Last i read the 'Cube API is openGL-like, but by no means just a straight copy. Nintendo were apparently aiming for a friendly interface which developers could get to grips with easily, so its likely its actually simpler than GL in parts.
Does anyone know what XBox Doom3 will be using? Its rumoured that Carmak said that an xbox version would be produced if MS produced an implementation of GL for it. A few months later the xbox version was announced… I can’t see iD putting up with a GL->DX wrapper, but since its an nForce chipset i’d assume that there’d only be a minimal amount of work needed.
This “easiness of porting” from PC to Xbox also shows a problem: a real games console isn’t just a PC but a dedicated games machine. Every component is optimized for its single purpose: to make games look pretty good and fast. It’s more complex to develop games compared to typical PC applications but this depends on your background. If you used to use great computers like the good old 8- and 16-bitters for example with all their nice custom chips, or come from a parallel processing IT sector, etc.
Ok, a Xbox is a PC. However on the same side this is a disadvantage, because it’s bound to PC architecture. We don’t see this very clear with the older PS2 and newer Xbox today (because PC game capabilities made a big step during the last years; although I think it already reached an end). However it’s been visible very well with PS1 compared to past PCs, and it will be very visible again with PS3 and Xbox2. Because PS3 won’t be bound to today’s PC architecture.
I don’t think a “all in one” solution will do the games job in a good way. I think a Symbian cellular phone does do a better job compared to WinCE, same applies to PS2 and Gamecube compared to Xbox.
(Indeed the - few - Gamecube games did impressive me a lot.)
Well, if this community has it’s way, developers won’t be going to a publisher saying they can do XBox and PC, they will say: “We can do 5 platforms including a wireless component for online tasks while away from the primary platform.”
Having said that, a lot of games coming from the PC to a console may not make sense, while a lot of others do make sense. It really depends on the game.
BTW, the whole argument of titles that are specific to a particular platform to take advantage of the console’s capabilities is starting to diminish. Look at the strides that companies like Criterion have made. Show me the real difference between Tony Hawk on multiple platforms, aside from play features, not drastic graphic changes. The same for GTA. Wait until developers get ahold of PS3 and see if they want to continue to go to the metal
-SG
P.S. That reminds me, and Jeff can chime in here, John Carmack stated at last year’s QuakeCon that he sees no reason to go to the metal anymore and will now only program to “standard” API’s. Of course, for id, that means OpenGL with port to DirectX.
The sentence “OpenGL with port to DirectX” I don’t understand, I’m afraid.
Does it apply to Doom3 in the way there will be a OpenGL version of Doom3 and a DirectX version?
Or does it mean that some ID engine games don’t use OpenGL on Xbox but DirectX? (Is there OpenGL on Xbox?)
I don’t agree on the PS3 matter, though. How could any developer exploit the multi processor and render farm features of the PS3 (alleged) in a native way on PS3 and … PC at the same time? Use Java with muli threads? If it didn’t make a difference to use these features then there’s no need for a PS3 … is it?
With regard to the PS3, without saying too much :), it will be a significant task for any developer to code to the metal. Not only do most game developers not have experience writing multi-threaded apps (look at Saturn and PS2 development - and people were calling the PS2 “Saturn Squared”), more and more developers are using high end visualziation tools and becoming used to having the middleware handle the “heavy lifting”.
It will be necessary to have middleware that helps developers exploit the PS3 architecture due to it’s complexity and the fact that a game developer generally does not have enough time to let their top coders spend 6 to 9 months becoming efficient multi-threading coders. Not with a rumored 16 core system…
HOWEVER! If there were some Java technologies, and a kick ass VM, that provided an interface to the architecture that did most of the work for the developer, different story.
-SG
P.S. The PC is a different story, but even here we find that bigger brains like Carmack are tired of writing to the metal. What I meant by the OGL/DX comment in my prior post is that id STILL backs OGL and if they wanted to support another API it would be a port to DX.
