Description I'm currently working on a puzzle/action casual game with a small team of developers, and after showing it to a few people at the GDC, the feedback we've gotten is that they would like to see a cellphone version of the game. Its currently in the beta stages and you can check out the most recent screenshots at http://www.agfrag.com/cp.html. Its also currently geared towards 4-player action, but we'll be scaling it down to 1-player for the cell phone version. Question: Is there currently a large enough percent of cellphones on the market using BREW technology that it would make sense developing the cellphone version using BREW and OpenGL ES? Or should I look into programming it with J2ME? Would porting the code to both BREW and J2ME be a better option? I apologize if these questions sound stupid, but I have no idea as to what technology is currently driving games on most cellphones these days. Our expected release is withing the next few months. Its a 2D game so I don't need 3D hardware support, but there's no use developing for outdated hardware, right?. Does anyone have an opinion as to what the best option would be for a game like this? I'm using OpenGL for the game, so I was leaning towards BREW with OpenGL ES. Also, I thought picking up a copy of "OpenGL ES Game Development" to get up to speed with OpenGL ES. Any comments about the book? I know Dave co-authored it, so its got to be good, right? Current Project: Puzzle/Action Game called Circuit Pyramid™ Download (Note: Beta Version) Screenshots

Depending on where you want to distribute your game I'd say either J2ME only (Europe), or both J2ME and BREW (America and the Far East).

Quote:Original post by wyrzy
Description
Our expected release is withing the next few months.

You can achieve that in J2ME (you'll probably need at least one person with experience in J2ME though so you don't get caught in the many "gotchas"). Doing both a J2ME and a BREW version (and the many many ports you'll need) in that timeframe is a bit of a stretch.

Quote:Original post by wyrzy
Its a 2D game so I don't need 3D hardware support, but there's no use developing for outdated hardware, right?.

There most certainly is use in developing for "outdated hardware" because it isn't that outdated. OpenGL ES supporting phones aren't very common and most of the phones in the market currently support the oldest versions of both J2ME and BREW (ie, MIDP 1.0 and BREW 1.x). In this business it's wise to develop for the most outdated hardware that can run your game.

shmoove

Quote:Original post by wyrzy
Its a 2D game so I don't need 3D hardware support, but there's no use developing for outdated hardware, right?.

Wrong. As was stated above there are few phones that support OpenGL ES. Also a large majority of people with phones do not have the high end phones. They have either the phones given away with their plan or a few levels above it.

I'd suggest you knock up a demo for a J2ME handset (initial BREW dev costs are high) and show that to potential investors.

Mobile development is neither cheap nor easy. Carriers will most likely ask for a number of handsets to be supported, all of which you'll need probably have to buy to test on, and may well shy away from a non-licensed game. The BREW/MIDP implementation will be non-standard in some way for each handset, meaning you'll need to work on each port individually (BREW is actually much better here - the java team at my company have a hellish time with poor MIDP implementations and manufacturer-specific features) You will need to write clean, optimised code designed around hardware with no GPU, FPU or even integer division support (forget about OpenGL ES - the people paying for games at the moment are mostly all using slow, underpowered hardware) You'll need to design your game(s) around very restrictive screen sizes, RAM, input methods, etc. If you're developing for BREW you'll need to pay for your game to be tested by Qualcomm's approved testing company (NSTL)

Putting together a demo on a J2ME handset is relatively easy and cheap, which is why I suggest it. To move on from that to develop the game for a collection of BREW/J2ME handsets is a whole different matter and WILL cost you $1000's in equipment, dev time, testing costs, and so on. That next step will require substantial investment and staff with the desire and ability to enthusiastically work on titles that will generally suck horribly compared to any dedicated gaming hardware platform.

Sorry if that sounds negative, but that's the reality of the industry. It CAN be a rewarding and challenging field, but it's one to which you will have to be wholly committed, and one in which the profit margins are often very small.