Java Game Console

Game Play & Game Design
#1

Short Version:

I have an idea for a Java Game Console. Go to http://jgcwiki.datadino.com for more info.

http://jgcwiki.datadino.com/jgcwiki/images/7/7a/Game_Console_(Thumb).png

Long Version

As of late, I’ve become infatuated with a few different and radical gaming ideas. One of them concerns the matter of porting Java games to consoles. It would seem that we are consistently stuck with a chicken and the egg problem: No one will invest in coding commercial games in Java until they see a successful commercial game done in Java. And no one can create a successful commercial game in Java until they can aquire a sufficient investment.

The traditional method of exiting this loop (independent developers) is no longer an option as games continue to grow to a level of complexity that requires tens of thousands of man hours to complete. This complexity makes the games unbelievably expensive to produce.

My solution is to go back in time a smidge and again empower the independent developer. i.e. Give developers a chance to cut their teeth on small-time, 2D projects.

As this idea has grown in my mind, I’ve been spending time trying to learn how to develop the necessary hardware. (Santa is going to be bringing me one of these this year. ;-))

The final hardware must be in the $25-$50 price range for the console, with individual games retailing for $5-$15 a piece. The console and games would sell on the toy shelf next to handheld games, instead of in the more “sophisticated” electronics section. Such low prices would encourage children and parents to make an “impulse buy” on both the console and software. The graphics would be on par with the Super Nintendo, and the media format would be extremely small.

I’ve given quite a few more details on this page:

http://jgcwiki.datadino.com

Feel free to add information and files to the Wiki. It’s open to the world so that anyone who wants to collaberate on the idea can do so.

I’m not sure how far I’ll manage to take this project. I’d like to take it all the way to a commercial product, but things don’t always work out that way. :slight_smile:

What do you guys think?

#2

I went through this cycle just recently, see http://www.java-gaming.org/cgi-bin/JGNetForums/YaBB.cgi?board=Offtopic;action=display;num=1093790338. The biggest problem seems to be getting hardware that will run reasonable stuff while not costing too much.

I got myself one of these http://www.asus.com/products/desktop/digimatrix/overview.htm that really looks the part and runs Java reasonably, although OpenGL drivers arn’t good yet. However, price wise your talking ~$800 all in.
Kev

#3

[quote]I went through this cycle just recently, see http://www.java-gaming.org/cgi-bin/JGNetForums/YaBB.cgi?board=Offtopic;ac tion=display;num=1093790338.
[/quote]
That’s pretty cool, Kev! Very reminicient of the X-Box or Phantom Games Consoles.

[quote]The biggest problem seems to be getting hardware that will run reasonable stuff while not costing too much.
[/quote]
Indeed. PC hardware is tremendously expensive due to the sheer amount of hardware in any given box. That hardware then begats more hardware because you need a place to store an Operating System, BIOS, and driver software for all that hardware. Amazing at it may sound, however, PC hardware can be produced in mass quantities for about $200 a unit. Consoles like the X-Box actually take a loss for every console they sell, then make up the difference in game licensing.

My solution is a bit different. Have you ever seen those controllers with old games that are sold at Walmart and Thinkgeek? They are complete consoles, preloaded with 5-10 games and sold for about $20 US. The way they are able to accomplish this is by designing and manufacturing a custom ASIC. If you were to pull one of those controllers apart, you’d find nothing but a PCB board with a single microchip epoxied onto the board!

That single microchip (ASIC) was designed with the ROM to hold the games, the processor to run them, the NTSC Video Encoder to drive the television, and the interrupts to listen to user input. Such a chip probably uses about 100,000 NAND logic gates and costs about $1-$5 a chip in quantities of 250,000 or more. That’s right, I said about $1-$5 a chip! :slight_smile:

For the JGC project, I’m recommending something similar. All the logic would be on one chip, thus reducing manfacturing costs to negligable amounts. For prototyping, I’m already looking at a parts list that includes one FPGA (a rewritable microprocessor) running at 50-100 MHz for ~$40, a PCB board that runs about $18 a board, a custom manufactured case (haven’t got a quote yet), and a cheap gamepad (about $9.95). Production hardware would replace the FPGA with an ASIC, and would cost about 1/5 of those prices due to quantity discounts.

