Life after death: meet the people ensuring that yesterday’s systems will never be forgotten

Jaguar

For James ‘Shamus’ Hammons, Doom and a limbless, white-gloved cartoon mascot were all it took to ignite a longstanding obsession with Atari’s Jaguar. Back in the early ’90s, he felt he couldn’t abandon his Atari ST for IBM PC-style computing simply to play id Software’s genre-defining FPS, but then the news came that Doom would hit his favourite company’s new 64bit console. The real clincher came later, however. “I saw a preview of Rayman,” he recalls, “which at the time was going to be a Jaguar exclusive, and it looked amazing.”

Today, Hammons – the lead developer of the Virtual Jaguar emulator – is just one of a growing scene of enthusiasts dedicated to sustaining consoles the world would sooner forget, earning them some respect for the niche they tried and failed to carve out. These emulator developers, amateur historians and digital archaeologists want you to remember Jaguar’s brief run at Nintendo and Sega’s dominance, and Virtual Boy’s daring attempt at 3D gaming almost 20 years too soon. And they’re desperate to preserve something of ambitious systems such as Nuon and Pioneer’s LaserActive, which were never widely known in the first place.

Hammons’ early enthusiasm for Atari’s Jaguar hardly matched the gaming world at large. Released in North America on November 15, 1993, and in Japan and Europe the following year, it represented Atari’s final stab at relevancy in a market rejuvenated and then snatched away from it in the latter half of the ’80s by Nintendo and Sega.

But Atari’s 64bit console was notoriously difficult to develop for, and neither Doom nor Rayman nor Jaguar-exclusive Tempest 2000 could prevent it failing spectacularly at retail against the older SNES. Humiliated, Atari pulled out of the videogame hardware market in 1996.

Doom brought the FPS to consoles long before GoldenEye on Atari’s Jaguar.

Hammons let go of his Jaguar, too, but the memories never faded and several years later he went searching for an emulator that could play the Jaguar version of Rayman. None could, although David Raingeard’s Potato Emulator was apparently making rapid progress. “And it seemed that he was doing his best to make it run most games,” Hammons says.

Raingeard abruptly stopped working on the emulator without explanation in mid-2003. He released it under a general public licence, though, whereupon a programming group called SDLEmu used it as the basis for its open source Virtual Jaguar project.

“Rayman still didn’t work properly,” Hammons says, “so I decided to take a look at the source code and see if I could somehow fix the problems with it.”

The code was a mess. “It relied on a non-portable, closed-source 68000 core written in x86 assembly language,” Hammons explains. It was also littered with game-specific hacks that made it good for only a handful of titles. Little wonder Raingeard abandoned the project: he had coded himself into a corner. Adding support for Rayman would be no easy feat, and it proved just the tip of the iceberg.

“I can’t remember what I tackled first,” Hammons continues, “but I was determined to get the codebase into a more sane and portable state.” He presented his changes to Niels Wagenaar of SDLEmu, and within a few years he became the primary, and then sole, developer on the Virtual Jaguar project. He and a handful of contributors continue plugging away at the emulator to this day, making progress mostly by writing test programs or finding software that doesn’t work and painstakingly analysing how it exposes weaknesses in the emulation. Hammons wrote a pipelined version of the emulator’s digital signal processing, for instance, because he discovered that Wolfenstein 3D needed that feature for its audio.

Every performance or compatibility gain is embraced by a small group of Jaguar fans on the AtariAge forums. Hammons says the Jaguar scene has undergone a mini-revival in recent years, spurred in part by his emulator, but also by new game development. Downfall, a simple game in which your goal is to stay within the bounds of the screen while falling, was released in 2011 by leading Jaguar developer Reboot, a six-strong team of coders. Downfall now comes bundled with Virtual Jaguar. Éric Chahi’s Another World was also ported to the console last year, courtesy of Sébastien Briais from programming group The Removers‚ who got the idea and Chahi’s approval after seeing a Linux port of the game’s engine in 2007.

Atari’s Jaguar fell over in a few areas, not least with its poorly designed joypad.

Hammons isn’t alone in his quest to keep the outcasts of videogame history alive. Some lead the battle for their system almost solo, trailing a handful of enthusiasts and homebrew coders in their wake. Others find strength in numbers.

Christian Radke founded Planet Virtual Boy in 1999 with the intent to preserve, document and develop new games for Nintendo’s unwanted red-screened stepchild. A surprisingly robust community soon formed around the site, where members share their discoveries and work together to learn more about the system.