For people who make their living shipping software this isn’t a big deal. Having to do a complete rewrite for a console is prohibitively expensive - hence the dawn of NDL, Renderware, etc. and its increasing adoption by studios. Publishers are concerned with making money (and rightly so) and they are not as concerned with the nuances of technology like we are - so for many of them, being able to port an existing application to the XBox is a huge win for them.
[quote]It should be entirely possible and indeed eminently worthwhile to produce an OpenGL driver written in pure Java which calls DirectX to implement its OpenGL functionality.
[/quote]
Wasn’t there one that already did that, called Java3D?
No, SciTech did it though, creating an actual Win32 driver which gave you OpenGL on D3D only cards, and it’s reasonably good, too. It is of course a completely brainless business move on their part because for the price of their driver software you can buy a new video card that’s already got drivers But as a pure Java library with a nice BSD license on it it’s got great merit.
D3D is a level lower than OpenGL; everything in OpenGL can be written in terms of D3D but not the other way round. You won’t be surprised to find out that the code is 95% common in the drivers, too, and that largely speaking, this is why there is no noticeable performance difference between the two.
Essentially, it can and should be done. This is a project which would be incredibly valuable to a lot of people.
I always find DirectX to be more laggy. Quake III and Half-Life ran so smoothly on my comp. But Ghost Recon and Black and White had a lower framerate.
The reason why there code is similar is because Microsoft used OpenGL as a model (as they always do with other open source things) for DirectX. I even read this interview: “The only model we had was OpenGL…”
I agree. However this middleware which has to address the special PS3 architecture, will be hardly useable for a PC-architectured Xbox2 - until it would be extremly generic which means slow again.
[quote]HOWEVER! If there were some Java technologies, and a kick ass VM, that provided an interface to the architecture that did most of the work for the developer, different story.
[/quote]
Yes. Java is different and of course much more than a “middleware”.
[quote]P.S. The PC is a different story, but even here we find that bigger brains like Carmack are tired of writing to the metal. What I meant by the OGL/DX comment in my prior post is that id STILL backs OGL
[/quote]
I see your point.
Some time ago I read some quite positive comments from Carmack about OpenGL 2.0 and his praise of a pre-version of a 3dlabs Ogl 2.0 driver.
Today’s games usually are made by managers with Excel sheets next to team, not my game enthusiasts anymore. This is why today’s games are mainly so dull and boring and always the same - they’ve just to sell well.
This is my impression. However there’s been several interviews in a well known British paper magazine for game developers (a year ago maybe?), which backed up this impression. (Team 17 has been there, Richard Joseph, and many more well known people).
[quote]This is why today’s games are mainly so dull and boring and always the same
[/quote]
Oh, come on now… it’s ALWAYS been like this. But guess what, you don’t remember the dull/boring ones from the days of Atari… why? Because they were dull and boring. You only remember the great games and fool yourself into thinking that was how they all were. I guarantee in another 15 years people are going to be saying the same thing…
“remember how great games were in early 2000??? They suck now.”
That’s not true in my experience. Games are still made by game enthusiasts just like they were before and they tend to have a lot more working for them in the way of money, marketing, and resources. Today’s games are dull because most people just aren’t that creative and end up making dull games. The truly good games that get made STILL have bean counters making sure that the game comes out on time and on budget - its just that there was a better game being made for the money. Nobody wants to put out shit if they can avoid it. Its not like every game developer out there is teeming with creativity that’s being stiffled by the business process. The sad truth is that most developers are just not good game designers. Its an entirely different set of skills. Few game developers make good game designers - and even fewer of those are good artists.
I don’t think so; I’ve read your statement on that topic before but still I don’t agree. I am an oldies but goldies fan, hehe. Of course this is subjective; as always with games and tastes.
Unfortunately I don’t know if that mentioned British paper magazine is on the net (don’t think so). So I could back up my theory with some nice articles.
The main point has been the changement in “who directs the gameplay of today’s games” and that the publishers try not to risk anything anymore.
[quote]But guess what, you don’t remember the dull/boring ones from the days of Atari…
[/quote]
Oh, I remember many of them because I usually played them too. In former times the number of games has been lower so this has been possible.