Almost all processor cores are coming from OpenCores with the primary CPU being the Java Optimized Procssor. All of these cores are written in VHDL, so testing is as easy as popping them on an FPGA. And before you get scared, here’s an example of a simple microprocessor:

library ieee;
use ieee.std_logic_1164.all;

--------------------------------------

entity OR_ent is
port(      x: in std_logic;
      y: in std_logic;
      F: out std_logic
);
end OR_ent;  

architecture OR_beh of OR_ent is 
begin 

    F <= x or y; 

end OR_beh;

That’s right, it’s written in software! :smiley: A program called a “logic analyzer” goes through the pain of actually producing the physical layout of the chip. Nothing much for the chip developer to do, other than write some code!

So what this all adds up to, is that $25-$50 is a very reachable price for the design I’m working on. I can’t wait until my FPGA kits shows up! ;D

#4

Thats pretty damn funky actually. I might well have to stop by and take a look…

Coooool

Kev

#5

[quote] Posted on: Today at 9:17am
Thats pretty damn funky actually. I might well have to stop by and take a look…
[/quote]
/me takes a bow :slight_smile:

I’m still working out a lot of the details, but here are a few points that you may find particularly interesting:

Planned Image Compression Technology (Guess that 4K contest was good for something!)

Graphics Hardware Design (Seriously Old Skool, dude! :D)

Choice in game media (Trade 'em with your friends!)

Links to inexpensive and easy to use manufacturing (PCBs, Plastic Injection Molding, metal works, laser cutters, etc.)

P.S. Every page on the JGC Wiki has a “Discussion” tab on it. You can feel free to flip to this tab and add notes or comments. See Wikipedia for examples of this in use. :slight_smile:

#6

One thing to say: Put together a business plan and try and get some “seed” or “pre-seed” funding for this as a commercial and/or academic project.

if you get a few thousand dollars funding to get started - not even to make a full prototype (although that might be good) - then you will be in an excellent position to get real serious funding later on. Believe me, you are MUCH better funding a prototype off a tiny amount of other people’s money than your own, it is pricelessly valuable in making VC’s and private investors consider any/all future business plans if it has been funded NOT privately to date.

Also, if you live in EU, the EU has a lot of grants and matched-funding schemes which you effectively ONLY get onto if you start with them before you even design your first prototype, let alone build it. In the US, its slightly similar in that (so I’m told - but I have no direct experience of this) it’s easier to get NSA etc grants right from the start than to mock something up and then go asking.

And, hey, if you can get even a couple of hundred dollars to help fund this, why not? :slight_smile:

#7

That’s some good advice, blah! I’ll have to consider doing that. AFAIK, getting a bit of seed money from the government is quite easy here in the states. I’ve never tried to access these funds myself, but my understanding is that they are allocated by capital hill to promote the start of small businesses.

One thing I’d like to add, however, is that I’d like the Java community to have a say in every step of the process. For one, I’m trying to build this so that people here could produce games that make money. To do that, I need to make sure that everyone gets what they need in a console design. (Thus the Wiki.)

What this means is that the more work that can be done on the console without corporate interference, the better off we all are. I’ve even been considering if it would be feasible to eventually turn this into a Co-Op instead of a corporation or LLC. Unlike a corporation, a Co-Op would have a responsibility to its members rather than its shareholders. That tends to give them a certain degree of freedom that corporations lack.