Intended as a pioneer for virtual reality gaming, the distinctive tabletop console’s red-on-black stereoscopic 3D visuals won it few fans upon its release in Japan and the US in mid-1995‚ leading to its discontinuation only months later. Nintendo probably would rather you forgot it ever existed; tellingly enough, even when the notoriously secretive company decided to provide sales figures to NeoGAF user Aquamarine for consoles stretching back as far as SNES, Virtual Boy was omitted from its figures.

But Radke, a professional web developer, is determined to prove the device’s worth and to record its high and low points for all the world to see. He first got wind of Virtual Boy in a one-page article in Germany’s official Nintendo magazine at the impressionable age of 13, although it wasn’t until 1999 that he got his hands on one. His fascination has since morphed from giddy anticipation to studied curiosity.

“One thing I enjoy almost as much as a good game or building websites is researching and documenting niche videogame systems and their history,” he says. “Searching through old magazines and websites, talking to former developers, looking at game binaries for cut content, unearthing previously unknown details, finding images of unreleased games – it’s like digital archaeology.”

Accurate sales figures for Virtual Boy are hard to come by, but the console lasted less than a year before being discontinued.

Radke and his Planet VB cohorts have developed entire Virtual Boy games from scratch. Radke is personally responsible for a Sokoban clone called Blox, along with a sequel, Blox 2, while others have released rhythm games, platformers, an unfinished and wholly unofficial Mario Kart sequel and more. They’ve tracked down marketing materials, including game screenshots and videos, along with official documentation, and dumped and distributed copies of unreleased games – namely Bound High and Faceball – on cartridge as well as in ROM form. The Planet VB community builds its own development tools to simplify the process of developing homebrew games and apps, and the site holds homebrew coding competitions roughly every two years, the most recent of which yielded a Game Boy emulator running on Virtual Boy.

There are few limits to these enthusiasts’ dedication beyond the time they have spare to commit. For example, one member, Guy Perfect, spent weeks putting together a painstakingly detailed tech scroll that breaks down the inner workings of the system, simply to fill out the smaller details that had been overlooked in other documentation, both official and unofficial.

It’s this obsession with preserving Nintendo’s console in its totality that unites the community, and Radke points out a quirk in the system’s appeal: “Compared to most other consoles, there’s a manageable amount of Virtual Boy-related things. Getting it all together is very possible, be it in the form of a complete collection of games [22 were released], hardware and merchandise, or complete informational coverage, like on Planet VB.”

This puts the system in stark contrast against the swell of thousands of titles that exist on more popular systems such as Sony’s PlayStation or Nintendo’s NES. Obscurity is both a blessing and a curse when it comes to preservation.

“The winners always write the history,” game development student and Pioneer LaserActive historian Max Krieger says, “but that creates a clouded picture. And I think people, and game developers in particular, deserve to know about the failures.” Krieger speaks with a wisdom beyond his years. A student enrolled in DePaul University’s games programme in Chicago, he discovered the LaserActive console while browsing around Wikipedia a few years ago.

LaserActive looks like an oversized DVD player, and its add-on modules are equally monstrous.

Pioneer’s LaserActive is a strange footnote in videogame history, released in North America and Japan towards the end of 1993 with a price as monstrous as its crate-like shell. Intrepid buyers faced an outlay of $970 for the base unit, which was a bare-bones LaserDisc player, plus around $600 each for the two modules that played Sega Genesis/CD and TurboGrafx (PC Engine) games as well as LaserActive-specific LaserDisc games such as the Taito-developed Hyperion. There were two other modules, too, for computer connection and karaoke, in addition to special 3D goggles you needed for rail shooter Vajra 2 and a few non-game multimedia titles.

Krieger has dedicated much of his free time over the past two-and-a-half years to unearthing everything he can find about the rare console. “I think the LaserActive caught my eye because it was so outlandish,” he says. “It was this massive machine that took like an entire entertainment centre to fit. It made a lot of noise [and] it cost a lot of money, so I was like, ‘How did this thing even get released?’”

What made it even more intriguing for him was that Pioneer’s LaserActive had no emulator. Popular consoles from the same era had functional and fast emulators before the turn of the millennium, and these days even the LaserActive contemporaries that failed have some level of emulation – Philips’ CD-i has CD-i Emulator, Atari’s Jaguar has Virtual Jaguar, and 3DO has FreeDO and 4DO, while Fujitsu’s FM Towns Marty has UNZ and Xe. Yet Pioneer’s machine didn’t even have a work-in-progress project. There was scant information about it or its games on the Internet, so Krieger started digging, and he hasn’t stopped since, creating the LaserActive Preservation Society along the way to record and share his findings.

