I just want to point you to an interesting discussion on the OpenMind mailing list:
http://sourceforge.net/mailarchive/forum.php?thread_id=5100642&forum_id=32309
OpenMind is an “old” (if you can say that) Java game engine and now seems to be restructured or dead.
Crisis of faith anyone? They seem to be going around at the moment. The world is a’changing, I think thats more evident than ever and it doesn’t seem to be going backwards to me…
Kev
There is a lot of talk about how Java is capable of producing high quality games. So far we did’t see any commercial products, and that is the main problem here. Community projects like open mind fail because there is no real proof that they can perform as advertised. They probably could have if only they could reach that stage of development.
Instead of concentrating on developing community general purpose API’s, developers should focus on the games primarily (commercial project would be the best). Something like Megacorps team is doing right now and I hope for their success. Showing off features and proving that they actually work is the key. Take a look at the Unreal 3 engine demo, they show few sequences with newest features and suddenly everyone wants to get a hold of that technology, I am pretty sure that the demo like that can be pulled with Java, it should deliver 5-10 fps less according to Sun and some other 3D Java experts. It’s all nice when we read how Java can deliver 85 % of C++, That’s fine, but where is the game or demo that delivers all that performace and features not to mention platform independance. Personally I think that no one would ever care about those 15-20% of performance lost if there was a feature rich game that flawlessly work on all claimed platforms.
So, until there is at least one extremely commercial game (that utilizes all the high end hardware features) produced with Java, Open Mind and similar projects will suffer and only hard core enthusiasts will contribute to them while majority of people will sit and wait for Java to prove itself. Off course there is a question, who will risk the resources to pull an “Unreal 3 class” project with Java.
Hi VeeJay!
I think you would take a look at http://chromethegame.com
That’s a commersial game written in java, and it’s pretty cool. 8)
Hi Don,
All this would be very nice if following was not the case, Here is what Chrome team said about the technology behind the game.
For more details you can refer to their discussion forum at :
http://forum.chromethegame.com/en/forumShTh.php?SeID=0&ref=&SeID=4&ThID=7937
I would like to see this kind of game based entirely on Java though
Ok.
I didn’t know chrome just used java as script language :)…
But in most cases Java is up to 200 % faster than c++
Hi,
:o
I would strongly like to believe that, any references, micro/macro benchmarks that prove the above? :-/
[quote]Hi Don,
All this would be very nice if following was not the case, Here is what Chrome team said about the technology behind the game.
For more details you can refer to their discussion forum at :
http://forum.chromethegame.com/en/forumShTh.php?SeID=0&ref=&SeID=4&ThID=7937
I would like to see this kind of game based entirely on Java though
[/quote]
It depends on the interpretation of “scripting engine”. If I understand it right, in Chrome just the 3d- and sound-engine is Cpp. The complete game logic (and more) is Java. I wouldn’t call this scripting engine. If others do, it’s just important to agree to mean the same.
The latest Chrome patch replaced nearly 2000 Java class files, which contained many (all?) parts of the AI, plus game logic, network modules, sprite and textur and HUD and sound (file) management, a nice console system, and a lot of more. All being done in Java.
For me that’s the main part of the game and I don’t think it’s important whether the framework around the logic is being done in Cpp which calls the logic via JVM, or the other way round. (In both cases it’s so called “dirty Java”.)
Since all these many Java classes run inside the JVM at runtime of the game I don’t think it’s valid to call it “scripting”. First because todays JVMs don’t just interpret Java byte code anymore (compared to Java’s beginning with real interpretation), they use JIT/hotspots and produce real compiled code. Chrome uses a J2SE 1.4 JVM. Second because Java really isn’t just another typical scripting language, it’s a real language (compared to dedicated scripting languages).
Well, so much for my investigation on Chrome. Corrections and additions welcome.
[quote]There is a lot of talk about how Java is capable of producing high quality games. So far we did’t see any commercial products, and that is the main problem here.
