Category: Ancient Web

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

Posted on by

Ancient Web: One Long Page Tried to Rescue Louis Vierne From the Organ Loft

Some websites answer a question. Some spend several thousand words trying to convince you that the culture has failed to appreciate a dead French organist.

I prefer the second kind.

Read the Louis Vierne page at Philosopher’s Wheel

Eric Meece’s page, mostly written around 2008–2009, is titled Louis Vierne (1870–1937): The Man and his Music. It is not a neutral encyclopedia entry. It is a sustained argument that Vierne deserves to be taken seriously as a major composer and that modern neglect of organ music has helped push him out of wider musical memory.

The page is huge.

Its navigation breaks out Vierne’s life, the history of the organ, all six organ symphonies, the Pièces de Fantaisie, the 24 Pieces in Free Style, orchestral and piano works, links and a discography.

This is a personal website behaving like a monograph.

Criticism with a point of view

Meece does something modern reference writing often sands away: he tells you what he thinks.

He argues that the organ’s association with church services has narrowed the modern audience for concert organ music. He explains the rise of the French symphonic organ, especially the instruments of Aristide CavaillĂ©-Coll, and places Vierne in a lineage running through CĂ©sar Franck and Charles-Marie Widor.

Then he gets into Vierne himself: nearly blind from birth, partially aided by surgery, a student of Franck and Widor, appointed organist at Notre-Dame de Paris in 1900, and a teacher connected to a remarkable generation of French organists and composers.

The page is especially good at showing why a specialist’s enthusiasm can matter. A short biography can tell you that Vierne wrote six organ symphonies. Meece wants to tell you why the instrument itself changed, why the French style became orchestral, why these works sound the way they do, and why he thinks you should care.

You do not have to agree with every judgment to find the archive useful.

That is the old-Web bargain: the author gives you sources, context, opinions, links and enough detail to argue back.

The page is visibly handmade and occasionally idiosyncratic. It includes personal commentary, historical digressions and judgments that would probably be edited out of a conventional encyclopedia. But those are precisely what make it a record of listening rather than a database row.

CacheRat’s 1,967 Ancient Web Domains research list is full of this kind of one-person scholarship—the page that exists because somebody cared about one subject far past the point where a normal editor would have said, “Okay, that’s enough about French organ symphonies.”

Fortunately, nobody stopped him.

Explore Eric Meece’s Louis Vierne essay

Posted on by

Ancient Web: A Collector Mapped the Strange History of Soviet Calculators

Sergei Frolov’s Soviet Calculators History is the sort of page that starts with an abacus and eventually leaves you staring at a family tree of electronic machines most Western computer histories barely mention. It is part collector’s guide, part technical history, and part artifact from the era when enthusiasts translated their own specialist research onto the Web.

Read Soviet Calculators History

Frolov begins before electronics with the Russian abacus, or schoty. One of the small but revealing details on the page is that a Soviet trade-calculation textbook published as late as 1986 still devoted material to performing calculations on an abacus. From there the story moves through slide rules and into the increasingly complicated world of Soviet electronic calculators.

The value of the page is not that it simply lists model numbers. Frolov was collecting the machines themselves. He describes a collection of more than fifty distinct Soviet calculator models, along with dozens of books, magazines, manuals, and related pieces of documentation. That physical collection gave the Web page a level of specificity that generic computer-history timelines often lack.

There is also an unusually visible translation history. The English page explains that it came from a translated Russian article and points readers toward the Russian original and other English translations. Frolov even asks readers to help correct the English. That is wonderfully old-Web behavior: publish the useful thing now, expose where it came from, and let knowledgeable visitors improve it.

For researchers in 2026, this matters because Soviet consumer electronics can be awkward to study across language, model-name, and documentation barriers. A collector who knows which machines belong together, what literature accompanied them, and how the technology changed over time can provide a map that would be expensive to reconstruct from scattered auction listings and isolated photographs.

The page also demonstrates why personal collections were so important to the early informational Web. The author did not need to establish an institution before publishing. Owning the hardware, studying it, finding the literature, and caring enough to explain it were sufficient credentials to create a useful reference.

That model produced some uneven pages, of course. It also produced resources that institutions never bothered to build.

The larger CacheRat 1,967 Ancient Web Domains research list contains many of these specialist corners. Frolov’s calculator history is a particularly good example because the subject is both technical and geographically easy to miss.

