By Qusay H. Mahmoud, June 2005
J2ME Luminary - Eric Bilange of MFORMA talks games in this interview with Qusay Mahmoud. Gain insight into the world of J2ME games, the technology and business. Eric is presenting "Developing Cross-Carrier Multiplatform Mobile multi-player Games" at JavaOne 2005, your opportunity to meet him in person.
Dr. Eric Bilange is the co-founder of MFORMA, a global publisher and distributor of mobile entertainment products. Earlier, Eric was CTO and co-founder of Indiqu, a leading supplier of wireless games and entertainment to 18 networks. Previously, he was director of product management Internet solutions for Alcatel, one of the world's leading manufacturers of telecommunications infrastructure technologies, based in Paris. In this capacity, he defined products and managed a mediation platform for ISPs, telcos, and portals to facilitate the access of wireless devices, PDAs, and set-top boxes. Before his stint at Alcatel, he was Netscape's director of professional services for Southern Europe, where he was responsible for managing the deployment of intranet, extranet, and e-commerce Internet solutions for major European companies, including France Telecom, Telefonica, Telecom Italia, Europe Online, Renault, Health Online, Unilever, and Barclays. Dr. Bilange graduated with honors from the Paris VI University and earned his doctorate, summa cum laude, in Man-Machine Dialogue from Rennes University, both in France.
Eric will present Developing Cross-Carrier Multiplatform Mobile multi-player Games at JavaOne 2005. In this interview, he gives us his insights into the future of wireless gaming, the Java 2 Platform, Micro Edition (J2ME), and more.
Qusay Mahmoud: Tell me a bit about MFORMA.
Eric Bilange: MFORMA is a world-leading publisher and distributor of mobile entertainment content. We deliver mobile entertainment applications to more than 100 of the world's leading wireless operators in 40 countries.
Our strategy is to produce the best quality mobile entertainment products, and to provide wireless carriers everything they need to succeed in delivering mobile entertainment to their subscribers. That means we not only furnish them with content – mobile games and other entertainment applications – we also provide the technology they need to manage and integrate the applications in their networks. We also provide them the leading brands and brand partnerships that make for high-quality, compelling entertainment products with wide and instant appeal. Our strategy is to extend this full-service, total-solution capability throughout the world. We already have operations in the U.S., U.K., Korea, and China, and plan to have a distribution center and a strong in-country presence in every key market.
QM: What is the most challenging part of your job?
EB: Technology! I mean the diversity of technologies that we have to understand and master to make our products easy to use and our games fun to play. If I can draw a comparison with the Internet: Back in 1994 there were two browsers, several simple yet powerful protocols – and this is pretty much still the case today. In wireless, you have a good dozen handset manufacturers each providing different OSes and you have many wireless operators using different network technologies and billing systems, providing different interpretations of standards and runtimes. From SMPP to 3D, there is a wide range of technologies to understand, learn, and follow.
QM: What products and services does MFORMA offer?
EB: We have one of the world's largest content catalogs, including more than a hundred applications, but more importantly it is one of the world's most diverse and high-quality catalogs. We have compelling products for virtually every mobile audience: games, sports, music, mobile 411, shopping, news, travel, astrology – with more coming.
QM: What are some of the most interesting widely used games that you offer?
EB: We have several great games. I'm always tempted to say that they're all great, but users are the ultimate judges. Our branded t itles are very exciting: Top Gun, True Crime, and several Marvel titles: Blade Trinity, Elektra Assassin, and the soon-to-be-released Fant astic Four. Our portfolio of casual games is also very attractive. It's amazing that a game like Operation or Connect Four can be so much fun and so appealing to users. These are among our most popular titles. You see, on a cell phone one wants either to save time or to kill time. We are delighted to see that we provide great time killers – media snacks – that fit the usage habits of mobile users.. Our multi-player titles also add the twist of communication, with in-game chat, challenges to friends, and so on.
QM: Is there a future for 3D wireless gaming?
EB: I'm a gamer myself and yes, absolutely, as a gamer I love realism. Certainly the traditional game market has shown that 3D brought dramatic improvements in games. Today's cellular phones have very limited capabilities, however. We have to be very careful how we use them, giving careful consideration to sparing the life of the battery and using image rendering, frame rate, and colors effectively. 3D can be overkill for some platforms today and possibly for a good while longer – and for a lot of games true 3D is not required. I am still amazed to see so many people playing solitaire on their laptops on airplanes. Casual games, casino games, and possibly other genres do not require 3D to be great entertainment.