Krieger spent $450 to buy and repair his second-hand LaserActive with five games and a Sega module, and he uses this setup to record footage that he posts to YouTube so that at least something of the console has been documented for the generations to come. The LaserActive Preservation Project’s channel has 19 unabridged gameplay videos at the time of writing, with more promised to come soon.

There’s now a hope for a more tangible kind of preservation, too, because a few clever souls at the SpritesMind Genesis homebrew development forums have figured out a way to dump the data contents of the LaserDiscs via a Sega-controller-to-USB adapter cord. This could pave the way for emulation through retooled Sega CD and TurboGrafx emulators, although that won’t happen right away.

“The next part that’s a real challenge is that LaserDisc video is not digital, it’s analogue,” Krieger explains. “It’s a lot like a vinyl record; it just reads the data as it finds it. It’s all analogue encoded, and that’s why you can’t really just put it 1:1 in a digital format. You [still] have to choose a format to store it in and, no matter what, you’re going to lose a little bit in the transmission.”

In the meantime, the LaserActive Preservation Society’s focus is on getting video footage of games as well as photos of documentation, packaging and marketing materials, and whatever else Krieger and a collector named Tom (whose surname we’re asked to keep private) can string together. Information comes in part through donations and inside leads, but primarily, Krieger says, “it’s just Internet detective work,” by which he means rifling through Google Books, eBay Japan and archived press releases, plus tracking down former LaserActive developers.

He considers this work crucial to developing a complete picture of videogame history. Without studying and learning from failures like Pioneer’s, he argues, we might repeat them. “I think it’s really important to remember why it didn’t take off,” Krieger says.

He also thinks it’s vital that students of game design consider what obscure games did right that more famous titles did not – his contemporaries might have a background in classic games, but few know about the really obscure stuff. “So they’re all working on the same page,” Krieger says, “but I just think that if they knew about some of this stuff it would totally change their perspective.”

Nuon, Jaguar and Vectrex mega-fan Kevin Manne can vouch for the redeeming qualities of failed consoles. “Getting a Jaguar is like finding the good in what everybody else thinks is terrible,” he says. “There’s always something fun and worth playing on any of these systems.”

You just have to find it. Which shouldn’t be hard at all when it comes to Nuon games, given that only eight were officially released (and over a dozen more were cancelled). The Nuon technology amounted to a graphics chip that added videogame and enhanced multimedia support to DVD players. It was included in a handful of machines from a variety of manufacturers between November 2000 and July 2002, following several years of research and development by much of the former Atari Jaguar hardware team.

Not a console but silicon that could be added to DVD players, Nuon was picked up by few hardware makers.

“It really was the PlayStation 2 idea before PlayStation 2 came around,” Manne says. “It was a DVD player and videogame console all in one – they were just doing it the other way around. They wanted to have the Trojan Horse effect and get this videogame processor into existing DVD players to replace the DVD encoding chip in there for a minimal upgrade cost to the manufacturers.”

It was a dismal failure. Manufacturers held back for more games while developers waited for a larger installed base. And not even Jeff Minter’s Tempest 3000, a Nuon-exclusive sequel to Tempest 2000 (which had enjoyed a PC release after the death of Jaguar), could sway a critical mass to adopt Nuon over PlayStation 2. Manne was there when it happened, watching his little fan community, Nuon-Dome, gradually shrink to nothing in the years after Nuon creator VM Labs filed for bankruptcy in late 2001.

An emulator called Nuance promised to revitalise interest, and indeed drove considerable spikes in activity at each release, but its sole programmer, Michael ‘Riff’ Perry, died of a brain aneurysm in 2007, aged just 32. “He was just a really smart guy who basically reverse-engineered this whole system and got it up and running on early 2000s PC hardware,” Manne explains. “With every iteration it got better and faster, and you could run games to a certain extent on it, but it was pretty slow at the state it was in when he passed away.”

Nobody else has stepped up to take on the task of emulating a system with half-a-dozen games and a few bits of homebrew, and Manne doubts that anyone will. The Nuon platform may die with its hardware, recorded in perpetuity by Manne’s web shrine, which he now updates only once a year, such is the scarcity of Nuon-related news and developments. Nonetheless, the page contains just about every last fact regarding Nuon’s short life.

