself is not always the easiest thing for modern crawlers to digest, but the individual documents remain scattered across the domain and continue to be cited by retro-development communities, game-history articles and technical discussions.

That is another preservation lesson. A personal site can become infrastructure without ever announcing itself as an archive.

CacheRat’s 1,967 Ancient Web Domains research list exists to surface exactly this type of domain—the places where years of experiments, scans and specialist knowledge accumulate behind a homepage most people will never see.

Covell’s archive is what happens when curiosity lasts longer than the hardware’s commercial lifespan.

Fortunately, that is how a lot of technical history survives.

Explore Chris Covell’s retro-computing archive

Posted on by

Ancient Web: Modulated Light DX Sent Audio 104 Miles Through the Air

There is radio DX, where people try to communicate over absurd distances with radio waves. Then there is Modulated Light DX, where Chris Long and Mike Groth document the equally reasonable decision to send audio across the landscape using beams of visible light.

Visit Modulated Light DX

The site is a deep archive of amateur optical communication rather than a single stunt. It reaches back through heliographs, early experiments, laser work, receiver design, atmospheric effects, and decades of Long’s own light-communication experiments. Several of the pages originally lived on Tony Sanderson’s Bluehaze site before being mirrored and updated here, which makes the collection itself part of the preservation story.

The headline achievement came on February 19, 2005. Long and Groth established an audio-modulated light path of about 167 kilometers, or 104 miles, between South Barrow peak and Mount Wellington in Tasmania. The equipment used high-output LEDs rather than treating a laser as the only respectable way to do optical communication.

That distinction is important. The site spends real time on the behavior of light through the atmosphere: turbulence, scintillation, beam spreading, receiver sensitivity, and the practical advantages and disadvantages of coherent and non-coherent sources. It is not merely a photograph of two people pointing flashlights at mountains.

The archive also reaches much further back. Long preserved material on optical communications from the nineteenth and early twentieth centuries, accounts of 1960s laser experiments, and his own work from the late 1960s onward. A 1979 article on modulated-light communication survives alongside later revisions and commentary. Read in sequence, the site becomes a history of a technical hobby that never developed anything like the enormous online infrastructure surrounding amateur radio.

That is exactly why a specialist Web page matters. Search for optical communication today and the results quickly become fiber networks, photonics companies, research papers, or modern free-space optical systems. Those are legitimate subjects, but they can bury the homebuilt receivers, scavenged optics, audio experiments, and hobbyist engineering that came before them.

For modern experimenters, the circuits and discussions still provide starting points. For Internet archaeologists, the site is evidence of how technical communities used personal pages as laboratories, notebooks, bibliographies, and museums at the same time.

The larger CacheRat 1,967 Ancient Web Domains research list is full of small sites whose value comes from exactly this sort of depth.

A 104-mile beam is impressive. The better achievement may be that decades of the work around it are still readable.

Explore the experiments, circuits, and optical DX history