Mario of Badlogic Games has agreed to answer a few of our questions.  Badlogic Games are the people behind the physics puzzler Newton, which has recently broken 10,000 downloads on the Android Market.  No mean feat.  As well as developing in Android, Mario writes a blog detailing his experiences with developing Android games.  As part of the blog, Mario has let his readers in on his next Android project, which is pretty exciting – an Audiosurf inspired procedurally generated game which uses the player’s music to generate gameplay.

What’s your development background?

I’ve been developing apps, games and libraries for 16 years now. I started with QBasic and x86 Assembler back in 1994 at the age of 11. My parents bought me a 40Mhz 486 with an incredible 8Mb of RAM. Over the years i got my hands dirty with some other languages like Pascal, C/C++, Java, C#, Mips assembly, a bit of Python and some nasty Php. At university i did some functional programming which was a neat experience but i wouldn’t recommend it for production if you want to keep your mental sanity. I’ve worked at a company in Graz, Austria, named Know Center, since 2006 where my job is mostly about programming in Java. It’s a research facility so i do some scientific things in the area of natural language processing and supervised/unsupervised machine learning. My number one love was and still is game and graphics programming as it is what i grew up with. My initial motivation was to work out how those nice little Game Boy games come to life back in the early 90ies.

Newton is an ongoing project, with additional levels being worked on at time of writing, but how long did it take to get the initial version of Newton to the Android Market?

Getting the initial version done took about 2 weeks. Most of the time i spent on writting a game programming library that allows me to run all my games on the desktop as well as on Android. All graphics, input, sound and file i/o is wrapped and uses the respective backend libraries for each platform. The game itself took maybe 4-5 days. Publishing it to the market was a piece of cake really. That’s something Google got right in my opinion.

Which Android device(s) do you use yourself?

I have a Motorola Milestone (Droid in the US) and a HTC Hero which i gave to my girlfriend Stefanie  who’s also part of Badlogic Games.

Is there anything you wish you’d done differently with Newton?

A lot of things. The presentation is kinda boring and unimaginative. Maybe going with the usual doodle style would have given a few better ratings. The 3D version doesn’t look all that impressive either. I wanted to get the game out as fast as possible and couldn’t come up with a better style in time. Apart from that i think the rest of the game is pretty solid.

You’re working on a very exciting concept for your next app, a music-adapting game that uses mp3 music to generate gameplay.  Can you tell us a little bit about it and what your vision is for how the final game will actually play?

I stumbled across Audiosurf a couple of weeks ago and was pretty impressed with the idea of procedurally generated levels based on music the player can provide himself. As oposed to games like Guitar Hero where you only get a set of predefined tracks the publisher thinks will work with most of the audience the procedural approach allows for a much more customized experience. I think this is essential for music games as music is a matter of taste. I want the game to adapt to a player’s musical taste as much as possible and with the procedural approach this becomes a possibility. It also removes the need for me to create levels myself which i pretty much suck at. I’m not exactly sure where the game will head but it won’t have any puzzler elements like Audiosurf does. I will probably focus on a more arcade style racing game. Maybe i’ll go a completely different route though, away from the racing theme. It all depends on how well the audio analysis stage performs in the end. I think Audiosurf did a good job for specific genres but fails horribly for others. If i can get that part done well enough it opens more possibilities.

From reading your blog, I can see the music-adapting game project you are working on currently involves a lot of research in to sound recognition and beat detection, how much time are you investing in research and do you always use such a scientific approach to development projects?

Researching solutions for the problem at hand always takes up a lot of the total development time. Usually there’s not a lot to research for games like Newton as i pretty much know how to go about stuff from the beginning. The audio project is a bit different. I have some experience with processing audio but not enough to get the job done good enough for this project. Audio processing is a mathematical problem so you won’t find a lot of the usual tutorials out there like you get for say OpenGL and the likes. As I do scientific research in my day job I’m used to digging my way through scientifc literature and there’s a lot of good stuff there on the problem of onset/beat detection. So going that route was a no-brainer really after realising that there’s just not enough material out there that is non-scientific and covers all my needs.

Newton is a free app, and is growing rapidly in popularity.  Do you have any plans to release paid for apps in the future or will you be sticking to freebies?

