A developer came to me a week ago with a project they’d been working on for over a year. The proposition of what they offered and the importance of what it would mean to historical software at Internet Archive was so compelling that within 48 hours, we’d announced it to the world.

The site is DISCMASTER.TEXTFILES.COM, and within its stacks lie multitudes of previously hidden software treasure, and a directed search engine that makes it a top-notch research tool.

More than a fascinating site, though, it represents some philosophies regarding the Archive’s stacks that are worth exploring as well.

The first thing that strikes a visitor to the site is either how strange, or how nostalgic it looks. The site is strikingly simple and references the first few years of the world wide web, when backgrounds were grey by default, and the width of the screen was almost always under 640 pixels. Same with the link colors, and use of (to the modern era) small icons next to the words and links. This is a version of the world wide web long gone.

However, underneath this simple exterior beats the heart of a powerful search engine and an astounding amount of processing that has analyzed millions of files to make them easy to interact with. If your area of research or interest is vintage/historical software, we’ve all been handed a top-class tool to discover long-lost files and bring them back instantly.

A Quick Reminder about CD-ROMs

From (very roughly) 1989 through to the early 2000s, CD-ROMs (and later DVD-ROMs) were one of the primary ways to transfer heaps of software or large-sized programs to end users. Instead of spending hours or literal days transferring software you may or may not have wanted after you received it, you could go to stores or on-line and purchase a plastic disc that contained between 600-700 megabytes of information on it.

The potential of this, in fact, was so strong, that there was an entire industry of providing databases, news summaries, and even all-digital magazines using this format. Booklets of CD-ROMs became resplendent, and libraries could allow patrons to check out these discs to do research with them.

Besides these more institutional compilations, an industry rose up of companies compiling software, artwork, music and more and selling them to end users. Companies with names like Walnut Creek, Wayzata, Valusoft, and Imagemagic would have catalogs of CD-ROMs to buy. Starting out with software from bulletin board systems and gathered from FTP sites, these CD-ROMs quickly ran out of easy-to-find material to fill, and an era of “shovelware” began, allowing these products to claim “thousands of files, gigabytes of materials” while pulling from more and more out-of-date sources.

As websites, torrents and other means of transport brought the era of physical media for software to a close, the world was left with a finite, contained pile of titles that had come out on CDs. And, as luck would have it, people have been uploading those out of date files to the Internet Archive for years.

The Final Piece

Therefore, sitting on the Archive, are tens of thousands of these CD-ROMs of the past. And for a very long time, it’s been possible to download a Disc image, analyze its contents, search for useful or potentially interesting items, and then find a way to make them work again.

That last piece, in fact, is the hardest – not just knowing where the files you’re looking for are located, but to be able to browse them without a massive host of helper applications scattered to the four winds. There are dozens of archive types, dozens and maybe hundreds of multimedia formats, and, even more frustrating, archives within archives – making everything that much harder to find.

DiscMaster has fixed this.

Within the search engine is the ability to find millions of files, categorized by type or size or date or extension, and then be presented them instantly. Three decades of computer software with layers upon layers of obfuscation are brought immediately to the top.

The developer wrote applications to grind through the contents of a CD-ROM and present them with previews that wouldn’t require anything but a browser to see. This can take hours to pull out of a single CD-ROM, but the results are breathtaking.

Audio and music files play in the browser. Flash, IFF, Bitmaps, Fonts and more display in preview. Macintosh, PC, Commodore, Atari and more are presented simply, without a mandate to track down the proper utility to figure out what they are.

In other words, vintage and historical software is back from the obfuscated darkness.

In the short time that Discmaster has been online, success stories are appearing. Authors are finding shareware programs they lost track of decades ago. Original versions of software that were thought impossible to track down just pop up in the search engine. And organizations dedicated to creating catalogs of now-dormant formats are suddenly handed a thousands-of-items to-do list on a silver platter.

