Category: Ancient Web

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

Posted on by

Ancient Web: Runo’s Cloth-Doll Site Kept Handmade Pattern Culture on the Open Web

Runo’s NunoDoll site belongs to a species of Web page that used to be everywhere: one person makes things, photographs them, explains how, and keeps adding pages.

Visit Runo’s cloth-doll gallery

The site describes itself simply as a handcrafted doll gallery. The categories spread outward from there: boys, girls, anime characters, game-related dolls, life-size dolls, kimono dolls, nekomimi figures, mermaids, mermen and stuffed animals.

Every doll is handmade by Runo.

The gallery links into a broader network of sewing pages with instructions and patterns for dolls, toys, clothes and other fabric projects. Instead of separating inspiration from technique, the site keeps them near each other. You can look at a finished character, then move toward the construction side of the hobby.

That is how a lot of craft knowledge traveled online before tutorial platforms standardized everything into ten-minute videos.

The handmade Web fits the handmade object

There is a pleasant structural match between the subject and the website.

The dolls are one-off objects assembled from fabric, thread, stuffing and small design decisions. The site is also hand-assembled from galleries, menus, old links and specialized subdomains.

Neither is trying to look factory-made.

That matters because craft archives gain value from variation. A polished commercial catalog usually shows the best result. A personal craft site can show experiments, odd characters, niche interests and techniques that would never justify a product line.

Runo’s pages also preserve a particularly early-Web international flavor. Japanese craft sensibilities, anime fandom, sewing instructions and English-language navigation coexist without needing to become a single branded content strategy.

The result is useful to researchers looking at Web craft culture as well as people who simply want to make a doll.

Modern social feeds contain vastly more handmade work, but they are terrible at behaving like libraries. Search for an old pattern five years later and you may be hunting through dead image links, deleted accounts and algorithmic sludge.

A little personal domain can be much sturdier.

The page still says, in effect: here are the dolls; here is how I make things; go look around.

That remains a good design.

Explore Runo’s NunoDoll pages

Posted on by

Ancient Web: The Salem Witch Trials Archive Put the Primary Sources Online

The Salem Witch Trials Documentary Archive does the thing historical websites are best at: it gets the primary material out from behind the summary.

Visit the Salem Witch Trials Documentary Archive

The University of Virginia project collects court records, record books, personal letters, sermons by Samuel Parris, the diary of Samuel Sewall, historical works, maps and archival images related to the Salem witch trials.

That structure matters.

A general article can tell you what happened in 1692. A documentary archive lets you move between the accusations, legal records, personal writing and geography that produced the historical record.

The site includes maps of Salem Village, Andover and Salem, along with a map of accusations. It also links material from institutions including the Boston Public Library, Massachusetts Historical Society and the former Essex Institute collections.

This is early digital humanities that still works

The project credits Benjamin Ray as director and the University of Virginia’s Scholars’ Lab and Institute for Advanced Technology in the Humanities for technical development and support.

That puts the site in an important category of the older Web: university projects that treated the browser as a serious research interface before “digital humanities” became a standard institutional phrase.

The design is straightforward because the documents are the point.

Court records are not wrapped in endless commentary. Letters and sermons are not buried beneath a slideshow. The archive organizes sources so researchers can follow names, places and events themselves.

That approach ages well because the underlying value does not depend on a publishing trend.

It also creates a useful distinction between historical interpretation and historical evidence. The archive includes introductions and contextual material, but the documents remain visible as documents rather than disappearing into a historian’s paraphrase.

For students, genealogists and researchers, that is the difference between reading about an event and being able to inspect part of the evidentiary trail.

The Web has lost plenty of ambitious university projects to redesigns, broken databases and expired grants.

This one still behaves like a reference desk.

Explore the Salem Witch Trials Documentary Archive

Posted on by

Ancient Web: Antique Bottles Turned Glass Trash Into a Reference Library

Old bottles are a perfect Internet subject because the difference between junk and history is usually one tiny molded word on the side.

Visit AntiqueBottles.com

The site grew around exactly that kind of collector knowledge: bottle shapes, embossing, regional merchants, glass colors, makers, liquor bottles and examples submitted by collectors.

One surviving section, the Antique Whiskey Bottle Hall of Fame, reads like a cabinet of extremely specific local history.

A clear half-pint flask from Anaconda, Montana is identified by the embossed Johnson & Burke Turf Exchange address. Another entry documents a Spokane liquor bottle associated with Jimmie Durkin and preserves a short history of his business beginning in the 1890s.

That is the interesting thing about bottle collecting.

The object is small. The research around it is not.

A bottle can be a business directory with dirt on it

Embossed names and addresses can tie a piece of glass to a saloon, pharmacy, brewery, mineral-water company or regional merchant that disappeared a century ago.

Collectors then compare form, manufacturing method, color, closure, lettering and provenance to narrow down age and identity.

Websites like AntiqueBottles.com were built for that distributed work. One person might own the bottle. Another has an old city directory. Someone else recognizes the mold style. The Web becomes a place where those fragments meet.

The site’s gallery format also preserves a pre-social-media collecting culture. Contributors send photographs and descriptions to be added to topical pages rather than posting them into a feed where they vanish beneath newer material.

That permanence is what gives the archive research value now.

A modern auction listing is optimized to sell one object once. A collector-built reference page tries to make the identification useful the next time somebody digs the same bottle out of a basement, privy pit or box at an estate sale.

CacheRat’s Ancient Web project keeps finding this pattern: ordinary enthusiasts quietly created searchable indexes for subjects that institutions barely touched.

In this case, the database is made of old glass.

Explore the AntiqueBottles.com reference

Posted on by

Ancient Web: Sloperama Built a Mah-Jongg Reference Library Out of FAQs

Tom Sloper’s Mah-Jongg site is what happens when a FAQ refuses to remain frequently asked and instead becomes an entire field manual.

Visit the Mah-Jongg FAQs

The archive traces back to late-1990s newsgroup material and eventually moved to Sloperama.com. Update logs show FAQ revisions in 1999, a move from thegameguru.net to Sloperama in December 2000, and conversion of material into HTML in the early 2000s.

The subject exploded from there.

The current index covers basic rules, competing rule sets, books, computer mah-jongg, terminology, tile materials, mystery tiles, restoration, strategy, etiquette, Chinese Official scoring, American rules, Japanese mah-jongg, tournaments, teaching and the history of the game.

There is even a “Rosetta Stone” of mah-jongg terminology because one culture’s perfectly ordinary game vocabulary becomes somebody else’s translation problem.

The history pages are unusually serious

Sloper’s historical sections distinguish documented facts from folklore.

The site discusses early sets, precursor games, Chinese Classical rules, regional variants and the uncertain shape of what it calls proto-mah-jongg. One section notes that no written rules survive for the earliest known tile configurations, making confident claims about the game’s exact “original rules” difficult.

That is a useful habit for any history site: say what is known, then say where the evidence stops.

The archive also preserves the culture around the game. Tournament trip reports reach back through the 2000s. Bulletin-board archives record years of reader questions. Strategy pages contain dated update logs and contributions from players in different traditions.

The site’s old architecture still shows through. Frames, bulletin boards and hand-maintained indexes coexist with warnings that the forced switch to HTTPS broke some older elements and that the site is gradually being made more mobile-friendly.

That technical friction is part of the record.

A modern knowledge base would probably flatten all of this into search results and collapsible cards. Sloperama instead lets you see how the reference accumulated, question by question and revision by revision.

That is useful for researchers because the page itself has a history.

Explore Tom Sloper’s Mah-Jongg archive

Posted on by

Ancient Web: NOHAB.hu Documented the Hungarian Career of a Swedish Diesel Icon

NOHAB.hu is the sort of railway site where a locomotive’s paint scheme, engine swap and last regular train can each earn their own paragraph.

Visit NOHAB.hu

The site centers on diesel locomotives built by the Swedish firm Nydqvist och Holm AB—NOHAB—and especially the M61 class used by the Hungarian State Railways.