The hopes of neglected systems such as Nuon rising from the ashes may rest on the work of arcade emulator MAME’s sister project, MESS (Multi-Emulator Super System). Having being in constant development for more than a decade, it has drivers in various states of completion for over 700 computers and game consoles, including many from small regional markets.  “There were always computers and consoles that were sold on local markets only,” MESS coordinator Miodrag ‘Micko’ Milanovic says, “and most people were never aware of their existence. Our goal is to provide as much info about them [as possible] and try to emulate them to run available software.”

For Milanovic, the initial draw was recording the existence of machines local to his native Serbia, such as Pecom 64 and Lola 8A. “For those, I did work from scratch and had [the] real machines beside me to help me figure out how it all works,” he explains.

The Lola 8A (left) and Pecom 64 (right) computers were produced and sold only in Serbia.

For some MESS contributors, though, preserving obscure systems is just a bonus. The real appeal for Wilbert ‘Judge’ Pol is simply figuring out how things work, a feat that he’s accomplished on several well-known systems as well as Game.com, Exidy Sorcerer, Bandai Super Vision 8000 and “probably a lot more”.

What unites the MESS team is a fastidious fixation on preserving in a single repository every computer and videogame system ever invented, and in making it possible for anyone to enjoy any machine, no matter how obscure. It’s an extraordinarily ambitious goal, but these are many of the same people who have spent the past 17 years emulating nearly every arcade game ever in MAME, and their focus is shifting. “During the past few years, we actually hit the wall in MAME,” Milanovic says. Little remains unsupported beside the newer, more powerful, and hence technically challenging, arcade systems released in the past decade or so.

Sega’s Lindbergh arcade system counts among the few examples. Its relative newness – it debuted in 2006 and runs games such as Virtua Fighter 5 and After Burner Climax – is part of the problem, but more still comes down to puzzling out the encryption module that features on its motherboard, which is going to take some time. Other systems go unemulated for all sorts of reasons. The most basic requirements for getting started are some example ROM dumps and knowledge of the system’s processor, which is relatively easy for more popular machines. Official documentation is seldom available for obscure systems, though, so the next stage involves trial-and-error tests during which the developers try to recognise what the system is attempting to do. This can happen quickly, but often it’s a gradual process.

Sega’s Saturn, its failed PlayStation and N64 challenger, has had plenty of interested suitors and multiple emulators under development for several years (the current best being SSF, followed by Yabause). But fast and accurate emulation of the ’90s console remains elusive. “Saturn is very annoying to make games on and even more annoying to emulate,” MESS contributor Angelo ‘Kale’ Salese says. “It has two standard CPUs plus another for audio, and then also two video chips, another few subsystems and a microcontroller that controls the whole system.” It’s incredibly complex, and hackers are still unsure how some components work.

Saturn’s complex architecture saw it struggle against PlayStation.

Saturn isn’t the only system for which emulation has proven an uphill battle. “For the Tiger Game.com, we had a general description of the main chip in the machine and how its peripherals were used,” Pol explains, “but nothing on the instructions of the CPU core besides a screenshot from a disassembler listing three or four instructions. I spent weeks figuring out the separate instructions by looking at the binary dumps of a few games.”

Apple and Bandai’s ill-fated Pippin console still remains unemulated despite the fact that it ran a stripped-down version of Macintosh System 7.5.2 on the same processor as low-end Macs of the time. The MESS team has only done preliminary work on a driver for the Old World Power Macintosh ROM that it uses, and nobody else has risen to the challenge of emulating Apple’s forgotten console, which saw fewer than 80 titles released, including Bungie’s Super Marathon and numerous small applications. You can, however, run some Pippin games in Mac emulator Sheepshaver.

Progress in the world of emulation, whether it’s of obscure systems or popular ones, is really just a matter of developer interest. If something gets too hard, or there’s not much known about the underlying hardware, developers are liable to move on to lower-hanging fruit. MESS is no exception. “What it usually comes down to,” Pol admits, “is [that] I’ll pick up what I feel like doing. If I start having goals and planning what should be done when, then it’d start to feel like work. It is a hobby, and it should remain that way.”

As for what drives his choice to work primarily on obscure systems, Pol is blunt. “It’s not like the more obscure computers and consoles had the greatest games or apps. There’s a good reason why those systems are not so well known. [But] the old media are deteriorating, and getting those games and applications preserved in some digital way is important to not lose those parts of history.”