The Philosophy of the Support Site

The ramifications and discoveries from Discmaster are going to be coming for a very long time – even if a researcher has a light memory of something they’re looking for, the search results will guide them in the right direction faster than ever before.

But beyond that, this site shows a different approach to the Internet Archive’s materials that’s worth seeing more of.

With over 100 petabytes of data, representing a mass of materials with all sorts of containers, metadata, and approaches by contributors, the Internet Archive has to be as general as possible. This generality extends to the presentation, search engine, and storage of the items.

It is a major effort to ensure the data stays secure, the metadata is searchable, and the ability to upload nearly anything results in a usable item details page.

But that’s kind of where it has to stop.

It’s asking an awful lot to both maintain an entity like this, and also design, say, a specifically-geared site for a relatively smaller set of people and needs. It can be done, but when energy and funding are limited, it’s sometimes best to stick to basics.

Discmaster shows one way it could be done. After working hard on its specific set (software from CD-ROMs), the entire site is constructed with its singular goal in mind. If it’s not obvious, the simple, almost-no-javascript and straightforward design lends itself to an entire family of browsers that run on those original machines. You’ll be able to download Amiga software through your Amiga, your Atari software to your Atari and so on. A thousand little touches and flourishes live easily on this custom experience – because it has the freedom to allow them.

Perhaps seeing Discmaster in action will encourage others to interact with the Internet Archive as a pool, a container of resources that could receive some of the powerful analysis along specific lines. If they can then be fed back to the Archive at the end, even better; but let a hundred supporting sites bloom.

Meanwhile, enjoy the history of software – it just got a lot easier to find.

A Small Addendum Regarding Emulation

After this announcement came out, a not-insignificant amount of people have come forward to ask some form of:

You’re the Emulation In The Browser People – will DISCMASTER allow you to emulate the programs that are found in these floppies and CD-ROMS?

The short answer is no, there are no current plans to do emulated previews.

The longer answer is that the wonderful emulation in the browser that the Internet Archive has covers over the amount of work that needs to be done in selecting, refining, and in some cases modifying original programs to make them work. If a program requires all of Windows 3.1 installed, for example, someone went through the process of determining that, configuring the item to know to load Windows 3.1, and then added custom settings in the item to ensure it would all boot up correctly. Often this work can be automated to a degree, but the time involved is considerable.

Multiply these issues by the dozens of platforms that are emulated, and you can see why it would be more trouble than it would be worth. Additionally, some programs just don’t make sense to be emulated – running a printer utility “in the browser” will probably just show a prompt and nothing else, as it is loaded in the background – many, many programs of the past don’t make sense without additional context.

A much more likely scenario will be DISCMASTER revealing long-lost vintage software that is so interesting and/or fun that it will get uploaded to Internet Archive separately and those configurations done to allow it to be played in the browser.

If you find interesting items along DISCMASTER’s millions, feel free to contact me, Jason Scott, or take a shot at uploading the program yourself and doing the configurations.

There are parts of technology history (frankly, any history) that are thought to be critical to telling the story, and utterly lost. Pieces and fragments will rise up out of the darkness, but a cohesive collection of what once made up a chapter will be thought gone forever.

Sadly, this happens a lot.

But in one special exception, the Computer History Museum found itself with an opportunity to seize the moment.

Compuserve is considered to be the first major online service in the United States. Founded in the era of “time-sharing” services (paying to use a computer during the main owner’s quiet hours), this subsidiary of Golden United Life Insurance moved from 1969 to 1979 in the kind of obscurity befitting a simple business-to-business service providing access to PDP mainframes.

This all changed in 1979 with the rebranding of Compuserve Information Service (CIS), which marketed itself to mainstream computer users, providing chat, games, and storehouses of information for an hourly fee to who ever could afford the phone bills and equipment to do so. It is here that Compuserve (and later services like The Source, America On-Line and Prodigy) brought a bulk of folks online for the first time.