Its historical starting point is 1960, when a 1,750-horsepower NOHAB demonstrator arrived in Budapest for trials. The locomotive used major General Motors Electro-Motive Division components, including a 16-567C prime mover, generator and six traction motors.

That demonstration led to one of Hungary’s most recognizable diesel classes.

The site then follows individual locomotives almost obsessively.

A railway photo archive with memory attached

The galleries are not anonymous train pictures. Captions record train numbers, locations, dates, overhauls, engine substitutions and changes in service.

M61 006, for example, appears on ordinary trains, track-renewal work and special movements. The page records photographs from the late 1990s and early 2000s, often with both 1024×768 and smaller 800×600 versions—an excellent little reminder of when bandwidth was a design constraint worth mentioning explicitly.

M61 010’s page documents both older and newer liveries and notes that its prime mover came from scrapped locomotive M61 003. A December 22, 2000 movement is described as part of the end of the regular NOHAB era, when the remaining locomotives were transferred away from Tapolca.

The site also spreads outward from Hungary. Separate galleries cover NOHAB-GM locomotives in Norway, Denmark and Sweden, including Swedish T41, T43 and T44 series machines and preserved demonstrators.

That makes NOHAB.hu useful to more than railfans looking for pretty photographs. It preserves fleet relationships, builder histories, component lineage and the everyday operational context that formal locomotive specifications rarely capture.

Old railway websites are often strongest exactly where official histories are weakest: what happened to this particular machine after overhaul, what hauled train 9709 that morning, which engine ended up inside which surviving locomotive.

Those details are too small for most books.

They are perfect for the Web.

Explore the NOHAB locomotive archive

Posted on by

Ancient Web: V8SHO.com Turned a Ford Engine Failure Into an Owner-Built Database

V8SHO.com looks like an old car-club site until you hit the camshaft pages and realize the owners built their own failure database.

Visit V8SHO.com

The site focuses on the 1996-1999 Ford Taurus SHO with the 3.4-liter DOHC V8. Among the ordinary enthusiast material—technical service bulletins, maintenance notes, parts information and owner stories—is a sprawling archive about camshaft sprocket failures.

The problem was ugly because the engine used tubular camshafts with pressed-on sprockets. When a sprocket slipped, valve timing could be lost and the result could escalate from a strange ticking noise to bent valves and a ruined engine.

V8SHO.com did what enthusiast communities often did before social platforms swallowed the conversation: it kept the reports on permanent pages.

The archive preserves both evidence and argument

A cam-failure index explicitly says the reports are raw owner information, not independently verified diagnoses. That disclaimer is important. The site assigns case numbers to reported failures but does not pretend each one was disassembled and confirmed by the webmaster.

Individual pages preserve mileage, symptoms, repair estimates, dealer experiences and sometimes VINs or technical observations. One indexed case reached Cam Failure #669 in 2006.

The archive also preserves Ford Technical Service Bulletin 03-14-1, dated July 3, 2003, which described an in-vehicle Loctite procedure for the camshaft tube-and-gear interface on 1996-1999 Taurus SHO models.

The owner community was not impressed.

V8SHO.com retains pages arguing that Loctite was not a permanent remedy and documenting failures owners said occurred despite attempted preventative work. Other pages discuss welding the cam sprockets, diagnostic symptoms and the mechanical chain of failure.

Whatever conclusion a modern reader reaches, the historical value is obvious. Manufacturer language, owner reactions, repair-shop reports and community troubleshooting survive together instead of being scattered across dead forum threads.

That makes the site useful beyond one Ford model.

It is a case study in how early Web communities built informal technical knowledge bases when official support did not answer the questions owners cared about.

A modern social-media group can generate more posts in a week.

V8SHO.com is still easier to research twenty years later.

Explore the V8 Taurus SHO archive

Posted on by

Ancient Web: Ju388.de Built a Digital Museum Around One Rare Junkers Aircraft

Ju388.de is what happens when one obscure aircraft gets the full personal-Web treatment instead of a paragraph in a general aviation database.

Visit Ju388.de

The Junkers Ju 388 was a high-altitude development of the Ju 188, itself descended from the Ju 88. The site breaks the aircraft into variants, technical systems, production history, drawings, photographs and the story of a surviving Ju 388 L-1 airframe.

That survivor is Werknummer 560049.

According to the site, American forces found it at the end of World War II in Merseburg. It was transported to the United States aboard HMS Reaper with other captured German aircraft, tested at Wright Field under the designation FE-4010 and later transferred to what became the National Air and Space Museum collection.

The aircraft survived largely because nobody had room to display it.

It remained disassembled in storage.

The site goes deeper than a spec sheet

Christoph Vernaleken’s archive includes exterior and interior photographs, cockpit views, technical drawings, historical images, version comparisons, a U.S. Air Materiel Command report and a map of production sites.

The technical section records a 22-meter wingspan, BMW 801 TJ-0 engines, a pressurized three-man cabin and a service ceiling around 12,800 meters for the Ju 388 L-1.

The page also preserves the weird engineering details that generic summaries tend to erase. Wing de-icing used hot air from the engines, while the tailplane used a separate fuel-fired Kärcher heater. The aircraft carried a remote-controlled rear gun installation rather than a conventional manned rear position.

The source list is equally valuable. Vernaleken cites 1944 and 1945 Junkers documents, technical drawings and German military material rather than presenting the page as unsourced enthusiast lore.

That matters because rare-aircraft history is full of repeated errors. A surviving primary-document trail gives researchers something to work backward from.

The site was last visibly edited in 2011, and its HTML still carries the flavor of the period: fixed backgrounds, image rollovers, hand-built navigation and separate English and German pages.

But the underlying research has aged much better than the markup.

That is the recurring Ancient Web pattern: old presentation, durable subject matter.

Explore the Ju 388 archive

Posted on by

Ancient Web: Nick’s Model Tanks Preserved a Lifetime of Scale Modeling

Nick’s Model Tanks is the kind of website you only get when somebody has been doing the hobby longer than the Web has existed.

Visit Nick’s Model Tanks

Nick Turner documents military models in 1/76, 1/72, 1/35, 1/87 and other scales, but the interesting part is the time depth. Some models were built in the 1960s and 1970s, then photographed, repaired or rebuilt decades later.

A British Mark I tank page describes an Airfix kit built around 1967 and restored in 1997 after the old flexible tracks began damaging the plastic. Elsewhere, Turner shows an Airfix ambulance built around 1969 with the glue marks still visible.

He leaves the scars in.

That makes the site more useful than a polished gallery of perfect finished kits.

It is also a record of how modeling changed

Turner explains old scale conventions, including the overlap between 1/76 Airfix armor and 1/72 aircraft, and notes how 1/35 opened much more room for detail. He defines hobby shorthand such as OOB — “out of box” — and talks about seams, join lines, brittle plastic, disintegrating rubber tracks, resin kits, photo-etched parts and scratch-built conversions.

The site repeatedly compares old builds with later work. A first 1/35 Abbot self-propelled gun from the 1960s survives alongside newer restorations and more sophisticated kits. A French AMX-30 model started around 1975 and was not completed until roughly twenty years later.

There are also hundreds of large photographs because, as Turner points out, modelers need to see the mistakes as much as the success.

One particularly old-Web section discusses the vanished AFV Interiors Web Magazine. Turner saved much of that site’s material before it disappeared in 2004, but explicitly refuses to republish the whole archive without permission. Instead, he offers to help individual researchers with specific vehicle material.

That little copyright discussion is itself Web archaeology: one hobbyist trying to preserve another hobbyist’s vanished work without pretending abandoned pages are automatically free property.

The result is not just a tank-model site. It is a personal record of techniques, materials, kit manufacturers and changing standards across more than half a century of modeling.

Browse Nick Turner’s model archive

Posted on by

Ancient Web: Rodent’s Vengeance Kept Rodent’s Revenge Alive in Java

Some old Windows games survive because a company remasters them. Others survive because one person refuses to let a mouse be eaten by cats forever.

Visit Rodent’s Vengeance

The original Rodent’s Revenge was developed by Christopher Lee Fraley for Microsoft and appeared in the 1993 Microsoft Game Pack for Windows. Its rules were simple enough to explain in a paragraph: move a rodent around a grid, push blocks, trap cats, and turn the trapped cats into cheese.

The problem was not the design. The problem was time.

According to the Rodent’s Vengeance site, the original game had effectively remained static for decades. Mike Reynolds, who had played it since the mid-1990s, rebuilt the idea in Java after completing a post-baccalaureate computer-science degree.

What began as an academic challenge turned into a preservation project with opinions.

The clone fixes the things nostalgia forgets

Rodent’s Vengeance adds features that old Windows software never had to anticipate: mouse control, alternate keyboard layouts, save-and-restore, adjustable board sizes, selectable color schemes and stronger cat pathfinding.

The site even documents the behavioral differences. The original cats mostly move directly toward the player and can be confused by obstacles. The Java version can give them a larger search area so they are less likely to behave like furry missiles with no concept of a corner.

The current download page offers Java .jar files for several visual variants and explains how to run them on Windows, macOS, Linux and other systems with a compatible Java runtime. Beta Release 2.77, dated December 15, 2023, added mouse control and a one-click scamper mode.

That makes this more interesting than a fan remake hidden in a repository.

The website explains what the old game did, what broke as computing changed, and exactly what the replacement changes. It preserves the gameplay logic and the reasoning around the preservation.

That is useful game history.

A screenshot tells you what Rodent’s Revenge looked like. A working reimplementation shows you how it behaved.

CacheRat’s Ancient Web research list contains a lot of sites built around this same instinct: if a useful or beloved thing is no longer maintained, somebody on the Web quietly keeps it moving.

In this case, literally.

The mouse still runs.

Explore Rodent’s Vengeance and its downloadable Java builds

Posted on by

Ancient Web: RetroArchive Preserved the Working Files of BBS-Era Computing

RetroArchive preserves an endangered species of computer history: the little utility ZIP file somebody absolutely needed in 1992.

Visit RetroArchive

The site stores old software, documentation and hardware material in directory structures that still resemble the collections they came from. One of the clearest examples is its BBS Toolkit CD-ROM archive.

A single RemoteAccess directory contains 227 files.

Not screenshots of the files. Not a nostalgic list of program names. The actual catalog: filename, size, date, extraction status and short description.

Entries include menu editors, callback verifiers, bulletin generators, file-list managers, language packs, message tools, protocol installers and utilities for RemoteAccess 1.x and 2.0.

The dates cluster around the era when running a bulletin-board system meant assembling the system yourself out of a main BBS package plus dozens of independently written tools.

Software archaeology works better when the filenames survive

Names such as RA_111.ZIP, BWAVE205.ZIP, RACE200.ZIP and LWMENU09.ZIP are not pretty metadata.

They are historically useful metadata.

Old software research frequently begins with exactly that kind of fragment: a filename from a disk label, a mention in a text file, a screenshot of a directory listing or a sysop remembering that a door program had “RA” in the name.

RetroArchive lets those fragments resolve into something concrete.

The same site also preserves manuals and technical documents. One surviving AMPRO Little Board manual describes a compact Z80A CP/M computer with 64 KB of RAM, serial ports, a parallel printer interface and support for 5.25-inch floppy drives. A Turbo Pascal tutor preserves programming material that assumes CP/M, MS-DOS and terminal-oriented development as ordinary environments.

That mix is important.

A software archive without manuals leaves you with mysterious binaries. A manual archive without software leaves you with instructions for ghosts.

RetroArchive keeps both sides close enough together that researchers can reconstruct how these systems were actually used.

The website itself is not especially interested in presentation. It exposes directories, indexes and files.

That is appropriate.

BBS culture was built from directories, filenames, archives and terse descriptions. Turning all of that into a modern gallery would make it more attractive and less authentic at the same time.

For the Ancient Web project, RetroArchive is a good reminder that digital preservation is not only about famous software.

Sometimes history is a 15-kilobyte utility whose entire purpose was to tell tomorrow’s caller that somebody had logged in today.

Explore the preserved software and documentation at RetroArchive