Of course, this is mostly speculative at this point. The project may never catch on, and I may eventually lose interest. But I think it would be VERY exciting if we could take it to completion. :slight_smile:

#8

Consider the LWJGL team to be right behind you :slight_smile:

Cas :slight_smile:

#9

[quote]Consider the LWJGL team to be right behind you :slight_smile:
[/quote]
Thanks Cas! You’ve probably got more real-world experience with the old gaming hardware than any of the rest of us here. (Err… at least me anyway. :)) So feel free to pop any thoughts you have into the Wiki, or discuss them here. :slight_smile:

#10

[quote]To do that, I need to make sure that everyone gets what they need in a console design. (Thus the Wiki.)
[/quote]
Govt grant people would see this as “good market research” and would probably even give you money to do more of it (e.g. to run the website, etc). The game is to get small amounts off them for everything - that way, you show that you already know how to be responsible with other peoples money, and build up a history of “NOT running away with grant money and disappearing” which, believe it or not, is very important to them :).

Makes their job easier (less worry of being shouted at for making bad investment decisions).

Go to the grants people. Say this to them. Ask them if its feasible etc. Dont use it as a reason NOT to get grant money. The history of 99.99% of projects suggests you’re better off making a research project with research grants and worrying about the community aspects later. There are many ways of making a company effectively run by the community. There are few ways (none?) of taking a community project and then getting govt money for it in arrears.

You can even get special grants that are only allowed for people making not-for-profit research etc “for the common good”…

#11

Ok, Blah. You’ve convinced me. I’m going to spend a little bit of time tonight on figuring out how to get Uncle Sam to cough up some dough. Thanks for the tips! :slight_smile:

#12

One good thing about using the LWJGL on this console is that the JVM need only be very small (no AWT port etc needed). What’s more you’ll have at least 5 top-quality games to launch with :wink: Very difficult to convince “real” developers to write games for the thing - how fortunate that we’ve already written them and they should “just work”.

Cas :slight_smile:

#13

Umm… Cas? There’s just one problem with LWJGL on this console. (Actually two.) For one, the hardware has no vector logic. This is purely old school sprite hardware, not 3D Vector hardware. The second problem is that LWJGL would take up a massive amount of ROM, and the textures would absolutely destroy the capacity of the smart cards.

A more powerful console could be built, but it would look a lot more like a Gamecube or PS2, and have the hardware costs to match. Sorry. :frowning:

(Unless you have ideas on how to add inexpensive Vector hardware, that is…)

#14

Inexpensive Vector hardware = all those old GeForce1 chips lying around. A GF1 is more or less as powerful as a PS2 I believe.

Actually how much does an original N-force mobo cost?

The thing is, 2D hardware is sucky and difficult to write for.LWJGL shows how easy both 2D and 3D can be in the same API. And it gives you a library of games to launch with. Why not make it the target? I’d buy such a console if it cost $150 or less I think. Can a GF1-based console be produced for $150? You bet it can. Dell kick out entire PCs for $250. In fact I think that if you’re really going to be successful in this venture you’re going to have to use commodity PC components and Linux for your OS, because primarily your problems are cheap common hardware, drivers, and a JVM.

My take on minimum specs is mysteriously similar to those of Alien Flux :slight_smile: eg. 64MB RAM, 500MHz chip, GF1. No CD ROM, just four USB ports to plug in a broadband modem and gamepads, mice, keyboards etc.

Then if you wanted to make any money to it, you license the console to ISPs and hardwire the internet connection to go to their special console homepage and deliver their licensed games to it. This leaves the problem of distribution of your console up to ISPs.

Cas :slight_smile:

#15

Whilst I don’t agree with Cas’s extreme demands :slight_smile: (I think he’s missing the target audience here slightly) it would be interesting to see just how cheap you could pick up old 1st-gen GF’s.

I’d suggest, in context, get the funding to research the 2d version. Then get funding to research how much it woudl “cost” (mainly in terms of effort!) to make a working one using GF1’s - it may not be as horrific as it first seems.

#16

It would be very interesting to know how cheap we can get 3D hardware. For a second generation console, that is. :slight_smile:

Cas, I’m not quite sure you realize what you’re asking. This console is going to be so cheap because most of the hardware will fit on the chip, and third party licensing costs will be minimal, if not non-existant. The cost of producing the console also has to leave a sufficient price margin to fit marketing and shelving costs.