This service flourished through the 1980s and 1990s, and in what should be considered a reductive and surface description of the situation, slowly broke apart via acquisitions, shifting priorities and the dominance of the World Wide Web providing many aspects of what Compuserve had previously done exclusively.

By the time this 2000s-era chapter was over, Compuserve was more a brand and a memory. But it had still reigned in the minds of many as the beginning, the launching pad for a lot of what people came to expect the online world to provide.

It was assumed most of the history of Compuserve was gone – the hardware, software and documentation scattered to the winds.

Not so.

As explained in this blog entry that literally reads like a movie script, the Computer History Museum has acquired, sorted, and added a major amount of Compuserve’s archives to their stacks. The collection had been sitting for decades, and was soon to be disposed of, when it was offered to CHM and they accepted.

Among these items that have been recovered are documents that are being given to the Internet Archive to scan and place online – instructions on how to operate a Compuserve service.

These opportunities to recover assumed-lost materials are extremely rare, but hope springs eternal that in rooms, attics and file cabinets around the world, there possibly lurk further discoveries, and happy endings. Almost like a movie.

11 years ago, on the site longbets.org, a friendly wager was made between two mavens of the web: Jeremy Keith and Matthew Haughey.

The bet, to be revisited a decade and a year later, would be whether the URL of their wager at Long Bets would survive to a point in the semi-distant future.

That is, this day, February 22nd, 2022, (2/22/2022).

As of this writing, the URL absolutely has survived.

Therefore, the Internet Archive shall receive a $1,000 donation from Mr. Keith and Mr. Haughey ($500 apiece), provided from an escrow account that has held the funds since the day of the wager. (We shout out to the Bletchly Park Trust, a worthwhile historical organization, who will not be getting the donation but who are deserving of yours.)

It would be easy enough to declare it a win for the idea of “the web” and that regardless of concerns brought up about the Internet’s ongoing issues, we can still find hope. So certainly, let us all applaud that things worked this way, and the URL’s 11-year consistency is a bright beam of light, online.

In many ways, however, the bet is at best a bittersweet victory, and at its darkest interpretation, a small oasis in a desert.

To understand The Long Bets, you need to understand The Long Now.

The Long Now Foundation is a non-profit meant to be an organization geared towards projects and approaches to thinking that chronologically leave the average human lifespan in the dust; focusing on 10,000-year timelines and solutions to problems of sustaining cultural contexts for a hundred lifetimes and beyond.

Currently, the Long Now and its ideals are expressed in both a very nice performance and drinking space called The Interval, and a number of stylish projects and websites to bring this realm of thinking into focus.

The most prominent and first major project was the Clock of the Long Now, a project to make a time-accurate clock that would function for 10,000 years. The as-yet incomplete project goal is to build a clock deep in a mountain range and set it off, ticking occasionally (but on time, doing so) for the next ten millennia.

Other projects follow this approach as well, ranging from delicious to provoking. A re-imagining of the Rosetta Stone, a language translation service, a manual for civilization, a mountain land purchase, and others in these themes.

Among these variant projects is Long Bets.

It is a facilitation of long-term thinking, of providing a neutral, fair and equitable way for years-in-the-future bets to be made between parties, each contributing funds towards the prize. It is traditional that the recipient of the prizes be organizations not run or controlled by either bettor.

Browsing the betting page, the bets range from the humorous to the aspirational, from specific sports outcomes to predictions around space travel, vehicle autonomy and economics. They’re a joyride of thought and conversation starters, as they’re meant to be.

And among them is Bet #601.

Jeremy Keith and Matt Haughey are both veterans of The Web as it has historically been described; each has had their voices heard to crowds online and off, describing the nature of websites. Their careers have (deservedly) benefitted greatly from the power of interlinked websites.

They both recall the start of the world wide web, as well as internalizing the rules and mores that followed its birth. They were well-qualified to debate on the longevity of URLs and the position that a specific URL would hold across time.