Finally, I believe that today the ratio of the cost for producing a 3D game to the selling price and reach of 3D games is not very attractive. I am confident that reach will increase very fast over the next 12 months and I will soon revisit my position. In the long term, for genres where 3D is a mandatory enhancement, yes, 3D has a future.
QM: What is the development lifecycle involved in producing a wireless game?
EB: It goes like this:
Development of reference builds for some target platforms. Generally we group handsets in tiers that are to some extent homogeneous in terms of screen size, memory heap, maximum size of downloadable binary code, and so on
Certification: testing all possible use cases as in the traditional game industry, then extra tests concerning incoming calls, SMS, low battery, backlight off, etc.
Port: addressing other platforms once you have certified builds.
Certification of ports.
Localization and customization to inject carriers' specific APIs, like a billing API, a game API to manage your nickname, and so forth.
Certification in context: all network elements in place on the client and on the server.
Many steps are common to software engineering methodologies in general, but, as you can see, some extra steps are needed to validate the g ame in context; that is, taking into consideration the device and network themselves, and their connectivity attributes. In providing this list of steps I am not saying that this is a straightforward, linear way of doing things. You can use techniques like extreme programming to optimize the development cycle.
We also target J2ME platforms to start with, as porting from J2ME is generally simpler. I am not going to dissert about the beauty of one language over another, but Java offers formalisms and abstractions that facilitates porting to other environment.
QM: What is the most challenging part in developing games for wireless devices?
EB: Developing games is challenging in general. It's amazing to see how people can be tolerant of the numbers of crashes they have using Word, for example, and completely intolerant of crashes when playing games.
Games must be fun, flawless, colorful, absolute best, so you must dedicate a lot of R&D effort to understand platforms and fine-tune your games for each individual device. Of course, there are many ways you can factorize and optimize your development, but there are still lots of tiny little details, or even a completely different design, to take into account for each platform. Cell phones never, or should never, crash. Games on cell phones must never ever crash.
How do you certify a game? In an emulator? Alas, not. You certify games for wireless devices on wireless devices. That means you load the game, you start the game, you test the game, and you take notes of what is going on. And, as if these weren't enough headaches, you have to add network connectivity tests. You need to get your hands on wireless devices on various networks, on all continents!
There are very few solutions to manage all that complexity. The good news is that the industry is starting to address this problem with better standards, carrier-hosted test centers, tools to automate testing, and so on.
QM: What about multi-player capability in wireless games?
EB: It's my favorite! I read and hear a lot of contradicting things about multi-player games. Cell phones are, first of all, communication devices, and, as I like to say, chat is the simplest multi-player game: players are making the storyboard and the rules. Multi-player on cell phones is a natural fit. It's also a good feature that compensates for the lack of a big screen, a fast CPU, and lots of memory, so it is definitely worth considering in games. We see that our multi-player titles are very attractive to end users. We even see that when people have the choice to download a multi-player version over a single-player, they prefer the multi-player version.
QM: What are the security and privacy issues in multi-player g ames on phones?
EB: Well, the same as multi-player in PC or console games – pretty much the same issues as in any online activity. In fact, the mobile-phone environment is more secure: Wireless carriers want to protect their relationship with their subscribers, so when you sell your games through them they want you to provide some guarantees. As far as I know, to date there has been no known issue related to privacy or security arising from any multi-player wireless game from any vendor.
QM: How big is the wireless gaming market and where it is head ing?
EB: In some countries more than 80% of the people have a cell phone. In the future everyone will have a cell phone and no plain old telephone line at all, other than for receiving fax and using DSL until WiFi/Max arrives. Projecting conservatively, everyone will play at least one game on their cell phone. A good fraction of the world population will certainly purchase three or four games a year. With 6.5 billion people on Earth, we can envision a total of at least 20 billion game purchases per year at $5.00 apiece, so a market of $100 billion. I am certain I'm totally underestimating this market when ever I look at my kids and their friends, who are already downloading several games a month!
QM: Is there anything we can learn from the WAP experience?
EB: Yes: tenacity pays but the right message must be delivered. I was excited in 1997 when SFR deployed the first European Unwired Planet (now Openwave) solution on an Alcatel phone. That was when I started to think that phones would be the gaming device of the future. WAP was starting, and has made tremendous progress since.
What frustrated me the most was the motto that WAP was bringing the Internet to your phone. That was wrong and did not make sense, and was not even the intention. WAP was there to bring us data services in a browser. And it did so and continues to do so.
I think hype killed the true potential of WAP – but through the tenacity of many players in this industry it now has been adopted… at least to download Java-based games, polyphonic ring tones, and other content.
You are bringing me to the next topic: MIDP is a great environment to bring games to cell phones. MIDP uses some Java concepts but not all of them. Some hype that was floating around frustrated a lot of developers. Nevertheless, with tenacity MIDP is evolving in the right direction and a lot of developers are becoming more and more attracted by it.
QM: How does MFORMA's developer program work?
EB: We have a small developer community we are working with. They develop games that we propose to them, or that they propose to us, and we publish them. Over the past four years we came to the conclusion that this network must be small but open to new developers. You can find out on our web site how you can propose concepts and services to us. We have matured a development environment and SDK over time that we share with our developers. We also give them tools and support to test their applications and games on cell phones. It is a very fertile community with a lot of talented people involved.
QM: What is the MFORMA roadmap for wireless games?
EB: We have a very exciting roadmap for games that goes in very specific directions. We have struck deals with great brands in the recent past and can now release action and casual games with great consistency, using our agreements with Marvel, Hasbro, Atari, Universal, and others.
We are releasing 3D titles and multi-player titles, but what seems to be the most attractive is that we are building fun and entertaining games that really correspond to what our audience is looking for. We have spent quite some time in doing market research and usability studies and believe that our coming titles over the next months and years will be outstanding.
QM: Tell me about MFORMA's content management platform.
EB: We are managing more than 20,000 SKUs a year, so there is indeed a serious need for managing our content. Thanks to our early beginning in the platform business, we have constructed over the year a solid asset management toolset that allows us to track any product at any time, in any stage of its development.
Our distribution platform has also evolved. We can sell products either directly to the consumer or through our distribution channels – wireless carriers. As an example, you can order titles in North America from our web site and get them from carriers' vending machines. In the rest of the world, we can distribute either through carriers or directly to consumers.
QM: How does your distribution model benefit content providers and developers?
EB: We can distribute content in many ways, hence "mobilize" brands throughout the world. Our distribution capacity and support of handsets is a value proposition that's very attractive to content providers.
QM: Who are MFORMA's competitors?
EB: If you look at our games business, you'll find the large game publishers like Jamdat and Gameloft. If you look at our application portfolio, you'll see a list of smaller companies in different genres.
QM: One of the key issues in wireless game development is testing with the range of different devices available. How do developers at MFORMA go about that?
EB: They test their applications on different devices! There's no secret. You cannot test a car without crash tests, driving tests – actually testing in all situations: wet roads, dry roads, high speed, low speed, different sets of tires, and so on. It's essentially the same for us.
There are a few things we have done to maximize our efficiency: We are capitalizing on knowledge, and sharing a lot across our development studios and with our partners. To address several problems we have also made available proven libraries that our engineering team keeps up to date on many devices. Lastly, with the help of companies like Mobile Complete, we have organized remote access to devices so we can test content in real contexts.
QM: What should developers be aware of to make their games work on as many devices as possible?
EB: Trust what you see, and experiment! Then capitalize on what you have learned to make sure that you are not repeating the same mistakes. Experimentation is not just "Try something and see whether it works or not." When we run experiments, we instrument them as much as we can so that we can profile devices and networks into a formalism that we can reuse over time.
You can also rely on very useful public initiatives like JBenchmark, which to some extent are also understating devices in a systematic and universal way. Accumulating your own experience with what is publicly available builds a solid understanding of what is doable, what isn't, and the kinds of pitfalls you must avoid.
QM: Some of the advanced APIs for gaming developed through the Java Community Process (JCP) may not be available on all J2ME-enabled phones. How do your developers deal with this problem?
EB: I am guessing that the common approach today is to run conditional compilations or preprocessing. It's too bad that Java never admitted autocode or preprocessable code in its syntax. That would have helped here!
The technical approach consists of doing things like: "If my target platform supports this API, then do blah blah." What is more problematic is the design phase: Game producers are not necessarily into technical details and once they hear about a feature, they naturally want to use it. Developers have to spend a lot of time educating producers, or product managers, about what various platforms support and what they don't.
QM: In your opinion, what is the greatest achievement of J2ME today?
EB: Its simplicity and ubiquity. OK, even if J2ME is still far from “develop once, run everywhere” the goal is set to be just that.
The Java programming language is easy to learn, and now taught everywhere, J2ME is definitely simple compared with other environments. It's a breakthrough from where we were several years ago, when developing an application or a game for cell phones was possible for only a few guys on this planet.
The massive support by handset manufacturers and wireless carriers is also a great achievement. Call it J2ME or WIPI, or something similar, "Java for cell phones" is a given in the GSM world and with some CDMA carriers.
QM: Do you see BREW as a strong competitor to J2ME?
EB: BREW and J2ME are not comparable, because their spirit and commercial approaches are totally different. J2ME is a community effort. Yes, it's still linked to a Java license, but it adheres to an open community process, JCP. BREW and other environments, including OSes, are private initiatives. BREW is also an end-to-end solution, which J2ME is not. We are not comparing the same kind of thing, so the word "competitor" does not strictly apply.
They are indeed two different environments for developers. Developing a game or an application for J2ME is different from developing it for BREW. I am not talking about differences between Java syntax and C++ syntax, but rather more fundamental approaches that require significant efforts.
QM: What are your thoughts on wireless peer-to-peer?
EB: What comes to mind is Bluetooth and eventually Wi-Fi-related technologies. Real peer-to-peer on conventional wireless carrier networks is a tough one for security reasons and for what wireless phones are today.
On security: There are legitimate preoccupations from wireless carriers here. After all, the majority of us see a carrier logo right on our cell phone, so if we have a problem who do we call? Our carrier. When you do peer-to-peer file exchanges with your PC, and you unfortunately download a virus, you don't call your ISP to complain, but with cell phones it's different.
Cell phones are not PCs. The only elements that are active listeners on cell phones are the telephony functions, including SMS. You can certainly download applications that can do more, and simulate peer-to-peer, such as in multi-player games: You can always have a server that mediates communication between peers, but it's not true peer-to-peer.
So, beyond what is doable in Bluetooth today, peer-to-peer is not going to happen any time soon.
QM: What are some of the challenges that wireless technology faces?
EB: Education and simplicity. It is way too complicated for people to understand all of today's data possibilities. First, there are lots of obscure acronyms and names: WAP, Java, J2ME, BREW, and many more. Second, finding content is a real challenge. The potential is great but as in any new technology, usability requires a lot of work.
This reminds me of when I first introduced web sites to my wife in 1994. She asked me simply: "Well, what is it for?" I couldn't convince her to go on the web until 1996, when I did some research on an allergy that our first kid had. Since then, my wife has become addicted to the web and does not understand why we sometimes have 10 minutes of outages. Fortunately, I never tried to introduce her to Gopher, FTP, or things like that….
I am having the same difficulties now with her on applications for cell phones. So it's a mix of addressing use cases that make sense for people and the right way of presenting information.
QM: What are your thoughts on wireless advertising?
EB: We've conducted several experiments. Of course, from the perspective of making money it is very attractive. On the Internet, you can tolerate some animated GIFs here and there or, more annoyingly, skipping some ads, because there is a large screen and you don't worry about your battery, and you are installed in a comfortable chair. On your cell phone, advertising has to be very smart, apropos, and unintrusive. Think entertainment, or saving time. For example, after a 411 search gives me the phone number I'm looking for, I don't mind receiving a text message like "Brought to you by…."
QM: Do you see a future for incorporating Flash in J2ME-enabled cell phones?
EB: Yes, but I have yet to be convinced. So far I have not seen Flash support for many phone models. Maybe it is pure speculation, but I have the feeling that, because Flash is a very rich environment, it cannot fit all phones. It's possible to imagine degraded versions of Flash for some families of phones, but then the beauty of Flash is gone – and in fact developing for Flash will become terribly technical.
The only Flash developers I know are more designers than programmers – which explains the kind of Flash applications you can see out there.
When it comes to games… Flash can be picked up for casual games. For action games, we need a more powerful development platform like J2ME.
QM: What do you envision will be possible in wireless gaming two years from now?
EB: Everything! Cell phones are the future of portable gaming platforms, at least in my opinion. But it goes beyond gaming as we know it: Native cell phone attributes will certainly be injected in future games: voice, location, messaging, presence awareness, etc. The context of use also will influence content. Often people joke about hitting a pole while walking and playing at the same time. It is not this context I'm thinking of. Your cell phone is always with you, in your pocket or purse. Your cell phone is your dear companion, much more so than your mailbox. You can expect new genres of games that take advantage of that closeness.
QM: Do you have any advice for developers of wireless Java-based applications?
EB: Be tenacious, think usability and mobility. We are at the very beginning of an era and there is still a lot to explore. Tenacity is required to understand all the technical and business details of wireless applications and games. It is not that, because Java is there and devices are small, developing and selling apps is a simple thing to do. On the contrary, it is awkwardly complex.
In all cases, though, the focus is to provide a great time killer or a very useful time saver. As everywhere else, the customer is king, and should be the main topic of discussion… the technology should be invisible.
About the Author
Qusay H. Mahmoud provides Java consulting and training services. He has published dozens of articles on Java, and is the author of Distributed Programming with Java (Manning Publications, 1999) and Learning Wireless Java (O'Reilly, 2002).