Now, Sun’s GTG should, I think, have an official public response to this prepared in short order.
What is Sun providing to compete with this fusion of major cross-platform technologies?
Cas 
Damn, pipped to the post.
Just define J2ME, Linux and Solaris to be the major game platforms and there is absolutely no problem
[quote]Just define J2ME, Linux and Solaris to be the major game platforms and there is absolutely no problem
[/quote]
Add to this Apple (*) and Windows, and there we are.
(*) For independent game developers who use Java, the Apple platform could be a great market chance. It’s not as huge as the Win32 market is, but much more indy friendly. At least this is what I read on Garage-Games forums etc. And Cas also says similar things. It’s a market many Win32 only indies won’t ever reach. PLUS you don’t need XNA or such. 
The two pieces missing from the puzzle are OpenGL and Playstation. All bases must be covered to provide a coherent response to Microsoft’s strategy. Unless there’s a geniune console competitor to PS2/XBox that runs Java and OpenGL that I haven’t heard announced at GDC yet.
Cas
From the looks of the article it seems like they regard ‘cross-platform’ as ‘both PC and XBox’. Doesn’t sound that innovating tbh.
And WinCE.
It’s a pretty bloody large market and a major boon to cross-Windows-platform developers. It will no doubt integrate nicely with the Zone too.
Cas
Sun needs a major pimped out partner in the form of Sony or Nintendo and after talking to both of them last week - neither (at the developer relations) level see any reason to pursue that opportunity. Nintendo just being a little more blunt about it
Good for them I say!
Now We have a cross platform product that hits Windows, windows hand held, and windows console.
Questions?
Q: Uh, yeah, aren’t all those windows platforms?
A: Of course!
Q: But, uh, where does the cross-platform functionality come in?
A: Ah, yes, our cross platform hist windows CE, Windows Xbox, and Windows Windows!
Q: Uh…what about MAC?
A: What?
Q: Apple, Macintosh, you know, the IPOD, the IMAC, the G4, etc…?
A: What?
Q: What about Linux?
A: What?
Personally, I think windows is a great thing. So crucify me. I play most of my games on windows. I develop on windows. Windows, windows windows.
Hey…java runs on windows.
Hey, what about the last cross platform programming tool MS announced? Wasn’t that C#? Why didn’t they use that? Could it be because it isn’t truly cross platform? Bingo! Is this new tech any more cross platform? Nope!
Geez folks, you are getting paranoid.
Java is so massively pervasive in the broader-than-gaming business market that it will always be here.
The big gaming companies will eventually move to Java because it is easy to use, and programmers are more expensive than hardware. Cut your staff in half, use ten percent more machine, and it means you have c/c++/assembly speed, at 3/4 the price. In addition, you are truly cross-platform, with linux, mac, etc. Tap 3-4 markets with one product.
The calls to “defend our java dreams” get repetitive here. They aren’t necessary. Java is not small.
Java is HUGE. it is pervasive. It is adaptable. It has the best !@#$ API doco on the web. It has several free IDE’s available.
It is the ONLY true programming language that is non-scripting( like php, perl) that is a www language.
it is the ONLY true cross platfrom language, bar none. Even PERL has system specifics that you need to be careful about when programming diff platforms.
Why are you worried?
I am not at all worried about the future of Java - I’m just concerned with the future of games development technologies.
Cas
Part of the worry I suppose is that (like .NET) MS are actually doing the “right thing” ™ and releasing their specifications for these new technologies.
Like .NET, you’ll see a linux version of XNA pop up.
As much as I love PC games, the console market is far more lucrative. I don’t quite get what MS are hoping to do here tho. Can you see Nintendo or Sony (or Company X) thinking to themself, I know, I’ll jump in and use that.
So, IMO, I think Cas is right. Sun could release a gaming toolkit, based on top of Java, sell it as a properly open standard and get buy in from Nintendo and Sony. Didn’t you suggest this “Games Toolkit” idea before?
Kev
[quote]Questions?
(…)Q: But, uh, where does the cross-platform functionality come in?
A: Ah, yes, our cross platform hist windows CE, Windows Xbox, and Windows Windows!
Q: Uh…what about MAC?
A: What?
(…)
Q: What about Linux?
A: What?
(…)
It is the ONLY true programming language that is non-scripting( like php, perl) that is a www language.
(…)
it is the ONLY true cross platfrom language, bar none. Even PERL has system specifics that you need to be careful about when programming diff platforms.
(…)
[/quote]
Very nice article, and funny, too.
I think Sun already tried the game toolkit idea, called JSR134, but it was wide of the mark. I think before, I suggested Sun put their full might behind OpenGL and OpenAL and get those technologies into the core Java specification, but they’re still soldiering on with J2SE and AWT.
I think what Sun really needs is J2ME+LWJGL in a profile. It’s almost Good To Go - we just need OpenGL ES support in the library. Then I think Adam’s Grexengine might provide some of the networky bits.
