Not at all, but it does rather limit my opportunities though…
Cas 
[quote]I suspect for the small companies it is partly using what you know, they come from a gaming background and are instantly productive in C++ as the know it inside out what what libraries to use. Java is an unknown to them and therefore a risk.
[/quote]
While that’s probably true for the traditional high profile games industry that can’t afford any risks, but I would have expected a trend away from C++ (hopefully in favour of Java) as C++ becomes less and less used in general.
In schools, people generally learn java to learn OO programming, so I would have expected more and more students to start developing (indie) games using it. Unless they’re tought that java is unsuitable for games? (just thinking out loud…)
I suspect that these people are using something even higher level, like python or Blitz. Or they stumble into various internet fora and get told Java is slow and that they should use C++.
In my experience, not many students try their hands on game development in the first place. I’ve been in two universities (one in the US, one in Belgium), and none of the people I knew had any ambition to write a game of their own.
I think indie game development comes mostly from die-hard grizzled old vets, who were used to assembler en PEEK/POKE on their Amiga. They grudgingly have taken the step to C (and I leave the “++” out intentionally, as C++ allows you to write plain old-fashioned procedural code), but they’re afraid to take another step, to a platform that has an image of being “slow”.
The flash community is a different story. I’ve seen guys with no programming experience do incredible things in Flash, shoot’m ups, adventures, animations, state-based stuff, and whatnot. But they refuse to switch to java, because it’s considered a programming language and they don’t think they “can program”.
So from java’s point of view, you have two communities who (generally) refuse to switch to java: the vets who stick to their old guns, and the flash people who are afraid of the word “program”. On the other hand, most java developers are more interested in the business applications than they are interested in games.
This is all, of course, generally speaking. Present company excluded
I suspect those people at university wanting to enter the games industry are learning C++ in their spare time.
For people coming out of the game industry C++ is what they know best as they have been doing it day in day out. PowerOfTwo games (another start up of ex- main stream people) makes the the point in their blog about prototyping http://powerof2games.com/node/11. Further illustrating the mindset and perhaps demonstrating why for them it is lowering the risk.
Once you remove the people coming out of the mainstream game industry and those wanting to enter it from the indie game development scene seems to use quite a diverse range of programming languages. It would be interesting to see statistics if they existed.
On a positive note Java has more high quality indie titles now than at any time in the past
Paul
Well, there are always going to be more titles today than yesterday aren’t there 
Interestingly, having this very argument with a diehard C++ studio elsewhere, “slow” never comes into it any more. Even the diehard C++ coders concede that Java is fast. The problems as they perceive it are:
- Steep learning curve and associated wasted time, not just the syntax, but the APIs and IDE
- No Direct3D interface, further steepening learning curve and reducing Windows compatibility
- No established 3D scenegraph middleware in the same league as, say Unreal Engine or OGRE or Irrlicht
- No console development possibility
- Existing C++ codebase and skills suddenly redundant
Cas
JVM Ports to Consoles would be a good leap for Java Gaming. Meanwhile, i think you can make a good money at PC Market.
jMonkeyEngine, Slick and other tools are earning features quickly. We just need few good games to show the power of Java Games to the world.
In my opinion, the “No console development possibility” is the main issue. Why game studios, not including casual gaming ones, would start using Java technology and the tools that come with when they can only develop for PCs, and mobile devices using Java? Those developers would have to maintain their now bigger skill sets in both C++ and Java. This situation is not appealing at all for developers.
if you’re worried about lack of Java adoption limiting your opportunity, you’re missing the real problem. no matter what language you use, whether it be Java, C#, C++, etc., you can NEVER place all of your eggs in one basket.
the only way to advance in the industry is adapt to it. sure C# is popular now, but what’s going to be popular next? do you think c++ will last forever? unlikely, but either way, the library you use with c++ will almost always be different
java has it’s place in the marketplace, notably servers and mobiles. if games isn’t going to be a dominating area of java, the best thing you can do is adapt to the current language that is dominating - and be prepared to learn the next language soon.
once you have found a job in the games industry, there’s nothing stopping you from showing off to your boss what java -could- do. puppygames would be an impressive portfolio to present, imo.
oh, and a little off the topic, but sony adaptation of java doesn’t mean shit, unless you specialize in DVD menus. they won’t be giving java access to their monster 8-core platform any time soon, if ever. not to mention PS3 sales aren’t exactly above par.
The PS3 sales are low because there was no plan to have them higher. Sony predicted a quarter of Wii’s sales for the PS3, they’re not stupid. As for Java in games development, most teams are likely to have existing coders proficient in c++; so you have to be proficient in c++ to join that team. Long story short you would need a completely new team of Java coders to create a game using purely Java.
I think it would be great if Java was made good use of; it can be done.
[quote]If the business industries managed to switch to Java so easily, why is it so hard for the games industry?
[/quote]
I’d say it was easy for business industries because they were given J2EE. There was and still is nothing equivalent for the games industry with good backing from for example Sun, so java doesn’t seem an obvious choice. C# is already a bit ahead of java in that respect.
I’d say project dark star is a step in the right direction for the server side of multiplayer gaming.
If mainstream desktop Java games happen, it will more likely come from the bottom-up rather than top-down. In other words, successful folks using Visual Studio / DirectX / etc are not likely to switch to Java. The biggest changes for them are the new versions of DirectX or some new physics library or 3D scene graph.
However, folks making Java ME games and/or casual games will continue to grow. Some of the Java developers will eventually become mainstream.
But, I won’t say Java itself necessarily makes game development more productive. A lot of it has to do with the game framework, IDE, tools, build scripts, libraries and the basic philosophy of the developer (design a game to be flexible to changes vs. design a game to a strict spec, etc). Although being able to create both Mac and Windows versions is a huge plus.
I don’t believe that DX is any more “windows compatible” than the techs I mentioned: OpenGL, JINPUT and OpenAL.
And XBox360? Don’t you think that the biggest problem is getting a JVM on the XBox rather than the APIs it uses? That’s a real red herring.
I think that more people using Java would result in better tools for java game development, better industry support.
Getting a JVM on XBox is actually a lot easier than it looks (GCJ, Jet, and the Open Source VM to name but three possibilities). Trouble is no one is going to spend any effort on it if there’s no actual advantage. If you still have to write all your rendering code from scratch because the XBox doesn’t support OpenGL then there’s no good reason to use Java. Therefore, we need a DX binding.
And yes, DX is more Windows compatible than OpenGL. By a factor of about 50%.
Cas
I’m not saying that it would be technically hard to get a JVM running on the XBox 360, I’m saying that it would be politically hard.
As for DX being 50% more Windows compatible than OpenGL (and I presume you mean D3D, not DX here), I don’t understand your point. I’ve coded in DirectX, D3D, OpenGL, and I don’t see any reason to claim that OGL is less compatible with Windows.
How are you quantifying the term “compatible”? I’m not sure we’re talking about the same thing.
OpenGL supports DX10 features in XP, hence being more backwards compatible than DX. If anything, OGL is more “windows compatible” than DX is…strange isnt?
DP
Quite :o)
I’d think an OGL to D3D wrapper for development on XBox would make more sense - then your code can still target Win32, Mac and Linux as well as XBox.
BUT… that still doesn’t get past the fact that M$ will not let Java apps run on their platform. They control access, and that’s one area they won’t let through.
About 35% of Windows computers don’t have OpenGL installed on them, as opposed to 100% with D3D. This figure appears to have remained constant for the last 4 years.
Cas
Which version of D3D?