The version of Newton on the market is a kind of test balloon to get a feeling for how the market behaves. It might not be ideal in that regard as Newton is not a game that appeals to everyone. Never the less i gained a lot of insight in publishing a free app with just enough content to get a user interested. I plan on releasing a full paid version for Newton as soon as Stefanie finishes of designing all the new levels needed to justify demanding money from people. Along with the new levels i already integrated a level editor and the possibility to share and download user generated levels in-game. The audio project will definetely be a paid app. However, i think it’s just fair to release demos for people to try out a game before they decide to buy. So there will of course be a lite version that will let you play a few of your songs to get a taste of the game. In any case i don’t expect to get rich but i’d love to get just enough money out of the paid versions to amortise the cost of my two Android devices

What are your favourite Android apps and games?  Is there an app you couldn’t live without?

The one app that i use most often is TaskPanel, a free task manager that allows me to kill processes. Also useful for me as a developer is SendLog which allows people to send me the logging output of my crashing apps. Apart from that i use the Mother TED app from time to time to view some TED talks. Astro is my file manager of choice which is pretty awesome. Apart from that i don’t have any other apps installed on my phone. On the gaming side i have lots of games that come and go. I mostly install and play games to check out other peoples work and compare the experience i have with the user ratings and comments for the game. It’s a bit of market research i guess. I haven’t yet played any game for entertainment or against boredom to be honest. I was impressed with the games by Polarbit but at the same time a bit disappointed that they just did straight ports of the IPhone games which are pretty old already. The games by Hexage have a nice quality to them which i just can’t describe in words. They just feel right. Robert Green’s Battery Powered Games are also pretty good. My girlfriend can’t stop playing Light Racer for some magic reason that eludes me.

How have you found developing in Android.  Is there anything you feel could be done to improve the Android development process, anything missing that would be useful?

Developing for Android proved to be extremely easy. The integration with Eclipse and the very good documentation let you get things done in no time if you already have a Java background. Doing some minor native coding can also be accomplished pretty quickly with the NDK. However, i wished that the NDK toolchain would use the standard *nix build process. It’s damn annoying to get some open source libraries that would just compile and work fine on ARM to work with the NDK toolchain. Apart from that i found a lot of bugs in the graphics drivers which is not Google’s fault be the phone manufacturers fault. OpenGL ES does some really weird things on many phones which makes testing for a small developer like myself really hard. Financially more potent teams have the advantage of being able to test their code on many more devices than i can, guaranteeing smooth deployment. A lot of my initial bad ratings had to do with bugs i just couldn’t reproduce due to me not possesing say a Samsung Galaxy. The emulator is no help there. It would also be nice to have access to the media file decoding facilities. That’s something i could have really used in my audio project and finding that it can’t be done without porting a couple of open source decoders for mp3 and oggs was kind of frustrating. The most frustrating thing was to discover that multi-touch on all Android devices that claim to support it actually doesn’t work. I shared this frustration with Robert from Battery Powered Games. Game developers can’t do multitouch games on Android like you find them on the IPhone. That eliminates a lot of genres, especially arcade titles which most often use a dual stick setup. Even worse, the reaction of the Google framework developers came pretty late and wasn’t giving any insights on how to fix the problems. The only urged us to not fill bug reports as this is a hardware problem. It’s really kind of sad. Other than that i love programming for Android. It’s an exceptionally well thought out environment.

How much potential do you feel the Android platform has?

I’m not much of an oracle i really can’t predict how Android will work out. Taking the viewpoint of a user i love Android and would wish it to be as successful as possible. The platform is really awesome and can compare easily with Blackberries or IPhones in its latest iteration. Seeing how a lot of major manufacturers will release Android phones this year i guess Android might well have a bright future. I’d love that to be the case but only time can tell. It will probably take up a lot of the market but the process will be very very slow.

Any advice for aspiring Android developers?

RTFM! Know your way around Java and Eclipse and don’t be afraid of using native code for better performance. Get onto the android mailing lists and ask intelligent questions there, help will be given to you. And don’t forget you are working on a mobile platform, you can’t do just anything there.

Thanks again to Mario for taking the time to answer our quetions.

Newton is available now for free on the Android Market.  The Badlogic Games blog can be found here and is well worth a read for anyone who is even remotely interested in Android development.