Loki Retrospective

The Major Linux Porting House Files for Bankruptcy
13 August 2001

By: Matt Matthews and Dustin Reyes

Two and a half years after the initial announcement of the Linux version of Civilization: Call to Power, Loki Entertainment Software, the respected Linux game porting house, has filed for bankruptcy. After a long series of departures of both development and technical support, most of the core employees have left for careers elsewhere. Certainly now is an appropriate time to look back on the events that transpired in the life of Loki: where they came from, what they achieved, and some thoughts on where they stand now.

A Brief History of Loki

On 26 January 1999 various news sites ran a story about Loki Entertainment Software porting Activision's Civilization: Call to Power to Linux. According to Scott Draeker, the president and founder of Loki, the idea to port games to Linux had originally come up during the summer of 1998, no doubt stirred by their understanding that there were 12 to 15 million Linux users worldwide at that time. Certainly, there seemed to be an opportunity providing games to a fast-growing market which had been mostly ignored, except for some efforts by a well-known maker of first person shooters.

For the first time, Linux users were treated to the experiences known all too well in the land of Windows games: a selection process for beta testers, a smattering of screenshots to keep the public in touch, a delay in the planned schedule, and eventually a final release date. Then, for the first time, on 19 June 1999, a reader of LinuxGames sent word that he'd purchased a retail version of a game for Linux at his local Electronics Boutique.

Shortly before the release of Civ:CTP, Loki announced that they would be producing three more ports: Bungie's Myth 2: Soulblighter, PopTop's Railroad Tycoon II and expansion pack, and Delta Tao's game Eric's Ultimate Solitaire. Soon these games, too, had their beta tests and began shipping to stores. Myth II had a demo before release, allowing Linux gamers to try it out before making a purchase. With Myth II available in August, RT2 shipping in October, and EUS arriving in November, Linux gamers had a lot of reasons to be pleased with how things were going for their underdog operating system.

Further, Loki had invited aspiring Linux game programmers to join them and show off their skills at Loki Hack 1999. The participants were given 48 hours in a secure setting to hack up patches to add extra features to Civilization: Call to Power. This not only got the community involved in Loki and their quest to bring games to Linux, but also helped scope out a few future Loki employees, including Ryan 'icculus' Gordon, John 'Reeves' Hall, Andrew 'prometheus' Henderson, and Daniel 'hotdog' Vogel. Patches were coded, the results were judged, and subsequently the prizes were rewarded with first place going to Christopher Yeoh's modification adding several new units and an enhanced spy unit.

The end of 1999 also saw Loki publishing Quake III Arena, id Software's latest first-person shooter. This move alone seemed to legitimize the world of gaming on Linux and support seemed high among Linux users. Shortly before Christmas, Linux gamers had their copies of the Linux version of Q3A. Even so, December 1999 brought even more titles from Loki itself. 3D0's Heroes of Might and Magic III and Raven's Heretic II also made it onto store shelves in time for the holiday season. In the space of a year, Loki had put out seven commercial games and the future looked quite bright indeed as the first days of 2000 dawned.

At the Linux World Expo and Conference 2000 in New York, Scott Draeker was understandably positive about the prospects for the new year. He predicted that Loki would publish 16 games by the end of 2000 and went so far as to indicate that a sports game, probably a racing title, was likely. And indeed several new games appeared on the horizon and in beta tests: Heavy Gear II, SimCity 3000 Unlimited, Soldier of Fortune, Descent 3, Sid Meier's Alpha Centauri, and Mindrover. With a lineup like that, Linux gaming was looking to establish itself as a real player in the world of computer games. Eventually, all six of these games made it to store shelves, although the winter would pass and spring arrive before they reached gamers' hands.

In August 2000, Loki picked up another contract, this one to maintain the Linux port of Unreal Tournament. With Unreal Tournament in house, the announcement of Deus Ex and Rune, both running the UT engine, came as no huge surprise. Added to the list of Coming Soon titles were Kohan: Immortal Sovereigns, a real-time strategy game being developed simultaneously for Linux and Windows, and the highly-anticipated Tribes 2. This last title was especially dear to Linux gamers since they had been denied the opportunity to take part in the Tribes frenzy back in 1999. Along with the announcement of Rune late in 2000 came the simultaneous announcement that FAKK2 was also being ported to Linux by Loki.