The machines may be obsolete. The collector’s map of them is not.

Explore the Soviet calculator archive and model history

Posted on by

Ancient Web: Grant Searle Rebuilt a 1974 Pong Machine Without a Microprocessor

Before a home videogame could be a small computer, it could also be a pile of logic chips whose entire job was to make two paddles, one ball, and a score appear on a television. Grant Searle’s Tele-Tennis page documents his reconstruction of exactly that sort of machine: a 1974 Practical Wireless project built without a microprocessor and without a line of software.

Visit Grant Searle’s Tele-Tennis page

The original design was published as a series by M. J. Hughes in Practical Wireless from July through November 1974. It belonged to the first wave of Pong-like home television games, when duplicating the basic idea meant constructing the video system itself. The magazine supplied circuit diagrams, component information, board layouts, and enough explanation for a sufficiently determined reader to build one at home.

Searle finally did that decades later.

The interesting part is what is not inside. There is no CPU running a game loop and no ROM containing game code. The machine builds its behavior directly from electronics: 7400-series logic, 555 timers, op-amps, diodes, counters, and separate circuits responsible for pieces of the picture. The paddles and ball are not sprites in memory. Their positions exist because the circuit generates the correct signals at the correct moments while the television image is being formed.

That makes the project useful far beyond nostalgia. A modern emulator can show what an early videogame looked like. This shows what one was at the electrical level.

Searle’s reconstruction also preserves the practical messiness that gets erased from neat histories of videogames. The original project occupied multiple boards. The 1970s design expected a UHF television connection. Searle adapted the output for a more convenient composite-video setup and put his build in an appropriately period-looking case. It is less a replica of a commercial Pong console than a reconstruction of the era when electronics magazines expected readers to manufacture their own entertainment.

For a webmaster, there is another preservation lesson here. The page is valuable because one person documented a completed build, not because somebody later summarized that such projects existed. Photographs, construction notes, scans, and firsthand decisions turn an old circuit into something another person can understand and reproduce.

CacheRat’s larger 1,967 Ancient Web Domains research list contains plenty of sites like this: small technical pages that quietly became historical evidence because somebody left them online.

Tele-Tennis is especially good evidence. The software layer is gone because there never was one.

Explore the Tele-Tennis reconstruction and original project material

Posted on by

Ancient Web: World War 1 Naval Combat Is a Hand-Built Map of the North Sea War

Old specialist sites are often at their best when the author tells you exactly what the site is good at—and exactly where it is a little crap.

Visit World War 1 Naval Combat

This site focuses mainly on surface warfare between the Imperial German Navy and the British Royal Navy during the First World War.

Its front page routes visitors into the Battle of Heligoland Bight, the loss of HMS Aboukir, Hogue and Cressy, HMS Audacious, Admiral Graf Spee’s cruiser squadron, Dogger Bank, Jutland, major warship losses, the Scapa Flow scuttling, warship reference pages, dispatches and image collections.

Then the author offers one of the more useful disclaimers on the historical Web: the maps are crude, should be treated as rough indications, and if the maps disagree with the text, the text is probably “the least inaccurate.”

Do not plan your holiday with them, he adds.

That is how you earn trust online.

A personal history site with scope discipline

The author openly explains why British and German forces dominate the site: they were the major naval powers involved in this part of the war, and he is British, so those navies are what interest him most.

That admission is more useful than pretending the archive emerged from nowhere with perfect neutrality.

The site is also a reminder of how much military-history material once lived in hand-built topic pages rather than giant databases. A visitor could start with one battle, follow individual ships, move into fleet losses, compare photographs and read dispatch material without being pushed into a recommendation engine.

The copyright footer runs from 1998 to 2026, making the site itself nearly as interesting as some of the material it preserves. This is not a frozen GeoCities snapshot. It is an old-style site whose basic architecture survived long enough to span dial-up, broadband, smartphones, social media and the present day without deciding that every naval battle needed a vertical video.

The design is not polished by modern standards. That is not a serious problem.

Historical reference work benefits from predictable structure, visible navigation and pages that remain at the same URLs for years. Fancy interfaces have a habit of becoming archaeological layers themselves.

CacheRat’s 1,967 Ancient Web Domains research list exists to surface sites like this: individual researchers who built durable knowledge structures before the Web learned to confuse redesign with improvement.

