XBLA 50MB limit inspires innovation
by Richard Mitchell 
{ Dec 14th 2006 at 4:30PM }
What the geek raised an interesting question yesterday: is the 50MB size limit for Xbox Live Arcade games a bad thing? The first and obvious answer is yes. After all, more space equates to bigger, better games, right? Conventionally, the idea makes sense. Look deeper though, and the question isn't quite so cut and dry. Yes, developers can do more with more space, but constraining their space forces them to find ways around it. Look at a game like RoboBlitz. It runs on the Unreal Engine and squeezes in under 50MB. What the Geek cites procedural generation as the next big thing in game design, and we have to agree. Honestly, it's something we've put some thought into recently.
Procedural generation is a kind of programming that uses sets of instructions rather than pre-designed content. The example that we were given by the RoboBlitz team was the Mona Lisa. You can compress a JPEG of the Mona Lisa and it will take up considerable space. On the other hand, you can give a computer instructions -- which take up hardly any space -- on how to draw the Mona Lisa. The catch of course, is making sure that the computer can draw well, which is where good programming becomes important.
It's possible that procedural generation could actually disrupt the development process as we know it -- where "disrupt" is used in a positive way. If Naked Sky can squeeze RoboBlitz into a tiny package, why can't a large developer do the same thing with a retail game? Not only could it save space, but it could save money and even development time. Meanwhile, companies like Sony are opting for a higher capacity format, Blu-ray, claiming that it's necessary for new gen gaming. It's only natural to think that more space is needed to drive innovation -- after all, when CD-ROMs were introduced, more space was the innovation -- but what if the opposite is true? This brings to mind the old adage, "work smarter, not harder." Huge amounts of space could ultimately lead to lazier programming, whereas limited space -- be it the 50MB limit of XBLA or the 9GB limit of a DVD -- could actually drive programming innovation. We could even jump right over the edge and say that games will actually get smaller, not bigger.
It's a lofty subject -- perhaps too lofty for us to tackle properly -- but it certainly fires up the imagination. What do you think, is standardizing procedural generation the next step in game design, or is it something else?
[Via Digg]
Reader Comments (Page 1 of 1)
J.Goodwin @ Dec 14th 2006 4:38PM
The 360 is uniquely designed to be an extremely capable "artist" in this sense. The main processor's three general purpose cores and shared access memory, as well as (IN PARTICULAR) the shared cache on the CPU that can be directly accessed by the video processor are all perfectly designed to assemble content on the fly. Not just textures, but models, variations in models, variations in textures, animation, etc.
Basically it breaks down to this: You can draw every frame of animation, or you can draw the key frames and outsource the rest to Korea. You can create 1,000 variations of a texture to represent battle damage, or you can let the computer do it for you.
If you do it all by hand, you have to store it somewhere, and then you have to find places to keep it in memory. If you do it the 360 way, you create it when you're ready to use it, and then you throw it away when it's not useful.
rickyyo @ Dec 14th 2006 4:52PM
Yeah I heard this idea back whe kkregier was being made and that game was the game that started it all I think if im not mistaken.... I really have had the opinion you guys have for a long time...
notpokey @ Dec 14th 2006 4:48PM
I'm happy to see the idea of 'generated' content being put to use in real games. My first experience with it was a demo called 'the .product' that released in 2000, for The Party demo competition.
It's a PC animation, with 3D video and sound and everything, lasting a good 15 minutes. And it's an absolutely mindblowing 64kb.
And as a glimpse of where the tech that drives RoboBlitz came from, it's still an interesting watch.
Link to details: http://www.pouet.net/prod.php?which=1221
edhe @ Dec 14th 2006 6:28PM
This is the next step for content imho, like cpus have gone from fast to fast & parallel content will go from detailed to detailed and dynamic without taking up more space.
Can only hope
the 50mb limit does drive innovation, it's a great idea.
Shamma Banks Jr. @ Dec 14th 2006 10:33PM
It may be good for innovation, but in the long run, I still believe that the limit should be raised to 100mb, when bigger hard drives come out.
But what really pisses me off is I know in the future I will have to buy another larger hard drive for my 360 for over $100.
Caffeinate @ Dec 14th 2006 6:53PM
I know this is a fanboy site, but you're all letting your fanboy beer goggles cloud yout vision.
Will this lead to better compression technology? In all likelyhood. Is procedurally generated content the wave of the future? Almost undoubtably. But trying to call what is really a restriction an advantage is very, VERY misguided. A more open approach from MS (encouraging smaller sizes, incentives for fitting on a memory card, etc) would have been less Draconian and would likely have allowed for some games to come to XBLA that we will not currently be able to see.
Don't get me wrong, I love the 360 and have no intention of purchasing another current-gen (and yes, they're all the current gen now) console for some time, but the old adage still holds; the ends don't always justify the means.
Objectivism works. Trust me.
killr0y @ Dec 14th 2006 7:04PM
This goes back to the days of the C64, where demoscene hackers would out-do each other by cramming as much music and hardware tricks into their 64k demo as they could.
Out of all that, these coders utilized every bit of every CPU cycle to produce graphics & sound that vastly exceeded the dreams of even the very hardware designers that built the chips.
