Linux Game Publishing: An Interview With Michael Simms
As we head into the weekend, it seems appropriate that we talk about something that a lot of us may do on the weekends - play games. That's an integral part of making Linux more fun. And so, today's post is about one of the people who makes these Linux games possible - Michael Simms, the CEO of Linux Game Publishing.
Michael, who has a background including 13 years of C programming, was a really fun interview and someone I hope to interview again sometime in the future when we have more time. As I understand it, Michael was busy coding in the U.K. while I was emailing questions from Panama, so the time difference made things difficult, but workable. On to the interview:
LG: As the Linux market is expanding throughout the world, it seems intuitive that the demand for games are growing and that this could be a market where developers for Linux could make some money. What are your thoughts on that?
Michael Simms: The market for Linux is definitely growing, but the desktop market is not growing as quickly as we had all expected. Almost every year we hear one or another Linux luminary declaring that 'this is the year of the Linux desktop'. but instead of the promised meteoric explosion of Linux on the desktop, we are seeing a slow and steady growth of numbers. Some of the big challenges have gone away now. The desktop for Linux is now almost as easy to use as the Windows desktop, and after a look at the new version of Enlightenment, I would say that it can look even better than even the best of the rest.
However from my perspective, the two big problems that stop mass adoption are Inertia and Fear. Inertia in that 'I have Windows, yeah it crashes a bit but not enough to make me angry enough to switch to Linux' and Fear in that 'But Linux is so hard to use'.
I see the problem of Fear going away as people see Linux and realise 'hey, it isn't all command line and techie stuff'. I also see Inertia going away, as more and more people are introduced to Linux as a first serious operating system, either by their school or their work who have adopted it as a better deal than Windows, or by peer pressure from their friends to 'give it a try'.
After that, I can see a big expansion of Linux on the desktop. Right now, the desktop isn't for those out to make a quick buck, it is for the early adopters. Those that want to generate brand loyalty and who have the foresight to know that as time goes on, this is going to be a market that really matters.
LG: Do you think games have a pivotal role in getting people to use Linux?
Michael Simms: Very much so. Things like a spreadsheet and graphics package mean that people can use their computer for working. Games mean that people can ENJOY their computer. If all you have is productivity apps, then Linux will be a fine OS for work, but who is going to really want it around in the home if all they can do on it is work.
LG: It seems it would make sense for some of these commercial Linux and Open Source entities to fund some game development - has there been any interest that you know of?
Michael Simms: None that we have heard of. We are completely self-funded and haven't been approached by any of the big players.
LG: Linux Game Publishing is looking for new games to publish; what do you expect of the proposals from developers? Should they have a working game with documentation, or can developers show you a beta and see if you think it is worth pursuing?
Michael Simms: We have had all kinds of proposals over the last few years. Sometimes people come up to us and say 'we have a great idea for a game, give us money'. We generally just ask them politely to go on their way.
If someone comes along with a genuine game, we will give them our honest opinion as to whether we see a market for it. Most games that we are approached with are too close to existing open source games for us to publish, but we have had one or two interesting proposals, one in the last few weeks that we are evaluating quite seriously.
LG: So games that are already Open Sourced are things you don't intend to compete with. What about trademark and licensing issues with existing ports of games from the operating system from Redmond?
Michael Simms: No, we have no real desire to compete with open source products. However most of the open source games are smaller and dont really fall into the arena of what we are aiming to publish.
When we sign a contract to port a game, we obtain full licensing rights for it, and that includes indemnification from copyright issues. As far as we are concerned, if the company we license a game from has a copyright or trademark issue, it is their problem. We pay them for the rights, and it is their responsibility to ensure the rights are theirs to sell in the first place.
LG: Do you look at how successful the title is on the other operating system to see if it's viable?
Michael Simms: To a degree, yes. We look at how successful a game was, but we also look at how good it is in our eyes. If we find a game we cant put down, then chances are, since we are fairly typical Linux users here, others will probably like it.
LG: What if someone has a good game concept but can't seem to code it properly - do you help them if the concept is good?
Michael Simms: We have tried a project along those lines and it didn't work out, so we are a bit reluctant to try that again.
LG: Was that a problem of vision?
Michael Simms: Well, we set up a group of people to work on a project, where they were guarenteed the game would be published when it was complete as long as
they did it well. We provided the group with a project manager and did a lot of work with them on it, but before long the group just fragmented and needed so much hand holding that it was not really time-economical. As soon as we stopped hand-holding every step of the way, the project collapsed.