Shortly after the beginning of 2001, the public became aware that programmers and support staff were leaving Loki to take employment elsewhere. Starting with the departure of Daniel Vogel for Epic Games, the core Loki crew left over the period of several months: Joe "tsaotsao" Valenzuela, Michael "Briareos" Vance, Andrew "prometheus" Henderson, Rafael "raistlin" Barrero, Stephane "MEGASTeP" Peter, Bernd Kreimeier, Ryan "icculus" Gordon, Andy "yoda" Mecham, Sam "hercules" Lantinga, and Mike "heimdall" Phillips. Loki maintained a mostly stony silence during this time.

Surprisingly, the departures preceeded the release of several games. Seven months had passed since the release of SimCity 3000, yet in April 2001 Loki finally shipped Sid Meier's Alpha Centauri and the long-awaited Tribes 2. In the following rush, a major Linux gaming reseller, Tux Games, had trouble keeping up with the orders that poured in from eager Linux gamers waiting to join the online fray; several gamers waited weeks for their back-ordered copy to arrive. Then in May, Mindrover also shipped and then Rune followed in June. Finally, at the very beginning of August 2001, Loki announced that Kohan: Immortal Sovereigns and Rune: Halls of Valhalla (a stand-alone expansion pack) would ship on August 15th. However, as of this writing, that shipping date is still two days in the future, and with the public news that Loki has filed for bankruptcy, it is not clear whether these last two titles will ship. Equally unclear are the fates of FAKK2 and Deus Ex, the former of which was in beta testing as of this writing.

After two and a half years, Loki has ported and published over a dozen games. Now, having filed for bankruptcy, Loki is protected from their creditors and able to continue doing business. Yet, under such circumstances, how can they be expected to produce any future ports?

What Loki Has Done Correctly

To outsiders Loki seemed to have been a successful company. From a technical standpoint, the job they took on was tremendous: the operating system they were porting to didn't have a standard desktop or even standardized libraries to depend on. In the face of that, the number of tools they worked on and built and then released to the public is nothing short of amazing.

The first aspect of a Loki product, other than the box and CD, that a Linux gamer was ever likely to see was the Setup tool. Just as much Windows software uses similar install software, Loki made Setup a tool that could be used to install software easily on a Linux system. Other companies have picked up on this nifty tool, including other game porting houses, to deliver their products from CD to a customer's hard drive. And although they will probably never be used as much as expected for Loki's games, the Uninstall, Patch, and Update tools should provide later commercial companies with a solid base on which to build their software maintenance and patching functionality. All of these, of course, are open source. Even their bug-tracking system is based on open source tools: Fenris is a fork of the Bugzilla bug-tracking system.

Even with these tools, there is still the job of porting a game, and Loki has put out some quality libraries to help in that area. Their SMJPEG and SMPEG libraries provide a way to deliver video and audio on Linux. OpenAL, intended to be a cross-platform equivalent of OpenGL for audio, was developed in great part by programmers at Loki. And SDL, the Simple DirectMedia Layer, was used by every Loki port in some form or another and there can be no doubt that without this commercial-grade testing of that API it wouldn't have advanced as far or as quickly as it has. Pushing 3D games on Linux, like Heavy Gear II and Soldier of Fortune, has also pushed the quality of OpenGL support on Linux. Many of the bugs in Mesa, Utah-GLX, and the DRI have been found as a result of attempting to play Loki's ports of 3D games. And now, as Linux is growing up, these libraries (like OpenAL, SDL, and hardware accelerated OpenGL) are starting to get included in the standard installs of many distributions. This paves the way for future game programmers to simply develop to the open standard and be confident that an end user's distribution will provide the required libraries.

In the mainstream of Pentiums and Athlons, it's easy to forget our friends in the PowerPC world. Luckily for the PPC folks, however, Loki made sure that they were included and made available binaries for all four of their initial games. You can even play EUS on Sparc and Alpha hardware. In one remarkable case, Civilization: Call to Power was available for LinuxPPC before it was available for MacOS; this demonstrates the kind of power that a cross-hardware operating system can offer. This kind of broad support for various flavors of Linux is yet another admirable feature of Loki's approach to porting.