What happened? 16MB RAM & 650MB CD-ROMS, that's what happened! When game musicians no longer needed to use the sound chip to synthesize instruments to make sound because they could just dump REAL music to a .wav file to play in the game, designers started getting lazy and filled up megs upon megs of space just for music tracks.
Back in the day, people actually drew the graphics & sprites PIXEL by PIXEL. You wouldn't believe how much work was put in to hand drawing ever last dot on the screen (granted there weren't nearly as many dots on the screen back then!). Utilizing tiny color pallets (16 colors to choose from on the C64), these guys hand-dithered the colors to produce the illusion of more colors than available...
Of course, if people designed games the same way that they did back then, it would take 10 years to produce a game (and oh what a game it would be!).
Still, the raw horsepower of todays consoles can be put to better use IF the tools are there to make it easier. I believe this is what XNA is the begining of. Right now XNA gets people's feet in the door of game programming. As XNA matures over the next few years, and adds more tools to its reportoir, we should start seeing a resemblance of the golden era of game coding.
killr0y @ Dec 14th 2006 7:16PM
Caffinate, try to not be an ass. What Richard is saying makes total sense if you actually tried thinking about it. If there is no reason for game designers to try unorthodox techniques, then its less likely for innovation to thrive.
Developers aren't going to willingly spend the time to come up with new algorhythms for procedural texturing, music, etc. Or come up with new approaches to game engine design, or even gameplay mechanics if they don't have to.
Putting a constraint on a process is the quickest way to get someone to start thinking outside the box. Giving someone too big a sandbox tends to lead to squandering those resources. Just look at the urban sprawl of most US towns & cities compared to their much neater European counterparts. Why is that? Because Europe is forced to utilize their limited space more efficiently. I'm not getting into a debate on Europe vs the US, there are pros and cons for both situations.. I'm just using this as an analogy of how forcing constraints, whether real or imagned, can lead to some very novel and surprising things.
Maybe if you tried a little unconventional thinking yourself, you'd realize this.
Pat Bateman @ Dec 14th 2006 7:20PM
I would so be more supportive of the 50mb limit, but it can have detrimental effects on these games. I would much rather have a larger game than be forced to deal with the ridiculous loading times in Assault Heroes. Anyone else willing to give up more HD space for better loading times?
tacklebox @ Dec 14th 2006 7:32PM
There is no question that we would have access to better games if the cap was at least 500mb. While it will become increasingly amazing what they do with 50mb of space, its not serving the gamer's best interest.
Caffeinate @ Dec 14th 2006 7:54PM
Hey, no need to get defensive.
To make Godwin's Law a reality here, my analogy will be Nazi-related. Let's say some wonderful technology was invented in a concentration camp, changing the world forever (cure for cancer, for example). Would this justify the Holocaust? Certainly not.
Now I understand that this is an extreme example and in no way, shape or form do I think Microsoft is the Nazi party. I LOVE my 360. But IMHO it is short-sighted to believe that ANY limitation on creativity is actually good for it.
As I said, it may very well improve technology, but again, ends and means.
Sorry if I touched a nerve, I'm just trying to bring up an opposing view.
What The Geek @ Dec 14th 2006 8:17PM
First, thanks to everyone for taking an interest in my article, and of course, thanks to Richard for bringing it to everyone's attention.
Second, I want to address the point made by Caffeinate, and contested by killr0y. You're both right. Currently the 50 mb size limit is causing devs to cut corners in addition to using better compression techniques. It has negative effects on the games that are out right now. In Major Nelson's most recent podcast the guys from Naked Sky admitted that they did have to remove some things from Roboblitz. So yes, at the moment the size limit is detrimental in some ways. In the future, however, devs will work toward fitting their XBLA game into the 50 mb limit WITHOUT cutting any corners, or removing any gameplay elements. Ultimately we'll see some very positive effects of the size limit, even though right now it seems to be handicapping developers.
Bear in mind that my article wasn't as much about what's happening on the XBLA right now as it was about what will happen in the future.
Jeramy @ Dec 14th 2006 10:58PM
i sorta agree with caffeinate
really, this is just justifying a restriction. granted, good things will come out of it, and i am a strong believer in the article's merits, but all in all, there will still be some drawbacks. hell, i'm sure they'd put halo on xbla if there was no restriction. an incentive program would be optimal, but i'm not sure how one could adequately implement it. i'm not a programmer though, so idk what i'm looking for.
but, all in all, i agree that the size limit will help things and benefit many. show those lousy ps3 fans why they wasted money.
perhaps ms could determine if a game was able to take up such a little space, and have tiered size levels. that's the only win-win i can think of. idk, screw it.
J.Goodwin @ Dec 14th 2006 11:09PM
Does the existence of Israel justify the Holocaust?
There are certainly many Zionists who act as if it does. Or vice-versa.
Some things you have to live with, and the results can be good or bad, depending on your perspective.
Carl @ Dec 14th 2006 11:11PM
I think the missing point here is that these are supposed to be arcade games. Not full out story-lined games. I've always looked at the 50mb limit as a restraint from making a game something I can jump into in 3 seconds and have a good time without having to press A over and over for storyline to continue. An arcade is designed for games that you throw a quarter in and try for the highest score. Simplicity is key.