A lot of it could have been with the vision. It was one persons idea and the rest of the team, some of them just werent as enthusiastic. Others were too enthusiastic and got disheartened when the whole group didn't spend every waking moment on the project, and others just wanted to take short cuts.
LG: Project management is tough. But one of the other challenging aspects of the Linux platform seems to be device drivers. Though things have become a lot better, what are your thoughts on the future needs of video card device drivers for Linux, such that Linux becomes more attractive to people clinging to Windows to play games?
Michael Simms: Well, one of the thing my developers tell me is a problem right now is the porting of the more recent additions to DirectX 9. The 3D technology of OpenGL is great, but it hasnt quite kept up with DirectX, leaving us with more work to do each time a 3D game is ported.
For most graphical things, however, SDL [Simple DirectMedia layer] is a great tool, and I strongly reccomend anyone who is looking at Linux development to look at it.
The other major stumbling block for Linux porting is networking. Microsoft has done a good job making DirectPlay very easy fro the Windows developers, and utterly useless on Linux. As a closed API that uses encrypted data streams, there is little or no chance that Windows games ported to Linux will ever be network compatible with their Windows counterparts. Some companies such as Epic and id have stood away from DirectPlay and their games are fully compatible between Windows and Linux.
LG: So Epic and id are at least seriously considering the Linux market.
Michael Simms: Yes, they are. Both of them realise that the Linux market is incredibly loyal, and they are reaping the rewards in terms of good PR, linux users hosting online servers for their games, even if not much in the bottom line yet.
LG: So, DirectPlay is a problem. Would it be possible to come up with a similar programming interface, along the lines of what wxWidgets has done for multiplatform code - which is *almost* a plug-in replacement for the Microsoft Foundation Classes? It seems that this might be a good way for game manufacturers to get something less tied to Windows, and give them some more control.
Michael Simms: Actually we are in the process of doing just that. I have a developer right now working on a wrapper around the LGPL'ed 'openplay' network layer, and this is something we hope will be able to just drop in to be fully compatable (at an API layer) with DirectPlay. Once complete this should make porting games with DirectPlay much easier.
LG: How will an evolution of the kernel be dealt with when it comes to the games?
Michael Simms: When we build games, we build using a completely compartmentalised environment, where the machine we are building on does not matter. We can actually build games for PPC computers on a standard PC, or a sparc, or an alpha. This closed environment has been very carefully selected over a number of years to be neutral to as many aspects of the kernel, the libraries, and everything, as possible. We do not see the kernel changes being an issue.
LG: So it is possible that you could publish games for more than one platform?
Michael Simms: In theory yes, but I, and all of my staff, are linux people at heart. We have no interest in porting games to the mac, or to anything else, and so we dont. Saying that, our latest release, Soul Ride, works on x86, ppc, alpha, and sparc Linux.
LG: There's the issue of how games licensed under Open Source and Free Software licenses can make money. It was only 20 years ago when games were designed, written and documented by one person or a very small team - but modern game development companies have sometimes 30 people working on the same game nowadays. It seems viable that games on Linux would have to be proprietarily licensed for the games to be able to afford such teams - unless someone commissioned the game for philanthropic reasons. What are your thoughts on game licensing in the future, in your experience?
Michael Simms: For years people have been asking me 'what about open source games'. And my reply is always the same. They work for the one person sized games, where someone in their garage sits there coding away for 3 months and then puts it on their website, or they work for games where the developers are copying an existing game, such as the freeciv project. For truly innovative large scale games, open source cannot do it. The reason is, the vision. When you are working on the kernel, it is there to do a specific job, and the famous situation of a developer 'with an itch to scratch' can be put to good use. We know what the kernel is there for, and so we all share the vision when we develop for it. Games on the other hand are very different. Games are based on imagination, and to get a good large scale open source game you would need a team of open source developers who all had the same vision of the game. When you are coming to a creative imaginitive work, that is just so unlikely to happen, and so far hasnt happened. When a project like freeciv comes along, the developers are working to copy the look and feel of a game they all enjoyed before. That is possible, they all share the same vision. When it comes to creating a new game from scratch, someone someone has the first vision, and everyone else is compromising their own vision to someone elses. In the end the interest fades and the project dies.
I would love to be proven wrong about this one day.
LG: This comes to a similar end, though - where the business interest replaces the developer interest. But a large project that demands support requires developers, and developers have to eat (contrary to what many people seem to think). Where the two coincide is where the developer may have to worry about paying the bills more than doing 'cool stuff in the garage'; do you think that some games that have been Open Sourced and where the developers have lost interest could be revitalized by a business? And if so, how - aside from closing the source - could this happen?