There are newer naval-history resources. There are prettier ones. There are certainly more comprehensive ones.

But there is something valuable about a site where the author has spent decades saying, in effect: here are the battles I care about, here is what I found, and please do not navigate the North Sea using my GIF maps.

Explore World War 1 Naval Combat

Posted on by

Ancient Web: Topher’s Castle Is a 1997 Homepage That Never Learned to Pick a Topic

Modern websites are repeatedly told to pick a niche. Topher’s Castle, established in July 1997, appears to have received this advice and spent nearly three decades refusing it.

Visit Topher’s Castle

The homepage calls itself “A Great Site for Everyone!” and then makes a serious attempt to prove it. The navigation includes advice and opinions, air-travel information, baby names, cast-iron collecting, Christmas material, famous birthdays, a Homer Simpson archive, kids’ pages, music, NASA, surnames, television, a voting booth, and a Weird Picture of the Week.

That is not a content strategy. It is a person having a website.

Topher’s Castle belongs to an era when a personal homepage was allowed to represent the full mess of somebody’s interests. A page did not need to satisfy an audience profile. If the webmaster liked television and family material and space exploration and cookware, all of it could live under the same roof. The architecture was the person.

That freedom produced some famously chaotic navigation, but it also created sites with personality that is difficult to reproduce when every subject is split into a different account, platform, or branded project. A visitor could arrive for one useful page and accidentally discover that the same person had built twenty other things they never would have searched for.

The site’s age makes the effect stronger. A personal homepage launched in 1997 and still reachable in 2026 is no longer just a pile of unrelated interests. It is a longitudinal artifact. Sections, terminology, link styles, and subject choices record what people thought was worth putting on the public Web when having your own domain or homepage felt like possession of actual territory.

There is a lesson here for modern independent sites. Focus is useful, especially when people need to understand quickly what a business or publication does. But the Web also lost something when every stray interest was treated as an error in positioning. Personal sites could be weird precisely because nobody had to explain the conversion funnel for the NASA page sitting next to the baby-name page.

Topher’s Castle remains fun to explore because the next link is difficult to predict. That used to be a normal feature of the Web.

Enter Topher’s Castle

After almost thirty years, its strongest organizing principle may still be the simplest one: Topher thought this was worth putting online.

Posted on by

Ancient Web: Mug Shots Has Classified A&W Root Beer Glassware Since 1995

Somewhere around the summer of 1995, Ken Polsson decided the Web needed a detailed catalog of A&W Root Beer mugs.

He was correct.

Visit Mug Shots

Polsson says he had been collecting A&W mugs since around 1980. When he first put the site online as Ken P’s A&W Root Beer Mugs, the collection was about 30 mugs, with images of roughly a dozen.

Then the Internet did what the Internet used to do best: other weirdly interested people found him.

Comments and collector information came in. The page expanded beyond one man’s shelf into a broader reference for A&W memorabilia, corporate history, known mug varieties and collector resources.

When a hobby becomes a database

The most impressive part is the classification system.

Polsson did not merely upload photos and call it done. Mug varieties are grouped by logo style, capacity, physical characteristics and country. Individual examples receive codes such as L01fC16b, allowing tiny distinctions in logo treatment, size, fluting and base shape to be recorded without writing an essay every time.

In 2006 he rebuilt the numbering system after roughly fifteen years of using earlier codes. That is database maintenance disguised as collecting.

The update log is a historical artifact by itself.

It records newly discovered mug variations, corrected classifications, logo differences, convention mugs, collector editions, auction prices, company history work and photography updates. A January 2006 entry even notes that Polsson re-photographed mugs without root beer in them because the brown liquid interfered visually with brown logos.

That is absurdly specific.

It is also exactly why specialist sites matter.

The collection’s introduction describes how prices changed from swap-meet finds costing pocket change to harder-to-find examples costing substantially more in antique shops. It documents how the hobby expanded into coffee mugs and Great A&W Root Bear memorabilia. It also records the site’s transition into selling duplicates to help fund further collecting.

By 2024, the site was still being updated.

This is what happens when one person maintains a niche database for nearly thirty years: details accumulate that no corporation would ever commission because no executive could explain the market size without being laughed out of a meeting.

And yet those details become useful precisely because nobody else bothered.

CacheRat’s 1,967 Ancient Web Domains research list exists for this species of website—the collector who accidentally becomes the archivist because he kept showing up longer than everyone else.

Mug Shots started with about 30 mugs.

Three decades later, it is part price guide, part taxonomy, part corporate-history scrapbook and part proof that one obsessed person can build a reference work by simply refusing to stop cataloging things.

Browse the Mug Shots A&W archive

Posted on by

Ancient Web: ThaiWorldView Became a Personal Atlas of Thailand and Hong Kong

ThaiWorldView is the kind of website that becomes difficult to summarize because the creator did not stop after answering one question. It grew into a personal atlas: photographs, places, religious sites, history, customs, architecture, and everyday observations spread across a large hand-built collection centered on Thailand and Hong Kong.

Visit ThaiWorldView

The site rewards browsing rather than scrolling. Its material is organized into subject areas and places, sending readers sideways through related pages instead of feeding them a single chronological stream. That makes it feel closer to an illustrated reference shelf than a travel blog.

The photographs are especially important. Tourist destinations are easy to find on the modern Web, but independent long-running sites often preserve a different kind of visual record: neighborhoods, religious practices, street scenes, buildings, local details, and places photographed before later development changed them. The value accumulates almost accidentally as the pictures age.

ThaiWorldView also demonstrates what happened when personal publishing escaped the limits of a printed photo album. A person interested in a country could keep adding pages indefinitely. One temple did not need to displace another. A section on history could sit beside everyday life. Hong Kong could expand into its own body of material without forcing the entire project into a new format.

That sprawling quality can look unfashionable beside modern minimalist design, but it has an advantage for researchers: there is room for specificity. Proper names, place names, traditions, images, and contextual notes remain attached to pages that search engines can surface individually.

For webmasters, it is another reminder that information architecture does not have to begin with a publishing cadence. A site can be organized around the subject itself. The visitor is not necessarily asking, “What did you post this week?” They may be asking, “What else do you have about this place?”

That question is where sites like ThaiWorldView become unusually durable. A social post expires quickly. A well-labeled page about a specific location can keep answering somebody’s search years later.

The larger 1,967 Ancient Web Domains research list contains many independent projects with the same quality: their usefulness increased because their creators kept adding instead of replacing.

Explore ThaiWorldView

It is less a snapshot of the Web than a reminder of what happens when somebody is allowed to build their own corner of it for as long as they want.

Posted on by

Ancient Web: Blackjack in Color Turned Card-Counter Math Into Visual Charts

Blackjack in Color looks like somebody decided that tables of gambling probabilities were not nerdy enough and therefore needed thousands of simulations, distributions, and colored charts. The result is much more interesting as a mathematical Web artifact than the casino subject might suggest.

Visit Blackjack in Color

The site is built around visual explanations of probability. Instead of reducing everything to a slogan about whether a particular system “works,” it shows distributions, frequencies, count behavior, and the uncomfortable fact that random processes remain noisy even when the underlying mathematics is understood.

That makes the site useful as an example of simulation-based explanation. A chart can show something that a single expected-value number hides: how wide the range of outcomes is, how often unusual runs occur, and how long short-term results can wander away from an average.

The subject happens to be blackjack, but the presentation belongs to a broader tradition of independent technical websites. Somebody has a specialized question, writes software to investigate it, generates data, then publishes enough of the machinery that other interested people can inspect the reasoning instead of accepting a conclusion on faith.

That approach was especially valuable on the older Web because specialist information was often scattered between books, forums, and private correspondence. A purpose-built website could assemble the charts, terminology, and explanations into one stable location.

There is also a good design lesson in the name Blackjack in Color. The visualizations are not decorative dashboard furniture. Color is used because the site is trying to make relationships in a large numerical subject easier to see. The graphics support the explanation.

For a reader in 2026, the interesting part is not a promise of easy money. Randomness does not become obedient because a website drew a graph. The value is seeing a niche enthusiast treat probability seriously enough to model it, visualize it, and document what the results actually look like.

That is the old specialist Web at its best: a narrow subject, a technically minded author, and a pile of work that exists because somebody wanted the answer badly enough to calculate it.

Explore the simulations and charts at Blackjack in Color

Even if you never sit at a blackjack table, the site is a useful museum of applied probability with unusually colorful exhibits.

Posted on by

Ancient Web: The Canadian Nuclear FAQ Has Been Answering CANDU Questions Since 1996

A lot of websites claim to be “living documents.” The Canadian Nuclear FAQ has receipts.