[/quote]
Um, problem for whom? When we did survivor on Xith, we couldn’t give a monkey’s what “commercial products” were using it, we had more intelligent demands like:
Sure, commercial successes do no harm, and help sway the consumers - which is great IF you need consumer awareness (but in mosts cases you probably don’t). They can also help with ignorant / scuttlebutt idiots like some of those you may meet when trying to sell to a publisher (there are always some employees who suck, but they could block you from getting to the negotiation table)
What you will probably find is that those with cash or resources who cry “I won’t do anything until I see a big commercial success” are one of two types:
They actually mean something rather different, for instance “I don’t trust this technology because it has too many bugs and no dedicated support team with a track record to comfort me that there are no showstopper bugs. I don’t care about a successful game, but what I really want to see is that the damn thing actually works - I don’t trust the authors as far as I could throw them.” But the statement about commerical success is a shortcut to them having to do their own due diligence on this issue. Doesn’t mean the game is necessary!
Ignorant fools or people who just like to open their big mouth. Often they know their own ignorance, so they don’t dare say something is good (because they don’t understand it) unless they have something to cover their butt, and when your big commercial success comes out they amend their words to say “Ah, but that’s only getting 100 fps where Doom3 just came out with better graphics and 120 fps. I’m not going to blah blah until there’s a game with D3’s graphics and at least 120 fps”.
IMHO.
By the original definitions of scripting languages, it’s fair (although still arguable…) to say that Chrome uses java for scripting. However, most people today (especially in games dev, where it’s being subtly re-defined) are ignorant on what “scripting” traditionally means and have vague hand-waving ideas about what it “sort of means”, which often come down to:
“scripting language = non-compiled interpreted language”
so that it’s a dangerous term to use unless you know the audience really understand it, and won’t publish it in a magazine or something, thus spreading misconceptions.
Also, the quote said that java is “functional” which is another phrase they should be more careful with ;D…
[quote]I just want to point you to an interesting discussion on the OpenMind mailing list:
http://sourceforge.net/mailarchive/forum.php?thread_id=5100642&forum_id=32309
[/quote]
Frankly, the OM thread appears to be a classic situation of some hobbyists doing some free-time work that they somehow imagine will be competitive with high-cost commercial tools and middleware. SF.net and the OS movement have made more people believe this is possible and set out on the path, but it doesn’t change the fact that back-room inventors don’t tend to beat large corporations at their own game unless the product at hand can reasonably be developed supported maintained and sold (even if the sales price is “0” it still needs to be sold…) by just a handful of part-time people.
So…the OM situation seems just what I expected, really.
Although it was a bit depressing to see Jacob Marner quoted, it’s good to see the old chestnut of “WHY THE HECK DOESN’T JAVA RUN ON CONSOLES?!?!?!!? SUN WTF ARE YOU DOING???” come up again…;D 
Just a few links to show what Java can do better than c & c++:
http://ei.cs.vt.edu/~wwwbtb/book/chap21/javalag.html
http://www.vinci.org/rlv/d/javaperf/docs/wentworth/spw98.html
(You will need PowerPoint for this one)
And a lot of more… 8)
Hi,
I found this one much more argumentative, and I think it pretty much reflects the reality and it focuses on Java game development.
http://www.rolemaker.dk/articles/evaljava/Evaluating%20Java%20for%20Game%20Development.pdf
My point in this discussion is not to prove that Java sucks or that it can’t perform, my point is that Java is capable of delivering high quality games with small performance penalty but huge production speed increase. There just has to be more effort in showing and proving this point to other people so that Java can make bigger momentum in gamming industry. I think community should focus more on developing games and demos than on trying to deliver projects like Open Mind.
I’ve changed my stance on the whole thing now. I don’t evangelise nearly as much any more because I don’t want any competitors 
Cas
[quote]I’ve changed my stance on the whole thing now. I don’t evangelise nearly as much any more because I don’t want any competitors
Cas
[/quote]
You know, I had begun to wonder 
The latest reply to this topic on the OpenMind mailing list is in a new thread: http://sourceforge.net/mailarchive/forum.php?thread_id=5107197&forum_id=32309
Some time ago I had the idea to centralize Java gaming API efforts a little (which was theoretically good, but practically impossible). In my opinion the Java gaming world is too small to have dozens of APIs. Every new developer should seriously consider to contribute to an existing API (Jogl, LWJGL, Xith3D, jME, Odejava …) instead of starting a new project. Otherwise you will probably end up spending more time for coding your API than your game (or you really have a lot of time and money) and the existing APIs will improve slower, because of small user bases.
Nah, we’re all doing fine, and the plethora of APIs keeps everyone happy.
Cas