Since we are talking about competing against AAA games, I was thinking more of games distributed on DVDs. Where java games will look the same as any other.
You know how I feel about webstart, so won’t comment further.
For a customer a java game have no advantage over a c game. Best option is to hide it is a java game and let it be judged on a equal basis, and not what language it was coded in.
Java’s only advantages worth speaking of to a customer are… well, there are none. Webstart is pretty dire as a customer experience, and even I don’t like many things about it.
All the advantages are for developers. In theory it should be reduced bugs, reduced time to market, reduced money - but the reality is working around stupid JDK bugs, being unable to distribute your games easily or at all, and tearing your hair out frequently at the unbending nerve of Sun to make your life difficult trying to earn a paltry crust.
(OT example: so I get Molebox working finally to get round stupid JRE size distribution problem leading to legal trouble. Then what happens? Won’t work with Armadillo (Software Passport) any more - big trouble getting my games onto the portals. I mean, I’m just fed up to the fecking back teeth of all the obstacles Java has put in my way when these are the precise obstacles (WORA anyone?) that should no longer be an issue).
Cas 
[quote]paying for content = fixing the symptoms
making your platform more interesting to developers = fixing the cause
It’s better to stop walking on broken glass than it is to take pain medicine to stop the pain of doing so.
Teach a man to fish, and all that.
[/quote]
Its not a matter of tech or capabilities its a matter of mentality. Are you going to fix the publishers mentality with that ?
Even if you got the biggest thing publishers wouldn’t pick it up if it was a Java game simply because they don’t need to take risks. But with Sun supports backing up good games the risk would simply disapear, just like magic, in the head of other publisher.
The option is to get your game to a less prophessional (sometimes desperate) publisher that will pick up anything and this is very dangerous, just look at Troika.
The only way to use Java in games is like they did with Vampire Masquerade as a scripting language. That way we can fool the publishors into thinking its a C++ engine and they can swallow Java as a scripting language. After all if they swallow python and lua why wouldn’t they ?
There is no problem whatsoever with using Java for a retail game. Publishers don’t give 2 hoots whether it’s written in Blitz Basic, Algol, Java, C or backwards in octal machine code. All they care about is that it does what it says on the tin by the date that they want it done by.
Cas
Maybe we are not speaking about the same publishers but im refering to games comparable to HL2 or Oblivion. Know any recent game that was in the list of a known game news site that used Biltz basic or Pascal ? How many games can you mention that succeded at the level of, im not going to say HL2, but let me see gamespy right now: for instance Swat 4 or Battlefield 2, and didn’t used c++/assembler mix.
If you believe that publisher risk management does not influence the choice of Java then we are contradicting ourselfs here. Developers make tests to see what tech is best for their needs. If its true that Java is so good and superior than C++ then why don’t we excellent Java games ? How much time are we going to wait ? There are enough ogl bindings to provide the bridge between Java and ogl. Results and proof of concept. Where are they ?
[quote]Java’s only advantages worth speaking of to a customer are… well, there are none. Webstart is pretty dire as a customer experience, and even I don’t like many things about it.
All the advantages are for developers.
[/quote]
Note: seeing as the conversation is about how to go forwards, and how Sun (or someone else) could/should get java into games development, I’m assuming basic things like putting some effort into fixing glaring simplistic bugs would be done in-passing.
That in mind, i.e. don’t just come back with “hey, but it has THIS bug”, you guys are seriously missing the point here: users DO have advantages (multiple) using java and they do, in fact, understand and appreciate them.
Am I the only one here who uses the MS Action Pack subs to get ultra cheap MS software for entire companies at once?
If not … have you never wondered why MS sells approx £20,000 of software to almost anyone for under $100, with FREE updates? Have you never wondered why there’s more brochures in there than CDs? (and FYI for those that haven’t seen it, there’s approx 100 CDs of full version MS software).
Have you not read the brochures?fYI: the brochures each start with: “Someone in my company says we should use linux instead. What do I say to them?” or “We’ve been told we’re not allowed to buy Office because it’s too expensive. How can I persuade my boss?” etc.
Wake up and smell the roses: real software companies that make serious money do so by making consumer “education” a priority: leaving aside training and certification entirely, just looking at the “convert our potenial users (they’re only potential: they just tried to evaluate the softwre and haven’t actually trued to buy it yet) into cheerleaders, salespeople, and hard-to-persuade MS lovers - exactly the kind of people that have an answer to every reason why the tech dept is trying to get them onto linux, and make it so hard that often they win out through sheer bloody mindedness”. This works. It is nothing special that MS does (it’s just that many peopel here ought to know this example), although MS has some of the best-of-breed marketing groups in the world and are a great example to follow on marketing.
And bear in mind that this literature is aimed precisely at the computer illiterate: hardcore techies are too hard to persuade, non-techies are easy.