That said, Keith was skeptical. Haughey was optimistic.

Like a lot of its neighbors, Bet #601 is too clever by half; the bet states that it is won or lost depending on the availability of the bet at the URL the bet is hosted at. That is, two situations exist to judge the outcome: Either the URL https://longbets.org/601 exists, at which point the bet is lost, or it does not exist, at which point the bet is won.

(Strikingly, if the bet had been won, the Internet Archive would possibly be the only place to browse the site in its original form, where it would have then helped prove the funds should go to Bletchly Park Trust. The continued reliance on the Wayback Machine as the vault of the Web’s lost memories would have persisted, in a very sharp and slightly less financially-beneficial way. Such is the price of memory.)

For the record, here are the statements made by each bettor about their arguments for the wager:

Jeremy Keith:
“Cool URIs don’t change” wrote Tim Berners-Lee in 01999, but link rot is the entropy of the web. The probability of a web document surviving in its original location decreases greatly over time. I suspect that even a relatively short time period (eleven years) is too long for a resource to survive. I would love to be proven wrong.

Matthew Haughey: Though much of the web is ephemeral in nature, now that we have surpassed the 20 year mark since the web was created and gone through several booms and busts, technology and strategies have matured to the point where keeping a site going with a stable URI system is within reach of anyone with moderate technological knowledge. My oldest sites are going on 13 years old at the time of this bet and the original URL scheme still functions via 301 redirects to a final format we selected about six years ago.

This should be it.

But it’s worth noting how the context of this bet has changed over time. And issues with the continued evolution of the web strike at heart of the point the bet was trying to make.

THE URL:

The Long Now Foundation, intending to maintain its footing for as long as absolutely possible, has a very vested interest in its URLs staying stable. Between hosting structure, the setup of the webpages themselves, and maintaining clean, static URLs (longbets.org/601 is a very simple address, lacking any ornamentation or dependence on programming language extensions or dynamic rendering). The domain name longbets.org is registered until June of 2022 as of this writing, but was registered in June of 2001, twenty years ago, which bodes well for continued survival.

If you’re going to bet that a URL is going to stick around, on a website run by an organization that expresses its character by the longevity of its projects, staking your bet on a specific URL from that organization is a pretty safe bet.

THE WEB:

Both of the parties in the bet clearly think of “the web” as being a set of interacting links between websites, but even by 2011, the idea of a “website” was beginning to experience direct collision with the ever-centralizing, ever-shifting audience of online life. Mobile access is a quirk in the 1990s, an oddity growing into a majority in the 2000s, and now, in the present day, phones with screens are the “home computer” of vast percentages of internet patrons.

In the interconnection of the world, it is harder and harder to think of a “website” where “platforms” rule the roost. A user is more likely to have an account name, or a public identity, than to ever utter the phrase “http” in their daily activity, or maybe even their year. The clear goal of many firms is to dissolve the consideration of the URI or URL, with many of the previous protocols of the earlier Web forgotten. The question becomes less of “will this URL survive” and more of “will the idea of the URL survive?”

THE IDEA OF LONGEVITY AND CHANGE FOR DIGITAL DATA:

Finally, the overarching fact of the situation is that sites like Long Bets are part of a philosophy of the web that is rapidly shrinking. Points of data and dependable signifiers of content and individuals were once the destination. That’s long changed; they are but stops along the way, flotsam and jetsam that ride in nebulous platforms that dominate online life. While Jeremy Keith and Matt Haughey maintain personal websites, they have rapidly become like homesteads that jut out in the center of towering skyscrapers and apartment blocks. Future generations will think of “the web” as much as they think of “the roads”; intensely interest to a few, below the watermark of consciousness to the rest.

As we move into this even-more-ethereal version of the Web, where objects, materials and locations possess data as much as pages and links we ever did, the Internet Archive will do its best to keep up and grow to match the challenge.