Let’s assume for a moment that NVidia licensed the GeForce 256 for ~$5 a chip. (That’s already 1/10 the final price, BTW.) We’d then need a 3D API to support it. Let’s go with OpenGL/LWJGL. Now because of OpenGL, LWJGL has thousands of symbols. Assuming we trim down the class file format a bit, that gives us about 100K for the symbols alone. Now we need a driver to implement the logic for each of those symbols. Probably the only option would be to license from NVidia.

NVidia’s driver is quite large, on the order of several MB. Now ROM is cheap, but not that cheap. Manufacturers never reveal their prices until they have a relationship with you, but it’s safe to say that an 8MB ROM chip will cost more than a 512K-1MB that was originally planned.

Now take the game size into consideration. 2D games can get away with small sizes because they look “cartoonish”. (See Image Compression Technology for more info on how I plan to take advantage of that.) 3D tends to demand more photorealism in the graphics. In addition, more graphics are needed, since they need to cover significantly more virtual surface area than the flat panel 2D games have. Obviously, the 8Mbit smart cards are going to need an upgrade. Using the Nintendo 64 as a guide, we’ll need at least 32MBit media. The N64 goes as high as 512MBit (64MBytes of ROM). So now our media busts the $5-$15 market price.

Of course, we need RAM to load all that data, so we’re bumping the ram up from 4-8MB to 64-128MB.

In short, 3D would make the console miss its $25-$50 price tange, the media to miss its $5-$15 price range, the console could no longer be sold as a toy, it would now compete with the X-Box/PS2/GameCube, and it would lose millions of impulse buys.

I have to go, so I’ll explain more on why the second generation might work later.

#17

I think that attempting to sell a console for $50 is probably a complete waste of time. Anyone actually looking for a value console will be looking at a secondhand PS or PS2. The market is thoroughly saturated.

Your only hope of success is in offering something different and on an equal price footing. Carve a lucrative niche into the market.

Who will buy a $150 console? Well, anyone who is interested in $10 games delivered digitally. Or anyone who would like to develop console games without having to pay for a dev kit and licensing. Nerds and hobbyists - the people who founded the home computer revolution in the 80s still exist, they’re just younger than us and they don’t have much simple stuff to play with.

On the other hand, who will sell a $50 console, subsidised by game revenue? Any ISP that wants to make a $10 cut on every game sold. And they could be talking hundreds of thousands of games. A nice little earner.

Cas :slight_smile:

#18

[quote]I think that attempting to sell a console for $50 is probably a complete waste of time. Anyone actually looking for a value console will be looking at a secondhand PS or PS2. The market is thoroughly saturated.
[/quote]
That’s why I’m not targetting the console market. :slight_smile: It’s a toy, Cas. It will sit on the shelf at Wal-mart next to these and these. You’ll find the little smart cards nearby, or in the checkout racks. Parents and children will see these things and say “I can afford that” or “That looks like fun!” And the form factor means that they can trade video game cards with their friends.

If these things end up anywhere near the electronics section, then it has already lost the market.

[quote]Who will buy a $150 console? Well, anyone who is interested in $10 games delivered digitally.
[/quote]
It sounds like a great idea, but you need a LOT more resources to make it happen. i.e.:
[]You’re breaking even on the manufacturing cost (and actually losing money due to operational costs).
[]You have to run servers and pay for staff to run those servers.
[]You need to figure out how the average person will connect this to the internet. (Is a computer required? Does it have an Ethernet and modem?)
[]The complexity of online connectivity means that there will be support costs for the device.
[]You need to have deals with major gaming providers for world-class games. (How else do you expect to convince a teenager to buy a $150 console over a $80-$100 XBox or PS2?)
[]The independent developer will again be unable to compete

[quote]Or anyone who would like to develop console games without having to pay for a dev kit and licensing. Nerds and hobbyists - the people who founded the home computer revolution in the 80s still exist, they’re just younger than us and they don’t have much simple stuff to play with.
[/quote]
Ah, but the console I’m proposing already IS accessable. Card readers and writers are quite cheap, as are smart cards. Complete dev kits (e.g. a modified console with debugger) could easily be sold for a hundred dollars or so. Because it’s in Java, development could happen on the desktop before the hardware.