Likewise, java in the games industry would get a huge boost through simple user education. This whole thread started around the premise of the power of Carmack to influence: no serious techie gives a **** what Carmack thinks because we know how little he knows about the things he talks about, we’ve followed him over the years and seen his naivety, ignorance, and poor general knowledge of development techniques and tools (sure he’s great at some thigns, we know that too)
…but the masses he’s leading are not us, but the players. So why then dismiss player education of why they benefit from webstart? And, in more general terms, why the hell are you developing in java if there’s no benefit to the players? Benefits to the developer must necessarily cause benefits to the players, logically. although of course you may have to work hard to work out what the actual end benefits to them are.
phew. . Sleep time, I think…
Firstly, the OGL bindings are all literally in alpha or beta stage, they aren’t at all mature and as part of any risk management process you’d discount them out of hand for not being at 1.0.
Secondly, the existing talent and even code bases are all in C++ and will be for another 5 years or so. When you’re doing resource management you look at the availability of the skills and I can count the number of Java developers with the skills and experience to pull of a AAA title on the fingers one one hand. So you’d discount it completely on that.
Thirdly, if you’re a Win32 developer you might be looking to get your game onto the Xbox in which case you don’t care for OpenGL or Java anyway as you’ll actually be making life considerably harder for yourself thanks to the incredible failure of SGI lobbying to get the API on Xbox and Sun lobbying to get Java on the Xbox.
All hearsay mind you.
Cas
[quote] Secondly, the existing talent and even code bases are all in C++ and will be for another 5 years or so. When you’re doing resource management you look at the availability of the skills and I can count the number of Java developers with the skills and experience to pull of a AAA title on the fingers one one hand. So you’d discount it completely on that.
[/quote]
That resource management is also reviewed by publishers.
So this declaration is a contradiction of your previous statements about publishers not giving a “woot”.
And it is in this statement that you ARE correct. Publishers do care what you use.
And on a personal publisher note, if the guy you are going to go pitch to has it in his head Java can’t do games, your Java game isn’t going to change his mind UNTIL it is done and ready to be delivered (but he will not pay for that development). So you’re going to have to do it on your own, case and point Puzzle Pirates and maybe even your work Cas…until the idea that Java can do games it universal
I’d make a pretty clear distinction between the “publisher” and “developer” here. The publisher really doesn’t care what technology goes into it, only that it works. It’s the developer that makes the decisions about the development teams, tools, and technology.
Cas
[quote]Maybe we are not speaking about the same publishers but im refering to games comparable to HL2 or Oblivion. Know any recent game that was in the list of a known game news site that used Biltz basic or Pascal ? How many games can you mention that succeded at the level of, im not going to say HL2, but let me see gamespy right now: for instance Swat 4 or Battlefield 2, and didn’t used c++/assembler mix.
[/quote]
http://www.puzzlepirates.com/
Java Indy game, now being published by UbiSoft…nuff said.
Publishers only use two factors when it comes to games: “How many copies can we sell? How much do we have to invest?” They don’t care about technology as long as it works. There will be more games in the comming months that will be at or approaching the level of technical sophistication/production values/marketability. This gets closer every day.
I am going to push my stuff as close to that level as I can. I know what it takes to make a game at the AAA level. I have the skills and artistic ability to do it. For me, the only question is do I have the time and patience to go that far by myself. At some point during this project I may start bringing in other game developers that I know from all over the world from working on mods and maps for FPS’s. Several of the people on mod teams that I have been involved have since gone professional (on games such as Jedi Knight II, Doom 3 and Quake 4), and I have done some contract work for Activision also. I am sure there are other people with a similar history/contacts experimenting with Java.
And of course, there are examples such as Megacorps Online by Agenency9 who are even closer to being able to produce something worthy of a top publisher. They themselves, or a dev group could pick up their tools/api’s and bring something to a publisher.
My point is, all the seeds are there. It has happened already with Puzzle Pirates. (You could consider Chrome Java as well, as it is mostly Java, with a binding to C++ for the low level engine).
Publishers care if they are paying for the development, they don’t care if they are buying. Even if they are paying for dev, if you can show you have the team and skills it can be negotiated. However, initially, for Java to have a full scale published AAA , it will probably need to be funded by a team (self or via investment), then brought to a publisher. Though, that may not be true now that there is the UbiSoft example, just harder.
[quote]I’d make a pretty clear distinction between the “publisher” and “developer” here. The publisher really doesn’t care what technology goes into it, only that it works.
[/quote]
After reading some articles at gamasutra and reading a lot of ex-devs interviews and portmortens, in particular the info about selling your product, i do believe that developers care about what technology you use and every other detail. In fact risk management is so complex these days that it takes months for them to aprove a game for publishing (read some of the interviews from Troika ex-devs). They will consult all sorts of advisors and specialists to evaluate among all the offers they have, what are those that have more chances of giving them their profit. In their scale they will account for things like gameplay, art quality, appeal to mainstream, linearity, compexity and of course technology. They don’t care if the game is in Java or not if they will allways put a C++ game in front of a Java game. How much the other factors weight on their decision may get them to guive priority to a Java game but don’t put much fait on it. There are many PC games being made but much more are refused and never get to be published.