So it inspires innovation in another way.. the way games used to be. Being different and making a fun unique game.
Boff @ Dec 15th 2006 6:17AM
I agree with several comments here. Yes, the 50meg limit causes some detrimental restrictions - but on the other hand I wholeheartedly agree with it also inspiring innovation.
Killr0y, you hit the nail right on the head for me. I remember the C64 and Amiga ShareWare/FreeWare/PublicDomain days and being blown away by what some of the coders were capable of. Then along came HDDs and CDROMs which also brought the old argument of BloatWare vs Innovation.
Also #15 has a good point - the whole point of XBLA is the 'pick-up-and-play' side of things isn't it? Surely if the limit was higher or removed altogether then the games WOULD get more complex and involved. I'm fairly certain most game designers are going to want to cram as much in as possible to make it that much better (in their opinion) than the competition?
Just my thoughts anyway!
Burnt Meatloaf @ Dec 15th 2006 6:13AM
The caps are only for bad developers. Screw them.
Innovation and creativity come from the desire to do different things. Talented people are naturally terse and efficient. Putting restrictions on people will generally challenge people who would otherwise prioritize incorrectly, but in the end, they are still lacking talent.
It's not about compression techniques. It's about not making bloat in the first place. Really good developers will probably make 10-15MB games, just because they know they don't need to waste space. If you tell people not to use more than 50MB, they will probably find ways to make them 50MB, rather than asking themselves if they really need 50MB of space in the first place.
Experienced developers (people that have made at least 1 Live title), should be offered more leeway. People should have their limits set by their talent. Until you prove yourself, you're not allowed to waste people's time and money. If you're good, you can make your own rules and let MS review your proposals.
*killr0y: "This goes back to the days of the C64, where demoscene hackers would out-do each other by cramming as much music and hardware tricks into their 64k demo as they could."
Very bad example. Keep in mind that small demos are about code, even if they rely on procedural sysnthesis. The 50MB limit is about graphics and sound. We don't write self-modifying code anymore or pull resources from specific versions of ROM.
64K demos were still focused on raw technology, took FOREVER to code, and were usually fraught will bugs and glitches that usually barfed on hardware released a year or two later. Playability is a different beast entirely, and that requires good, structured, maintainable programming practices that are not always compact, even on game consoles.
Brian G @ Dec 15th 2006 9:34AM
It only took 10 posts for this topic to bring up Nazi's. Wow.
Innovation comes when pre-set conditions are issued and developers have to work to meet them.
Yes it's true that the 50mB limit is a limitation which is limiting what a developer can do, but if the limit were 100, 200, 300, 400, or even 500, people would still have to make concessions in order to build within the constraints.
A 50MB limit means that you have small games which people can have a little fun with, which don't take a lot of resources to download, and which don't require a ton of developmental resources (all of which justify the relatively low price points).
Let's remember that Xbox Live Arcade games are not meant to be full fledged gaming experiences, they're little things you can pop into or out of. They're the whipped cream on your sundae, not the sundae itself.
Johnny @ Dec 15th 2006 10:50AM
normal mapping.
dont overlook that one.
it came out of last gen, we wanted more detailed models, but we coudlnt afford the polygons.
so instead we stored the information from the hi-rez version of the model in a damn texture that can be applied to a lo-rez model and make it it look near exactly the same.
today we like to call that Gears of War.
XLA developer @ Dec 16th 2006 9:49PM
It looks like I'm a little late to comment but . . .
I make downloadable console games and I think the 50m restriction is reasonable for now.
The size of a game can effect the end user the following areas:
Download times
Load Times
Storage Space
Game Performance (Frame rate and smooth game play)
Without limits, developers might not work hard to get the size down. There are many things that can reduce the size of a game without the end user even noticing a difference. I would suggest that many developers might not take the extra time to weed out the unnecessary size if a limit was not imposed. The end user pays because they wait longer to download, they wait longer for the game to load and they have less room on their hard dive for other games. I don’t care how big your hard drive is or how patient you are for a download, if you could get the same thing (or nearly the same thing) smaller and faster wouldn’t that be better? If your games are 10X as big your hard drive needs to be 10X as big to hold the same number of games.
Is Blast Factor a Better game than Geometry Wars or Mutant Storm BECAUSE it is so much bigger? Could it have been half the size without really effecting the quality of the game?
Obviously the 50M size limit does limit the end result and there are some cases that a better game could be made with more space but, at the moment, the budget sizes of these games are a much bigger restriction. You don’t get downloadable games that are Halo quality not because of the size limit but more because of the budget limit. What you get is creative games and not just bad halo clones. This is a great example of budget restrictions resulting in innovation and great $5-$15 games.
I believe that the future IS in downloadable games and eventually you will be downloading AAA titles but even then size will still be an issue (even if the limit is 2G instead of 50M). Techniques we have learned along the way will help those games be better. Limits will cause innovation.
This article never stated that the 50M limit was always a benefit. It just said that sometimes it might be good. I guarantee that this is the case.