Visit the Canadian Nuclear FAQ

Dr. Jeremy Whitlock says the site was established in April 1996 as an unofficial, privately maintained collection of frequently asked questions about Canadian nuclear science and technology, especially CANDU reactors.

The homepage was still being updated in August 2026.

That is three decades of one technical website refusing to die.

The FAQ format when FAQ meant something

The structure is gloriously old-fashioned.

The main index is divided into subjects such as CANDU technology, the Canadian nuclear industry, costs and benefits, safety and liability, waste management, security and non-proliferation, uranium, research reactors, other research and development, and further information.

Then come actual questions.

What does CANDU mean? How does a CANDU reactor work? What is heavy water? How does on-power refuelling work? Why does the design have a positive void coefficient? How is nuclear waste managed in Canada? What happened in the 1952 NRX accident? Can CANDU reactors use alternative fuel cycles? What role has Canada played in isotope production?

This is what a real FAQ looked like before the phrase got reduced to five accordion boxes written by marketing.

Whitlock is also unusually clear about provenance and bias. The site states that it is unofficial and that its answers do not necessarily represent Atomic Energy of Canada Limited, Canadian Nuclear Laboratories, the Canadian Nuclear Society, or the nuclear industry generally.

That disclosure matters because Whitlock is not a random hobbyist. His biography says he earned a PhD in Engineering Physics from McMaster University specializing in CANDU reactor physics, worked for AECL and later Canadian Nuclear Laboratories, served in nuclear safeguards, and from 2017 to 2025 worked at the International Atomic Energy Agency.

So the site sits in an interesting category: personal, unofficial publishing backed by deep professional expertise.

That combination used to be one of the Web’s great strengths.

A subject-matter expert could maintain a public reference indefinitely without waiting for an institution to approve every paragraph or redesign the entire thing because somebody discovered rounded cards.

The page even retains a fossil from an earlier Internet layer: a Nuclear InfoRing WebRing section. The technical content kept evolving while some of the surrounding navigation culture froze in amber.

CacheRat’s 1,967 Ancient Web Domains research list is built to surface exactly these long-lived specialist references—the places where expertise accumulated slowly enough that replacing them with a generic search result would be a downgrade.

Thirty years later, the Canadian Nuclear FAQ still does the job implied by its name.

That sounds ordinary until you remember how little of the 1996 Web survived long enough to update its footer in 2026.

Read the Canadian Nuclear FAQ

Posted on by

Ancient Web: NYU Put Archimedes’ Cattle Problem on the Web

Archimedes managed to make counting cattle unpleasant enough that mathematicians were still talking about it more than two thousand years later. An old NYU mathematics page preserves the statement of the Cattle Problem in exactly the kind of university-hosted Web format that deserves to survive.

Read Archimedes’ Cattle Problem at NYU

The setup sounds almost harmless. The problem describes cattle belonging to the sun god, divided by color and sex, and gives relationships between the sizes of the groups. In mathematical form, those relationships become simultaneous Diophantine equations: equations whose solutions must be integers.

Then Archimedes makes things worse.

Additional conditions involving square and triangular numbers transform the exercise from a system that can be handled algebraically into a famously difficult number problem with an enormous smallest solution. The pastoral language is camouflage. Underneath the cows is serious number theory.

The NYU page is interesting partly because of the mathematics and partly because of how little ceremony it needs. University Web servers once accumulated thousands of pages like this: lecture notes, demonstrations, source texts, problem statements, datasets, and small research tools placed online because a faculty member or department thought somebody might need them.

They were not always designed as “content.” They were closer to shelves.

That model was extraordinarily useful. A search could land directly on the statement of a centuries-old mathematical problem without requiring an account, a viewer application, a learning platform, or a twelve-step navigation path through an institutional homepage. The URL itself points into an Archimedes directory. You are looking at the organizational habits of an earlier academic Web as well as the mathematics.

For a modern reader, the page is a useful starting point because it preserves the original shape of the challenge. You can see how an ancient word problem becomes a set of integer constraints and why the problem’s reputation is not based on merely having a lot of cattle to count.

Old university pages are fragile because they often survive outside whatever system the institution currently considers official. That makes finding and documenting them worthwhile while they are still there.

Open the Cattle Problem statement

Archimedes did not have a Web browser. He did, however, leave future computers something substantial to chew on.