But what a challenge it shall be. Bet on it.

Continuing our tradition of releasing new sets of emulated items around the holidays, the Internet Archive has added a new collection: THE JACKPOT LOUNGE.

Previous sets of items, including arcade machines, handheld toys, computer software and flash animations, all represent thousands and in some cases tens of thousands of individual items from history, all playable in the browser.

The Jackpot Lounge is much more focused and refers to one specific group of coin-operated games: Gambling Machines.

Not too soon after video games began replacing mechanical coin-operated games on the midway, bars and other locations, games with a gambling theme or straightforward gambling payouts began to arrive. The same arguments for video games taking over (cheaper maintenance, more dependable, easily upgradable) applied to these potential replacements for slot machines.

While such games would exist in a grey zone for a few years, Poker and Slot/Fruit machines of a video game nature quickly fell under the jurisdiction of gaming authorities in different countries. In some cases, however, it is clear looking through the code of these games that a select few cheated or ended winning runs quickly.

Still, there was little to compete with the complete automation and dependability of these machines, and over time they took over gambling houses and areas where games of chance were legal. Even the venerable and charming mechanical horse-racing games began to fall compared to a simple horse-racing videogame like Status Fun Casino:

As more and more pressure came down to ensure the slot and poker machines had dependable random number generation and payouts, the games became more homogenized, only reflecting changes in sound and visuals while letting the rules stay dependably the same.

From those early beginnings, the increase in quality of sound and video with videogames extends to gambling machines – the colors are bright, the sounds intense, and all are intended to catch your eye and bring you over to play “just a few rounds”.

Reflecting that we are now closing in on a decade of emulation at the Internet Archive, it seems a good time to bring in this class of machines, which have at the start of each game a very complicated start-up process. Reflecting that actual money could be involved with a machine and the tallies of what it was paying out for the day would be very important to a casino, these games have very complicated processes for starting up. By pressing these keys, you are simulating Audit and Jackpot keys being inserted, sign-off buttons are being pressed, and that the metal “cage” door of the machines were closed and locked. Thanks to a set of volunteers, however, these extremely un-intuitive keypresses have been documented for you.

Because of the nature of this genre of machine, many of the 500+ machines will look rather similar: made by the same companies, with only minor modifications to the code to reflect different rules or providing compatibility with gaming commission chips (which would not be manufactured by the game companies, but supplied to them under authority of the commission). Others, however, will seem like strange one-off creations that you can’t imagine attracted anyone to play them – except they did.

One particular fascinating set of machines exists in this collection: Stealth Gambling Machines, meant to look like one kind of video game, but secretly playing another, based on when a secret switch (or secret coin slot) was used. In countries where hosting a gambling machine of any sort held severe penalties. For example, this game acts like Breakout but is actually a poker machine.

And finally, let’s just remember that sometimes, the game is out to get you. The game Tetris Payout will wait until you are winning a bit too much, and then intentionally throw the worst pieces at you to ensure your game comes to an end early.

So go ahead, check out the lounge, and walk through decades of electronic coin-operating casinos. You can’t win real money…. but you can’t lose it, either.

A very large thank you to Xarph and others, who played through these hundreds of machines to learn the by-machine instructions for getting them into a standard playing mode. The process was weeks of work and incredibly appreciated.

If coming into contact with gambling machines causes stress or issues for you or your family, contact Gamblers Anonymous.

An arbitrary but still amazing milestone has passed:

The Internet Archive now has over one million manuals, instruction sheets and informational pamphlets in its Manuals collection. They range into every field of study or product and extend back, in some cases, well past a century!

People have been uploading manuals of every stripe into the Internet Archive for well over a decade, not to mention instruction booklets and related works in collections of partner organizations which are not in the main collection. In the past few years, efforts to mirror both manual repositories out in the world as well as documentation collecting by individuals have put our materials into the stratosphere of this staggering number.