I find it an interesting difference in mode of operation with Sun and Microsoft: M$ sees a great technology that some pipsqueak startup makes, and instead of trying to copy it, they just slap £10m down and buy them, lock stock and barrel, and rebrand it as M$. For example, Sun need some whacky multiplayery Javified game servery technology, so they punt £10m over to blahblahblah and buy his company off him. This is how DirectX came into existence.
But Sun do suffer from a terrible, terrible “not invented here” syndrome. How much bought-in technology is there in the Java product range?
Cas
Ok, here we go…
Cas, you bring up some very valid points, but you bring them up within the context of a VERY narrow view of the Java world. It would be great is we could just run off, do what ever the hell we wanted to Java, called it whatever we like and be doen with it. God KNOWS it would be so much easier than having to answer to a board, build proper TCKs, document the APIs, and position the technology as a proper Java technology. but we have to. Java is not controlled by Sun, despite what a few of you here may think.
Sure, from your point of view, LWJGL and J2ME would be a good “games platform” solution. But that is no suprise as LWJGL was started by you, so you suffer from the same NIH that you claim Sun suffers from. Why? Because you believe you have a better solution. i have no problem with that, BTW. With regard to JOGL or lwjgl, if both were submitted to the jcp, which would pass as being compliant? You want a complete games solution that carries the Java brand? We want that too.
About the server prototype. If you would care to read the press about the server prototype, you would have noticed that it wasn’t Sun technology only, but incorporated GameSpy and TimesTen as well. Why? Because their technologies are very good and so we used them. Doesn’t sound like an NIH problem to me.
The reason we developed the prototype is because we don’t believe that there is another soluiton out there in the games industry that compares to the COMPLETE server stack we demonstrated. Vertical AND horizontal scaling, 100% failover between servers and multiple games within the same stack. I don’t think any other server technology we have seen does what we built. Oh, and the RESOUNDING agreement of major game companies who want to understand more about what we did seems to echo my belief.
And the comment about “networky bits” implies that you don’t think that network technologies are difficult and I expect to hear that from one who does not build network technologies. If it was that easy, don’t you think Sun, or another network technology company, would have done it? Networking is hard. Networking to scale is harder. Check out utility compute models and think about how an EA or Blizzard may benefit from this.
On to XNA. So here is the official word on this from me (and not Sun ok?): it’s a total hack and a desperate move on MS’s part. They feel the PC market shrinking for them and continually try to branch out to other devices.
Didn’t it seem a bit “thrown together” to you? WOW portability between XBox and PC. Developers do that porting now! Please… Tools are important, but need to address mote than just one platform. Every developer we talked to couldn’t understand what the hell MS was trying to pull off there.
One last thought. I can’t understand why everyone is so scared of MS. Name one market outside of Windows that MS dominates. ONE! And no, Office is not a market, it is hand in hand with Windows. Servers? No. PDAs? No. Mobile? No. Set-top-box? No. Game consoles? No. Web games? No.
Look, Windows has roughly 85% of the desktop market. Curent Java technology is shipping on 70% of all NEW Windows desktop systems. It also ships with Linux and OSX. Java technology has shipped in over 250 MILLION handsets in just 4 years! By comparison, Nintendo has shipped just over 150 million units of GameBoy branded devices in 16 years. Java technology is shipping in set top boxes, available for both PalmOS and PocketPC, in servers of all makes, and as far as game consoles are concerned, we are working on it
-ChrisM
P.S. About JSR-134. it was not a games tool kit. It was, basically, DirectX for Java. Please read before commenting.
Want to see how much outside technology we bring in? Check out this link:
http://search.sun.com/Search/press?qt=acquire&nh=10&qp=%2Bcollection%3Apress&col=press&survey=0&rf=1
- ChrisM
First of all, you are correct that MS does indeed just purchase a whole lot of stuff. And how much of it ever makes it to market and stays or is successful? Very little. Hrmm. let’s see, specifically with XBox, RARE acquision = “Grabbed by the Ghoolies” HUGE seller for them eh? Ah yes, OddWorld. Where is Abe now? MS spends a huge amount of $$$ on investment. Not necessarily to use what they buy, but to crush it.
Secondly, if we thought there was a technology that did what we did, don’t you think we would have looked at acquisition? Oh yeah, as a publically traded company we are not allow to leaglly TELL you what we look at for acquisition, are we? There is a reason we have an investments group as Sun you know
Lastly, DirectX did not come into existance that way. talk to Alex St. John and see what he says Or, you can just read this link:
http://www.seattle24x7.com/people/alexstjohn.htm
-ChrisM
I do like to poke you with the odd controversial comment on the offchance you’re feeling eloquent and dying to tell us stuff See, we we 3 posts in a row and over a page of insights!
[quote]Curent Java technology is shipping on 70% of all NEW Windows desktop systems
[/quote]
Aha! It’s gone up since last we spoke That’s good.