It would be difficult to fault Loki for their interaction with the community. Other than LokiHack '99, for which we are aware of no comparison in the Windows world, Loki has maintained an active beta testing program, a USENET server with a group for each game and project, and two IRC channels. Through these communication channels, any Linux gamer could chat with the programmers or support staff, to discuss gaming and bugs and any problems they had getting their system to play Loki games. Naturally, Loki also provided email and phone support, for those less connected to the 'net. The Fenris bug-tracking system also found itself heavily used to keep gamers and programmers apprised of the each little glitch. Then there are all the end-user content creation tools that Loki went out of their way to provide for their fans: Fear and Loathing for Myth II, Heroes Map Editor for HOMM3, GTKRadiant for Quake III Arena, and the BAT for SimCity 3000. These tools brought absolutely no revenue to Loki (and in fact probably cost money in developer time and resources) yet they were provided with no strings attached for the users that bought Loki's games. As for the media, they provided those of us here at LinuxGames with the treatment afforded to other gaming sites by Windows game companies: automatic inclusion in beta tests, review copies of games, open communication with the programmers, and access to information before the public at large was notified.

While executing well on each of these counts and securing some top-notch titles to port to Linux, it seems difficult to see how Loki could have failed. Yet, they are now seeking respite from their creditors, and we're left asking: Why?

Potential Problems for Loki

While it's difficult to say precisely what brought Loki to this state, we can at least point to a few things that probably contributed.

The most compelling answer to the question "Why did Loki fail?" is probably that they overestimated the size of their audience. Despite recent claims that Linux has surpassed Apple's Mac hardware in its share of computer desktops, there are probably still not enough willing to stick to Linux full time to keep Loki's sales high. Factor in the price premium sometimes attached to Loki's games, and it seems like a losing proposition; gamers with x86 hardware can go right out and get the Windows version of a game more readily and probably at a cheaper price by walking into their local software store. Add in the reported poor sales of the Linux version of Quake III Arena and it becomes more believable that Loki was losing money to poor sales. That most of the games they published were released several months after the Windows version couldn't have helped either.

Which brings up another distressing aspects of Loki's whole enterprise: lack of penetration into brick-and-mortar stores. The public will blithely remain ignorant of your existence if you don't procure and maintain a section of the shelves of your local software shops. While Loki's games were spotted in several retail locations, including Babbage's, Electronics Boutique, Software Etc., and Fry's, the selection and placement have typically been suboptimal. (From my own experience, I've only seen Loki's games in Electronics Boutique, where they got in one shipment around Christmas 1999 and never restocked, and Babbage's, where a large selection of games showed up just around Christmas 2000.) We were told at one point that there were prospects of getting Loki's games into CompUSA and Best Buy, but to date that has not happened. Perhaps it was a case of failing to break the chicken-and-egg problem: it is difficult to get into stores unless you can show high sales numbers, but in order to have high sales numbers you have to have had something on store shelves that sold well.

Along with that comes the almost complete lack of advertising. While Loki never advertised in any of the mainstream game publications, they did place a single advertisement in Maximum Linux. It is fine to use interviews and online publicity to keep in the public eye, but it is probably not wise to rely on those avenues. One wonders what effect a series of well-placed adverts in some of the Linux magazines might have had.

At its peak, Loki had a dozen people working in porting and support. Given the relatively small size of the community, this team seems huge. It wouldn't be unreasonable to think that Loki bet on a fast-growing Linux desktop market and lost when the users didn't appear with wallets in hand. Then, caught with a heavy staff, lots of overhead, and little cash flow, things fell apart.

Where We Go From Here

Many people will be confused and saddened and angry over this change of fortunes for one so bright and promising as Loki. The question thus arises: What now?

With protection from their creditors, Loki may be able to turn its business fortunes around. For example, it could receive funding to port and publish new games, ration its expenses, and become a normal and profitable business. There is also the possibility that Loki could be bought by a larger company, and continue to exist under their umbrella. Loki's creditors could assume control of the company, in which case a large group, possibly including some who have no stake in publishing Linux games, would determine if and how Loki should continue business. A final option is that Loki could simply shut down, paying off their creditors as best they can.

When asked to comment on the situation, Loki Scott Draeker made this reply:

"We filed a Chapter 11 reorganization. This will allows us to deal with our creditors fairly and equitably and at the same time continue to operate the company. We are still shipping products and porting new games and expect to be doing so for a long, long time."