[quote]It’s the developer that makes the decisions about the development teams, tools, and technology.
Cas
[/quote]
The developer is the game studio in the case of games. Don’t forget that developers also want to sell their games. There are two aproaches to sell a game to a publisher: the good one and the bad one. ;D WARNING: SPEECH AHEAD
The bad one we go talk to the publisher and be naively honest with the guy. Hey dude i have this amazing game made in Java. We know that Java is not that much used in games but let me explain why Java is a superior technology and will cut down development time. Two mistakes: the guy doesn’t even know what a programming variable is and won’t trust or is interesting in earing any tech explanation, he has their own trustable advisors for that; second mistake you told him that Java is not that much used in games and, in fact, youhave just told him not to buy the game. The game is now at the bottom end of his list of games from which he will pick one or two and throw away a douzen of offers.
The not so bad aproach for Java. We make a proof of concept alpha demo. We focus on selling the game and telling what he wants to ear. But we won’t make the game based on what he thinks is best. We just soften it so that he recognizes selling characteristics in the game. We make the game we think its fun. The guy doesn’t trust you and is not an expert in making games or gameplay. He just wants to see how you aproach him. His advisors will evaluate the game for him. The game has a C++ engine and uses Java for scripts. Hey dude Java beats Python right ? In fact Java beats C++ but you wont tell him that. The game may even be 1% C++ and 99% Java with the bulk of the game being all made in Java and the basic engine funcionality in C++.
The good aproach. It allways works be Java or not. If you know some people who really have some talent for art/writing/gameplay and are willing to work for fun or a share of the game, make 50%-70% of the game or the entire game. Show the publisher the game almost complete and polished (no bugs or broken things, remenber he is going to examine it). Follow the selling techniques in the last paragraph and wait for his answer. Its almost impossible you won’t get someone to publish the game for you but most likely it will be one of those desperate publishers who don’t do any quality testing at all on the products they sell so you will have to make sure the game is very polished with no show-stopers or visible bugs.
The best aproach or the Carmack style. Do what you would have done in the previous paragraph but don’t go see a publisher. Instead make the entire game, offer 1/4 of the game for free to all the games magazines and create a site for the game. If the game is real good and you have a number of site visitors big enougn, you have a very big chance of being a second apariton of Carmack on the game scene.
I made a univ work, last year, on selling tech (in particular games) and what i told you here is what i have researched on gamasutra and their games ring and many other interviews and post-mortens available on the net. So feel free to reserach it yourself.
Yet another take on the blog from the one of the top dawgs of the GTG
http://blogs.sun.com/roller/page/dtwilleager/20050328#doom_on_cell_phones
Well, there we go. Doesn’t really say much that we didn’t already ken.
Cas
From the blog:
Well maybe he is exageration a little on the stability of opengl bindings but all the rest is there.
The LWJGL stuff is pretty much production ready We’re now really in beta I think… time to stop calling it alpha.
OpenAL the most commonly used cross platform sound library!? Methinks it’s FMOD actually…
Cas
[quote]Paying for content = kickstarting your platform
Atari, Sega, Nintendo comes to mind
And where are Atari and Sega now
Conclusion based on the data presented: 2/3 of the companies that try to “kickstart their platform” fail.
[/quote]
Nonsense. :
The reasons why Atari and Sega failed has absolutely nothing to do with them having payed for content. Do you think we would even remember Atari if they didn’t make sure there was enough interesting content?
[quote]paying for content = fixing the symptoms
making your platform more interesting to developers = fixing the cause
[/quote]
You really think that the developers’ interest decides the fate of a platform?!
Remember, the Atari 2600 was not a pleasant machine to develop on with only 128 bytes of memory and nightmare-ishly quirky video hardware; they certainly didn’t please developers.
No, developers worked on it because either Atari paid them, or because they knew the money was to be made from the 2600 otherwise, because the 2600 was hot.
I don’t think any professional dev-team starts development on a platform, just because they the platform. They do it because they believe they will make money.
Exactly my point.
Sun needs to make people believe they can make money by making games in java.
They can do that by either fixing java so it’s viable to make games in it (spending a lot of money on integrating proper opengl support into the core java distributions would be a good first step. No, jogl isn’t good enough), or by giving every single developer money for chosing java.
If they have to pay the developers, the developers will go away as soon as sun stops paying, and sun will never make money off java games.
Considering that Sun has always claimed to be “just a hardware company” it still beggars belief that they’ve not gone the way of the XBox and commissioned a piece of proprietry hardware to run Java games on.
Cas
So now we’re back at the old problem that java doesn’t run on consoles
I believe Sun has already done some effort for achieving this (PS2), but it didn’t work out for technical reasons.
I would say that if at one point they would succeed at having java running on at least one console, it would be a clever idea support development for at least one good title as a proof of concept.
I have no idea how Sun would benefit from this though, money wise. Java not being a games console and all…