While we gaze upon a nine-digit mountain of manuals, let’s talk about why this is so important and what it means.

Since we’ve had technology and tools, we’ve had instructions on how to use them. Passed along by demonstration, discussion or written forms, the critical link between having possessions and making the most of them has been the unit of knowledge called The Manual.

It’s a very general idea, and one that can range wildly in terms of depth, quality, and approach. Some items hand you a scant single printed page with poorly constructed diagrams. Others are multi-volume tomes that give exacting detail down to the smallest theory of operation. (The manuals for a Boeing 747 airplane weigh more than a 747!)

Considering the manual, one impression might be that they’re only of interest to someone in possession or with interest in the specific item or procedures being addressed.

But manuals are much more than just instructions of operation, or a listing of the components inside a product. They’re windows and insights into the priorities and approaches that companies and individuals take with the tools and goods they sell. And in many cases, the artistic and visual efforts to make an item clear to the reader has led to truly breathtaking visual feats.

With this many manuals and documents, a person trying to find the exact manual they need for something they own will always be a little difficult. Here’s some quick suggestions.

• First try searching for the product name, with just the company and the model number. If the company does not make that many items (as opposed to a large multinational), just putting in the company name may be enough.
• Sometimes there will be just one manual for an entire product family. If so, searching for the company name and the product type (like “Angelcorp CD Player”) will put you on the right track.
• There is also a “Full-Text Search” under the search box you can select which will look inside the documents in the manuals, allowing you to search for model numbers or information in a much deeper way.

Naturally, a collection of this size is rather hard to browse through – there are manuals coming in by the thousands every month, and we are working to get them into proper sub-collections and headings, as well as improving metadata. The work is never done with a project like this, but the joy and wonders never cease.

So enjoy a million manuals, and we’ll see you at two million!

A shout out and deep thank you to all the different communities and individuals worldwide who have assembled manuals mirrored from external sites like manualzzz, manualsbase and iFixit, as well as by uploading in some cases scanning and uploading thousands of manuals to the Internet Archive personally, for being the reason this collection has grown so large.

The recently released video game documentary High Score includes a sequence in the third episode about a game called GayBlade. GayBlade is one of the first commercially-sold LGTBQ-themed video games, a role-playing romp for Windows and Macintosh occasionally referred to as “Dungeons and Drag Queens”. Once thought to have been lost, the game’s software was recently discovered and preserved—and is now available in the Internet Archive!

Although LGTBQ people have been creating video games since the earliest days of the industry, there were very few games before the 21st century that explicitly had LGTBQ themes. Game creator Ryan Best hoped to change that with GayBlade, remarking, “This game gives lesbians and gays—and straight people—a chance to strike back at homophobia from behind our computer screen.”

The game is definitely political, racy and unafraid to make waves, as it definitely did in 1992 when it was released. Players are tasked with exploring a deep dungeon filled with homophobic enemies, trying to rescue the Empress Nelda and return her to Castle GayKeep. Best (and co-creator John Theurer) filled the game with humorous spells, items and antagonists while still keeping it all within the traditional role-playing genre. There are over 13 levels and 1,300 different rooms in this dungeon, reflecting the remarkable amount of work put into it by its creators—truly a unique work of art.

After being lost in a move from Honolulu to San Francisco, the game was thought to have disappeared forever. In High Score, creator Ryan Best laments that he was unable to find any of the game files, and was not very hopeful he would ever find them. But that’s not the end of the story—between the close of filming and the release of the documentary, Best discovered another copy of his game. Thanks to efforts by the LGTBQ Game Archive, Strong Museum of Play, and Internet Archive, it was preserved.

If you want to experience GayBlade for yourself, it’s available in our emulated games collection. You can play it directly in your browser if you’d like, or download the original source code. Additionally, an even earlier LGTBQ game called Caper In The Castro, a mystery adventure dating to 1989, is also emulated in the archive. So hit play and take a look at a little-known slice of LGTBQ history!