Cas
[quote]I do like to poke you with the odd controversial comment on the offchance you’re feeling eloquent and dying to tell us stuff See, we we 3 posts in a row and over a page of insights!
Aha! It’s gone up since last we spoke That’s good.
Cas
[/quote]
Ah! So we LIKE poking the bear in the eye huh?
Cas, you do make for some great discussion GGGRRrrrrrr…
- ChrisM
[quote]Ok, here we go…
[/quote]
Thanks for the comments, Chris. Hope Cas didn’t poke too hard :).
Both are low-level techs; there are plenty of mid-level and high-level techs in this area (including some owned by Sun). Between them GS and TT are so very far from providing a complete game solution as to be unexciting for many people - which is not to say that they aren’t exciting for many other people - but I suspect that most of the excitement you saw was because you were SUN: One of the few major IT companies with the word “network” in their name. Any move by Sun into this area where you show something concrete will generate excitement - even when the “something concrete” is itself too small to be exciting on it’s own.
That’s disingenuous; I can name several that fulfil all those criteria (I’m sure you can too). They aren’t even the important features anyway. Failover and multiple-game stacks are something you could do with any decent mid- or high-end middleware (even J2EE).
You can have it one of two ways: either you’re doing NIH in the full knowledge that there are other far more advanced systems (most of them too far ahead of you for you to catch up without a huge budget). Or you’re treading the same path as several big names before you, and just don’t really understand what the needs of online game development actually are, so that you think what you’re offering is somehow really exciting in and of itself (i.e. without the Sun brand behind it). I think you’re all very smart and that you know a lot about the industry, so I’d guess you just want to do it yourselves.
FYI the reason we aren’t demoing anything at GDC is partly because these bits in themselves are so easy and so valueless to developers, and partly because as a small company we can’t handle excessive interest (unlike you guys, who can handle it). I’ve actually been out there shifting licenses, and have a good idea what the sticking points are. A lot of the interest we had back in 2001 I’m guessing was similar to what you’re getting now - it’s from people who will seem really excited, even try to give you money for something you tell them isn’t for sale yet, up to a particular point and then they find you’re missing something that’s a deal-breaker for them and they’ll wander away. We knew this from the start, and resisted serious pressure from magazines etc to have a big launch.
(Aside from the obvious - that I agree entirely, having been dealing with a similar level of interest for years, and it’s only been getting bigger :)) I’m sure they’re delighted that a second “big-iron” company is getting into this market. Your presence re-validates IBM’s investment from 2 years ago (just what they need if shareholders question their involvement).
Sorry to be pedantic, but networking isn’t particularly hard. If you want me to prove it, get me 20 million users and I’ll build you the networking parts of a system to handle it. On my own (assuming I’m allowed to buy the appropriate off-the-shelf hardware!). The problem is the application processing; the nearest you get to a networking problem is usually unacceptably high application-induced latencies, or app-induced fatal non-determinism (almost prevented TSO from shipping!).
E.g. everything GS does is not particularly difficult when compared to developing an online game. IMHO GS has a fine suite of software for games developers. The sad truth is that - even with all their services on the side - they deal with less than 10% of the problems of developing large scale online games. I’m guessing the other 90% (which companies like Grex have good coverage of - and some even have complete coverage of) is the stuff that Cas was thinking of.
Integrate:
[] easy-to-use OGL binding
[] scenegraph API
[] grexengine architecture
[] client-profiles for all clients, not just mobile devices
…and you have something akin to a MS product: a cradle-to-grave solution that is extremely easy to develop with, and yet almost infinitely powerful.
NB JOGL is far too difficult to use; this is not a criticism of JOGL, but if you want to offer solutions rather than a library with “minimal conversion cost from C”, then you need to do the kind of simplification and rationalization that LWJGL has done. One of our USP’s is that we have invested a huge amount of work in doing this for the server side - and it really shows. You can develop new games very quickly (NB: so much so that we’ve been tempted to stop licensing our tech and just write new games; for some games we can shave more than 60% off development time, and that saving keeps increasing with each successive build)
EDIT: when I said “minimal conversion cost” I meant for games developers who’ve previously used OGL in C. I.e. JOGL minimises the amount of re-learning they need to do; the method calls they’re used to are all the same, etc. Whereas something like LWJGL does a lot of rationalization and merging and redundancy elimination.
Incidentally, since mid February we’ve been getting cold-called by VC firms. The way things are going we could be about to start turning them down, just so we can get on with growing our sales rather than be distracted and lured into endless meetings and presentations. How things have changed since the “VC drought” of 2001/2002!
Anyone considering VC money, I suggest now would be the time to start looking.
Was this a rhetorical question? I’d be interested to know the actual answer… Certainly, I recall that ORO waged a (bitter) public fight with Sun for years trying to get regular expressions into the JDK. This was back before the reforms of the JCP, when IIRC it was much harder (?) for anyone but giant firms to contribute anything non-trivial to java.