Michael Simms: It would be hard to do. Most of the time, the way to make money with open source is to have people pay for support, but people arent going to pay for support on a game. I cant see a way to take an open sourced game and making money from it.
LG: Do you think it would be viable for a Massive Multiplayer Online Roleplaying Game to be written and licensed through Open Source licensing, allowing for companies to run a server with the game and charge a subscription fee - as Everquest and other games do? Or would there be a risk of cheating?
Michael Simms: I dont thing cheating would be the problem. I have a great belief in the open source leads to great security and I think that an open source MMO Game would probably be the most cheat-free game out there, once they have gotten past the teething troubles they would have. I think the problem would be persuading companies to adopt it. A company will want to see they have a clear advantage over their competitors before they invested money into providing the content that a massively multiplayer game needs, and with an open source engine and client, the content IS their only advantage. It is too much of a level playing field for companies to risk in todays money-driven world.
LG: So you don't think cooperative code, like Apache, would work in such an instance because of the 'clear advantage' issue?
Michael Simms: well, companies use apache to drive awareness of their products through their websites, or to provide access to their products through their online store. Apache rarely drives the product itself, so I think that that is a bit like apples and oranges.
LG: As a game publisher, how do you see the issues with Blizzard Software and Battle.Net in the context of Linux? Was that a special case, or is that a good indicator of how the gaming industry does business - and if it is, why does it have to do business that way?
Michael Simms: I think Blizzard made a mistake. There has never been any kind of open source threat to any of their current titles, there have just been fans emulating their older games, or in the case of battle.net trying to play
their purchased games in a better way.
All Blizzard has managed to do is alienate some of its most loyal fans and supporters, aka their best customers.
Blizzard is definitely no friend to Linux or the open source community. Sure they make good games, but thats about it. There is a Linux version of the hugely popular World of Warcraft, and Blizzard canned it, without warming or explaination, even though it was functionally complete and ready to go, and after a discussion of a
support agreement with LGP. It would have risked nothing for them to make the game available, and they chose not to.
LG: Have you considered a distribution of Linux with all your games?
Michael Simms: No, We dont have the resources to keep up with all the changes that go into a Linux distro. We have had conversations with Linux distributions about bundling our games with their distributions, however.
LG: Cool, please keep us in the loop on that. Do you have any words of advice for people wishing to develop their first game? What resources would you point them to?
Michael Simms: SDL is the most important resource for a game developer. Without a doubt. It provides a strong API for just about anything you want to do.
I think though that the best advice would be - dont re-write the wheel. We have 300 versions of Ttetris, 200 versions of Breakout and Mahjongg, think of something new.
Look at some of the most popular open source games out there, and you will see that all the best are the unique ones, frozen bubble for example, its different, and its popular! Generally, as a company that needs to at least break even on a title, we are very unlikely to look at publishing something that is a clone of an existing game. Partly for legal reasons, and partly because, well, if the other game is open source, what does your game offer players that the open source one doesn't? There needs to be a 'must have'.
LG: Do you think that people need to take any special classes? As I understand it, there are degrees related to Game Development available -are these worth pursuing?
Michael Simms: There are arguments for and against this. I can see some cases where it would be useful as a game developer to have a good grounding in that before they start, but on the other hand, having a good imagination and the ability to code is all you really need. There are issues that arise more in games than in other applications, such as using double buffering to eliminate screen flicker, timing to ensure that games run at the same speed regardless of what computer it runs on, how to maximise frames per second, and what exactly is a 'vertex shader' anyway. However if you are a good programmer, you can find out these things anyway.
LG: When you're not busy doing CEO work, what game do you play the most?
Michael Simms: When Im not busy doing CEO work... well, that would be back in 2002 I think when I seem to remember I did play one game of Majesty...
No seriously, I do enjoy a good game of Majesty, and I am very good at Descent 3 too. Our new title, about to go gold soon, Ballistics, is one that I like to sit down and play when I get the chance. I don't really enjoy First Person Shooters, but I love strategy games. However, work keeps me busy usually for 12 or more hours a day usually 7 days a week. More likely in my free time Im going to collapse infront of the TV with the news or an episode of something sci-fi based.
LG: Thanks for your time, Michael. Maybe we'll be able to talk some more about your new release when it comes out. Everyone who's interested in playing games should probably take a look at the Linux Game Publishing website - and if you have a good idea for a game, maybe you should talk to them as well.