Yet one difference between a $150 console and a $50 one, is that the $50 one allows the hobbist to become a commerical game developer.

[quote]On the other hand, who will sell a $50 console, subsidised by game revenue? Any ISP that wants to make a $10 cut on every game sold. And they could be talking hundreds of thousands of games. A nice little earner.
[/quote]
Now you’re getting a bit fanciful. No company is going to eat a $100 loss expecting to be subsidized by $10 games. Even if we assume that the ISP gets a huge chunk of the pie (say 20%), he’s only going to make ~$2 off of every sale! He’d need to sell 50 games per customer, just to break even! That’s about what a customer can be expected to purchase over the life of the console! (I don’t have the exact reseach on hand; it costs about $2,500. But these numbers should help paint a picture for you.)

Basically, such a business model is doable, but it’s not for a guy in his basement to make happen. And as to your console cost, there’s one more part you need to consider. If the games are to be delivered over the internet, the console will require a large hard drive. A 40gig hard drive retails for about $80. Pulling a guesstimate out of thin air, that would result in a part cost of ~$40. That’s 1/4 the price of the machine! Plus you’ll have to develop hard drive management tools, game initialization and execution tools, installer standards, etc.

Kevglass had it right when he said that such a path is “too expensive”.

Now second generation hardware…

Consider that this “toy” is retracing the steps on Nintendo and Sony. Once the console and 2D games are on the market, the company that produces the console will have clout with manufacturers. Anyone who sells 100,000 units or more is going to be someone that manufacturers will want to work with. That means that cheaper licensing arrangements can be made, and custom hardware can be demanded.

e.g. If I was in that position, I would send a request for bids to Sharp, Samsung, Sony, Kodak, SanDisk, and other electronic storage manufacturers for a scalable memory device with a maximum of 512MBits of capacity, small form factor, inexpensive labelling ability, and a low unit price. They would then compete to produce the winning device. That would solve my storage problems.

One thing I would also do, however, is contact the major game publishers, and invite them to work on the machine. That may seem counterproductive, as the independent guy will get creamed with such competition. The reality is, however, that a competitive 3D game is outside the reach of your average coder. He can usually produce the basics of the game, but all the details and art required by the 3D game rend to be beyond his capacity to produce. That’s why most developers take their partially completed game to a publisher for final release.

#19

If the console idea gets far enough along I could help with some things.

I have 20 years exp in the architectural field, commercial and residential buildings. This means I have lots of cad experience and could do the drawings required for emachineshop. I have experience with AutoCad and Lightwave.

I’m also an artist. I could do package art / graphic design work, or get some professionals to do it. I know several good graphic designers.

I also worked at a tradeshow company, ExhibitgroupGiltspur and I could organize any presence at E3.

Of course thats way down the road, but at least now you know someone with those skills :slight_smile:

( Are you sure those console dimensions on the wiki are right? )

#20

That’s great to hear Tzan! Thanks for offering your services! The case “look” was one thing I was particularly worried about. From the render, you can see that I can design straightforward stuff without too much difficulty. What I lack is the more “artistic” side of modelling that makes customers drool over the various consoles. :slight_smile:

As for the case dimensions, those are correct. (Or at least I hope I can make them work…) There’s no real reason to build a large console, and a small device helps re-enforce the idea that the device is a toy. Of course, the final hardware may be of slightly different dimensions. The height is probably the

BTW, the 3D rendering was done with 1 unit == 1 mm, so you’re seeing more or less exactly how it would look. I just wish I had a scaled model of some keys or something to show how small it actually is. :slight_smile: