Category: Ancient Web

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

Posted on by

Ancient Web: The Wererat’s Lair Kept Its Mid-1990s Web Design on Purpose

The Wererat’s Lair does not merely look like a mid-1990s website by accident.

It has been actively kept that way.

Visit The Wererat’s Lair

The site mixes pet-rat material, personal pages, role-playing games, photo galleries, old web tutorials, MIDI, and years of updates into one hand-built maze.

The navigation literally tells visitors to click on stones to enter the Lair.

That is the correct energy.

Hand-coded in 1996 and still being maintained

A 2019 update includes the line: “why did I ever code this by hand back in 1996?”

That dates the site’s basic architecture and explains a lot about what survives.

The update log stretches across pet rats and gerbils, D&D and Pathfinder campaigns, local web-design projects, family losses, galleries, and site maintenance. In 2020 the author noted that keeping the main Lair looking and sounding like the mid-1990s took actual effort. Old MIDI files and Java applets had stopped working, so parts were converted or replaced to keep the experience alive.

There is also a practical layer beneath the cave graphics.

The site maintained rat-care resources for decades. A 2024 update says the “Rat-help” section had finally ceased after nearly thirty years, while classic answers were left online. Elsewhere are rat-book recommendations, galleries, memorial pages, and older care information.

The web-design material is its own time capsule. One tutorial walks through a JavaScript photo album using arrays, form fields, and numbered image files—the kind of thing personal-site authors once copied, modified, and learned from directly.

The Wererat’s Lair therefore preserves more than appearance.

It preserves a way of making websites: write the HTML yourself, bolt on a script when you need one, collect whatever subjects interest you, and let the whole thing become structurally weird over time.

CacheRat’s 1,967 Ancient Web Domains research list includes personal sites like this because genuine old-web continuity is rarer than retro styling.

Anyone can imitate 1996.

The Wererat’s Lair was actually there.

Return to The Wererat’s Lair

Posted on by

Ancient Web: P-Car Built a Porsche 993 DIY Library One Repair at a Time

P-Car’s Porsche 993 DIY section has the tone of somebody who likes driving the car but may like taking it apart almost as much.

Visit P-Car

The author says half the fun of owning a Porsche 911 is working on it, and then proceeds to back that up with a very large collection of repair and modification pages focused on the 993 generation.

The archive is not a glossy buyer’s guide.

It is a garage notebook with hyperlinks.

The repair database grew job by job

The DIY index is divided into basics, scheduled maintenance, repairs, and modifications.

Routine jobs include oil changes, brake work, spark plugs, transmission fluid, distributor caps and rotors, belts, filters, and brake-system flushing.

Repair pages go deeper: alternator removal, clutch problems, spoiler mechanisms, seat-drive shafts, DME relays, steering racks, wheel bearings, starter replacement, odometer repair, air-conditioning diagnostics, and more.

The modification section is equally specific, with suspension work, aero-kit installation, harnesses, lighting, oil coolers, gauge faces, Recaro seats, radio controls, and other owner projects.

Many entries are dated from the early 2000s. Some were contributed by other owners, which turns the site into a small community technical library rather than a single person’s notes.

The individual pages preserve exactly the details that disappear when a forum thread breaks: tool sizes, part numbers, difficulty estimates, photographs, order of operations, and warnings about what not to drop into the engine bay.

One oil-change article specifies roughly eleven quarts of oil, two filters, drain-plug seals, and the exact Porsche part numbers involved. Another contributed procedure explains how to clean the idle stabilizer valve with basic hand tools and solvent.

That specificity is why old enthusiast sites remain useful.

A broad article can tell you that a 993 is maintainable.

A page written by somebody who actually removed the part can tell you which 7 mm socket you need.

CacheRat’s 1,967 Ancient Web Domains research list includes enthusiast sites because the best of them quietly accumulated practical knowledge for years.

P-Car is one of those garages that happened to have a web server attached.

Return to P-Car

Posted on by

Ancient Web: Jef Poskanzer Made an Under Construction Page About Under Construction Pages

For a while, one of the most common statements on the World Wide Web was that the page you were looking at was not finished yet.

Visit Jef Poskanzer’s Under Construction page

Jef Poskanzer turned that cliché into a page of its own at ACME Labs in August 1996.

The idea was simple: collect examples of the little signs, workers, barricades, animated GIFs, and warnings people used to announce that a website was still being built.

In other words, document a design habit while it was still happening everywhere.

The phrase became infrastructure

A surviving quotation from the page records Poskanzer checking how many pages contained the phrase “under construction.”

Using AltaVista, he noted roughly 400,000 results in August 1996 and about a million by December of that year. The count kept climbing into the millions as the web expanded.

The numbers are less important than the instinct to count them.

Most early-web decoration was treated as disposable. Poskanzer noticed that a small visual joke had become a recognizable language of the web and started preserving examples.

ACME Labs itself gives the page additional context. Poskanzer’s update archive records his first web page in July 1994, the launch of acme.com in November 1995, the Under Construction collection in August 1996, and a long chain of software releases and small web tools afterward.

That continuity makes ACME feel less like an old website than an old workshop that simply never closed.

The Under Construction page is funny now because the phrase eventually became unnecessary. Modern publishing systems let unfinished work stay private until launch. Hosted platforms supply polished maintenance pages automatically. The hand-picked animated construction worker largely vanished.

But in the 1990s, unfinishedness was often public.

People built pages directly on the live web, apologized for missing sections, and asked visitors to come back later.

CacheRat’s 1,967 Ancient Web Domains research list includes artifacts like this because web history is not only browsers, protocols, and companies.

Sometimes it is a million tiny GIFs saying the place is not ready yet.

Return to Jef Poskanzer’s Under Construction page

Posted on by

Ancient Web: John and Mary Kormendy Watched One Texas Yard for 22 Years

A backyard bird list becomes something else when the same people keep it for twenty-two years.

Visit John and Mary Kormendy’s House Birds page

From spring 2000 until summer 2022, John and Mary Kormendy lived near Dripping Springs in the Texas Hill Country and kept track of the birds around their house.

By the end, the page recorded 125 species seen and two more heard.

That is already useful.

The comments make it better.

Natural history at household scale

The page does not read like a generic species guide. It records what happened at this particular place.

Some birds were breeding residents. Others appeared only during migration or cold weather. Some once-common visitors became rare. A Rufous hummingbird overwintered at the property for multiple years. A Crested caracara appeared only twice in roughly two decades. Golden-cheeked warblers passed through even though the property did not have enough mature oak habitat for them to breed there.

The notes are full of small observations that field guides cannot provide.

Mary once saw a roadrunner leap from the roof to catch a Black-chinned hummingbird. Black vultures sat together on the upstairs deck. Woodpeckers hunted insects from the railing. Ringtails occasionally arrived and moved through the trees “like ballerinas.”

The Kormendys say they updated the pages heavily during 2020, when pandemic restrictions kept them home and paying closer attention. By late that year they had photographs of 64 of the 125 species on the list.

The archive also records change.

The couple moved to downtown Austin in 2022. The page says one final update remained and then the house-bird site would be “finished forever.”

That phrase is strangely perfect for a personal web page.

The site did not fail.

The observation period ended.

CacheRat’s 1,967 Ancient Web Domains research list includes personal natural-history sites because they preserve long spans of attention that are difficult to reconstruct later.

One yard, watched long enough, becomes a dataset with stories attached.

Return to the House Birds page

Posted on by

Ancient Web: Canadian Soldiers Built a Research Library Around the Army’s 20th-Century History

CanadianSoldiers.com is the sort of site that becomes valuable when the subject is too large for one article and too specific for a general encyclopedia.

Visit Canadian Soldiers

The site is built around the history of the Canadian Army, especially the twentieth century. Its research sections bring together official histories, bibliographies, campaign material, unit information, and links to digitized documents.

The useful part is the structure.

A bibliography that keeps opening doors

One section traces the official histories of the Canadian Army through the First World War, Second World War, and Korea.

The First World War story alone is complicated. The site explains that Colonel A. F. Duguid was tasked with producing an official history after the war, but only one volume appeared. Gerald Nicholson later produced the single-volume Canadian Expeditionary Force, 1914–1919 in 1964.

For the Second World War, the site points readers toward the three-volume official history, including C. P. Stacey’s Six Years of War and G. W. L. Nicholson’s volume on the Canadians in Italy.

Its bibliography pages then fan outward into general histories, campaign studies, memoirs, and specialized works.

That is the old-web advantage here.

A modern search engine can find a book title. A hand-maintained military-history site can explain why that book belongs beside five others and where the official record fits into the larger historiography.

CanadianSoldiers.com also links outward to digitized copies hosted by government and military-history projects, so the site acts partly as a map to material distributed across the web.

The tone is not flashy. It assumes the reader came looking for something specific and is willing to follow references.

That makes it useful in 2026 for exactly the same reason it was useful when it was built.

CacheRat’s 1,967 Ancient Web Domains research list includes specialist history sites because many of them function less like articles and more like annotated card catalogs.

That kind of site does not need a redesign nearly as badly as it needs to stay online.

Return to Canadian Soldiers

Posted on by

Ancient Web: Bubble Bobble Fans Dumped the ROMs and Built a 1990s Reference Site

The Bubble Bobble Info Pages are what happens when fans stop being satisfied with “here are some tips” and start treating an arcade game like a machine worth documenting.

Visit the Bubble Bobble Info Pages

Compiled by Stephen Tjasink and Paul Rahme, the site collects information about Taito’s 1986 Bubble Bobble: enemies, items, hidden rooms, cheat sequences, scoring behavior, version differences, endings, and level structure.

Some of it came from playing the game over and over.

Some of it came straight out of the ROMs.

Fan research before game wikis became default infrastructure

The level pages include positions, delays, enemy counts, hurry-up timing, special-item locations, bubble speed, and other values dumped from game data. The site even carries information for levels 101 and 102 beyond the ordinary hundred-stage progression, noting that level 102 appears to be a demonstration level.

Elsewhere, the authors decode hidden messages from secret rooms, document the title-screen sequences for “Power Up,” “Original Game,” and “Super Game,” and compare behavior across arcade, PC, PlayStation, Saturn, and other versions.

The special-item pages go deeper still, tracking not only what an object does but the conditions that make it appear.

This is reverse engineering with smiley faces.

The front page says the project was an attempt to collate as much Bubble Bobble information as possible. It also openly credits other FAQs, players, and people involved with the MAME driver when information came from elsewhere.

The original front page carries a 1998 last-change date, while later updates have kept some links and details alive. Tjasink’s domain itself says its index page has been a placeholder since 1995.

That mixture is useful: an old fan reference that was not sealed in amber but was maintained just enough to remain navigable.

CacheRat’s 1,967 Ancient Web Domains research list includes specialist fan sites because they often preserved technical knowledge long before anyone thought to call it a database.

A game can be cute and still deserve a ROM dump.

Return to the Bubble Bobble Info Pages

Posted on by

Ancient Web: SegaSoft’s E3 1998 Photo Dump Caught the Trade Show Before Social Media

Before every trade show became a flood of livestreams, influencer clips, corporate recaps, and phone video, somebody from SegaSoft went to E3 with a camera and made a web page.

Visit SegaSoft at E3, 1998

The result is still online.

The page says a group from SegaSoft attended E3 in 1998 and that the author took enough pictures to build a day-by-day record for people who could not be there. There are 103 images in total.

The archive begins before the show opens.

Setup is part of the history

The index separates the trip into setup days, a side trip to Six Flags, the three public E3 days, and the final breakdown. Dates run from May 24 through May 30, 1998.

That structure matters because polished event coverage usually starts when the doors open.

This page keeps the boxes.

It keeps the half-built booth, employee jokes, travel downtime, equipment problems, and the general mess around the presentation itself. The sample captions include lines like “Lots of HEAT.NET boxes” and “Lamont, does it work yet?”

Those are not historically important sentences by themselves.

Together with 103 photographs, they preserve the texture of a game-company trade-show appearance in a way a press release cannot.

SegaSoft itself is now a historical subject. The company was Sega’s PC and online-gaming arm in the 1990s, and HEAT.NET was part of its multiplayer push. A surviving employee photo dump therefore catches both a particular E3 and a company from a transitional period when PC networking, online services, and console culture were converging.

The page says it was last modified on December 3, 1998.

It still looks like it.

That is a compliment.

CacheRat’s 1,967 Ancient Web Domains research list includes pages like this because informal event documentation often becomes more valuable after the event stops being current.

Nobody needed a formal archive plan.

Someone just had a camera, a directory on a web server, and 103 reasons not to delete it.

Return to SegaSoft at E3, 1998

Posted on by

Ancient Web: Tim’s Home Page Turned One Man’s Life Into Hundreds of Travel Pages

Tim’s Home Page is what happens when a personal website is allowed to keep growing for decades instead of being replaced every five years.

Visit Tim’s Home Page

Tim says he has been online since the early Q-Link days. His site eventually accumulated family pages, a pictorial biography, local information about Indiana towns, television pages, and a travel section large enough to become its own archive.

The scale is the interesting part.

His travel index lists 70 trips dating back to 1980. Individual trips contain photographs, maps, daily travelogues, and video. The site index describes more than 300 travel pages and over 1,500 vacation videos.

That is not a blog category. It is a life log built before that term became normal.

The web as a filing cabinet for one person’s memory

The travel pages move from ordinary Midwestern road trips to Antarctica, Australia, New Zealand, Alaska, Europe, South America, Africa, India, China, and the Middle East. The presentation is simple: dates, locations, pictures, links, and commentary.

There is no attempt to turn every trip into a polished magazine feature.

The accumulation is the feature.

The old TV Showcase section adds another layer of web history. Tim explains that files from the original collection were deleted by his hosting company and that he had no interest in rebuilding the whole thing. A smaller selection survives, including pages for The Dick Van Dyke Show, All in the Family, The Andy Griffith Show, Frasier, Seinfeld, Gunsmoke, The Honeymooners, and others.

That blunt note about lost files is almost as useful as the surviving pages. Personal websites were archives, but they were also fragile. One hosting failure could erase years of work without producing a press release or obituary.

Tim’s site survives because enough of it kept moving forward.

CacheRat’s 1,967 Ancient Web Domains research list includes personal sites because they preserve material no institution would have thought to collect: one person’s vacations, favorite television shows, family photographs, and local observations, all held together by HTML and stubbornness.

The result is not tidy.

It is better than tidy.

It is specific.

Return to Tim’s Home Page

Posted on by

Ancient Web: The SS Richard Montgomery Site Built a Private Archive Around a Dangerous WWII Wreck

A few masts sticking out of the Thames Estuary are enough to support an entire private web archive.

Visit the SS Richard Montgomery Matter

The SS Richard Montgomery was an American Liberty ship that grounded and broke up in the Thames Estuary in August 1944 while carrying wartime munitions. Much of the cargo was removed, but explosives remained aboard and the wreck has been monitored ever since.

That much is established history.

The website built around it goes much further.

A research pile with an argument attached

SSRichardMontgomery.com collects an unusual mixture of primary material, technical documents, photographs, maps, press reports, parliamentary material, Freedom of Information links, and personal interpretation.

Among the material linked from the site are original ship records, an accident report, photographs of the wreck and unloading operation, risk-assessment documents, navigation charts, historical newspaper material, and a scanned ship log. One surviving log scan is visibly marked as a declassified United States merchant-vessel record from 1943.

The site also makes much stronger claims about the scale of a possible explosion, government secrecy, surrounding infrastructure, and other hazards. Those claims should be read as the site’s argument, not automatically as independently established fact.

That distinction is important because it is also what makes the page such a good old-web artifact.

This is not a neutral museum database. It is one person assembling evidence over years because he believes an official story is incomplete. The result is messy, repetitive, opinionated, and full of material that would otherwise be scattered across reports, newspaper archives, government pages, PDFs, and dead links.

The homepage still behaves like a hand-maintained research desk. New notes sit beside material from decades earlier. Technical references sit beside television clips. Serious archival documents sit beside blunt editorial commentary.

CacheRat’s 1,967 Ancient Web Domains research list includes sites like this because the old web was full of people who did not wait for institutions to organize a subject for them.

Sometimes that produced excellent indexing.

Sometimes it produced arguments.

Often it produced both on the same page.

Return to the SS Richard Montgomery Matter

Posted on by

Ancient Web: TextileAsArt Turned Antique Textiles Into a Hand-Built Web Catalog

TextileAsArt is the kind of website where the navigation itself starts to resemble a museum storeroom.

Visit TextileAsArt

Its catalog ranges across Chinese textiles, Pre-Columbian material, Central Asian ikat and suzani, Japanese embroidery, European textiles, Coptic fragments, Persian silks, Islamic textiles, Kashmiri shawls, costumes, rugs, tapestries, books, and related decorative arts.

The homepage announces more than a hundred added textiles as though the visitor already knows exactly why that is exciting.

That assumption gives the site its character.

Catalog first, polish later

TextileAsArt mixes several functions that a modern website would probably split apart.

It is an educational resource, dealer catalog, reference library, appraisal contact point, restoration-and-preservation directory, archive of previously displayed objects, and large collection of image-heavy subject pages.

Its site map reads like a taxonomy assembled by people who actually handle these objects: separate paths for textile regions, costume types, shawls, ancient materials, embroidery, rugs, carpets, Persian art, bibliography, services, and archived inventory.

The archive pages preserve older objects by inventory number even after they rotate out of the main galleries. That gives the site a useful second layer. It is not only presenting what is currently featured; it leaves traces of what passed through earlier versions of the catalog.

The site’s own “About” material argues that textiles can be used to teach history and art because cloth connects ordinary necessity, technical skill, trade, status, ceremony, and design.

The website accidentally demonstrates the same point.

Its categories cross geography and centuries because textile history refuses to stay inside a single museum department.

Visually, the site belongs to an older commercial and educational web: image slices, dense link groups, hand-built indexes, and pages that expect the visitor to browse rather than be guided through a funnel.

CacheRat’s 1,967 Ancient Web Domains research list includes many sites where a collector, dealer, or specialist gradually turned working knowledge into a public reference collection.

TextileAsArt is messy in exactly the productive way those sites often were.

There is always another drawer.

Return to TextileAsArt