[quote]“remember how great games were in early 2000??? They suck now.”
[/quote]
Only if you think that the technical aspect of a game makes the gamplay, but I don’t.
I remember in the late 70s/early 80s. Jump Man on the Vic 20. It was revolutionary. The Doom of it’s day. Suddenly, there were about 100 Jump Man clones a year. I still maintain that the mentality of the industry has changed very little over the years, just gotten larger. Graphics have always been important, and technology has always driven the game industry. I’d say the quality to crap ratio is what… 20|80, 10|90? Probably was very similar during the “golden age” (early 80s). Just that instead of 10,000 games released, they had 100.
And THAT is how we come to understand Sun… wait… you are right, this is totally off topic now. Heh.
Edit:
[quote] Only if you think that the technical aspect of a game makes the gamplay, which is usual today, but I don’t think so.
[/quote]
I remember E.T. for the Atari 2600 being slammed for poor graphics (among other things). Why do you think they had screenshots on the back of boxes since the beginning. Graphics have always held great importance.
I don’t think so. In the mentioned British paper mag (with focus being on that topic we discuss) the boss of Team 17 did say that his company (and a few more) are the last ones from the old style. Nearly all the other game companies which have been very sucessfully on 8/16 bit computers and consoles didn’t survive.
He didn’t say this is bad or good, but that the game industry changed totally.
[quote]The sad truth is that most developers are just not good game designers. Its an entirely different set of skills. Few game developers make good game designers - and even fewer of those are good artists.
[/quote]
That’s true.
However from my experience “top designers” (on the E3 two years ago they interviewed many of them) are not always top game designers.
Anway, what we and SUN needs are some very nice Java games. I think that mainly independant game “studios” will go for it at first.
Understand that those of us in the Game Technologies Group are GAMERS FIRST. Jeff’s background is from the game development industry, I am a rabid gamer (27 game systems, 3 PC’s, 1 arcade machine, run large LAN parties), Doug has done some ports of games to Java for the pleasure of doing it, etc.
Understand that us gamers here now have the executive backing of Sun Microsystems and the charter to make this work! There is only so much information you can get by reading press releases and news articles folks. Please trust that we are going to continue to try to do what is best for the community overall.
Lastly, for the few of you that were here since the beginning…you will remember that JavaGaming.org was a little site being run by 2 people on a mission. The only way that we have gotten the organization to where it is today is through pure passion for doing what we believe is best for THIS community. And what is best for this community, we believe, is best for Sun as well.
I believe that we have delivered on most everything we have promised to this community, agreeing that it’s taken longer that intended in some cases ;), but we will continue to fight and push to make Java a viable and important technology for the video games industry.
[quote]we will continue to fight and push to make Java a viable and important technology for the video games industry.
[/quote]
Sounds pretty good. A bright future for Java game developers could be on our way.
You are entitled to an opinion. In this case you are probably wrong :), although your conclusion is quite sensible given what people outside the industry can see.
IME, the actual reason behind dullness has a lot more to do with the fact that the probablility of a game studio + game surviving until it hits the shelves is inversely proportional to how exciting/fun/groundbreaking it is.
Hence, you see lots of dull games just because there are 3000+ games developed each year, but a disproportionately large number of those that are dull make it to release, and the percentage of exciting ones that survive is so low it seems they don’t exist. They certainly do - and probably outnumber the dull ones - but so very few survive.
If you dig deeply enough, you’ll find oodles of different developers pointing out that “getting a game finished is hard. No, REALLY hard.”.
My own experience backs up your latter statements (about good games STILL having bean counters - especially IME at any studio with a high consistency of hit-titles/AAA).
However, I don’t agree about dullness. I meet quite a few people who aren’t that creative in the job - but they tend to be the ones about to leave the industry, struggling to get in, or working at a one-shot-wonder studio that will implode before it makes it to release. Admittedly, I tend to meet a lot of people at shows like GDC, where its unsurprising that you find really creative types, so this is just IME.
But as a pure Java library with a nice BSD license on it it’s got great merit.