Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

OnLive's Epic Plan For a New Type of Video Game

timothy posted about a year and a half ago | from the it's-everywhere dept.

Cloud 137

An anonymous reader writes "OnLive's had a tough twelve months any way you look at it, but as a new profile of the cloud game streaming service points out, throughout it all, service never dropped, and the number of platforms it's on keeps growing. Up next is the tiny Ouya console, but in a wide-ranging interview, OnLive's general manager talks up plans to bring MMOs to the service, and even a whole new type of video game, one that will run on many servers, not just one PC: 'Look at how CGI has changed cinema over the last few years — you can do CGI essentially realtime. It could completely change what a video game looks like. That leads us to new technologies. Then game designers say, "What could I really do with a computing platform that is so powerful but also available across so many devices?" You're no longer constrained by computing power — that has tremendous opportunity.'"

cancel ×

137 comments

Sorry! There are no comments related to the filter you selected.

Not constrained (2)

mog007 (677810) | about a year and a half ago | (#42786517)

Yeah, OnLive isn't constrained by computing power, but they're still constrained by bandwidth.

Is there a big enough market for their service in the few areas that are able to use their service?

Re:Not constrained (5, Insightful)

h4rr4r (612664) | about a year and a half ago | (#42786617)

Forget constrained by bandwidth, the real problem is latency. Unless they can put a data center in every city they plan to service they can basically forget about it.

Re:Not constrained (4, Insightful)

heson (915298) | about a year and a half ago | (#42787063)

Correct, and they aim for the wrong type of latency demanding games. To survive they must go for games that can handle bad latency. OnLive has a huge potential but only if they stop selling an impossible product and start going for achievable goals.

Guess: They have sold a lie to investors and are stuck in it.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42787631)

Right so everytime this comes up we all comment about the latentcy issue, which is good because it pretty well kills the idea that this sevice is going bring to market any real time games that won't suck.

Now this GM is saying they are going to bring new types of games to market, going to leverage the processing power they have for CGI and to then push to a handheld console?

Hi Rez animated chess/MTG/Zork anyone?

Re:Not constrained (2)

Luckyo (1726890) | about a year and a half ago | (#42788087)

Turn based-anything comes to mind. Turn based strategy for example.

But that's pretty much it. Anything real time is dead on that latency.

Re:Not constrained (2)

Dins (2538550) | about a year and a half ago | (#42788265)

An MMO could handle the latency issue better than most other types of games, but it would still suck. Back when I used to play MMOs, my ping was routinely 200ms. It only really got annoying when it was above 500ms. But then the rendering was all done locally and only my position and commands were sent to and from the server, so the game seemed somewhat responsive even when latency was going on behind the scenes.

Absolutely forget any sort of FPS.

Re:Not constrained (1)

Luckyo (1726890) | about a year and a half ago | (#42788701)

One of the main reasons why WoW is so popular is that its so damn responsive due to very permissive ruleset on what client can do mostly independent of server. Things like moving around for example.

Having latency on your every motion is utterly horrible in MMO. It goes against one of the core reasons why WoW is so successful.

Re:Not constrained (1)

halltk1983 (855209) | about a year and a half ago | (#42788371)

If everyone is affected by the latency, then it's a bit more of a level playing field. Like the days of dial-up quake team fortress. When I started there were almost no LPBs on broadband. So everyone had a 300ms ping time. It was difficult. It was challenging. It was sometimes frustrating. Then the first LPBs started really rolling in with their 30-60 ms times, and they really got a benefit. But the point here is that if the latency is large, but even, then it doesn't provide a benefit for some.

Re:Not constrained (2)

Luckyo (1726890) | about a year and a half ago | (#42788725)

There is a reason why people don't play football while wearing 10kg shoes on each foot. By your argument, that doesn't matter as long as all parties wear 10kg shoes.

Has it ever occurred to you that benefit has NOTHING to do with enjoyment if no one is enjoying it due to retarded lack of responsiveness due to latency?

Re:Not constrained (1)

exomondo (1725132) | about a year and a half ago | (#42790299)

But the point here is that if the latency is large, but even, then it doesn't provide a benefit for some.

But the gaming experience sucks for everyone then, otherwise you would get gaming servers introducing artificial delays to make sure every player had pretty much the same latency.

Re:Not constrained (1)

mlts (1038732) | about a year and a half ago | (#42788805)

What OnLive might do is consider making commodity boxes packed full of GPU power that are made to sit on LANs. This way, game commands for rendering are coming from a server at most a hop away.

Plus, people would more than pay for central render/streaming server than have to upgrade each PC's/devices graphics card each time a new Crysis hit the stores.

Re:Not constrained (1)

LordLimecat (1103839) | about a year and a half ago | (#42790973)

The problem is that your keyboard / mouse latency to the OnLive center, and the graphical latency coming back, add up; on a residential line you are looking at ~30ms each minimum, and likely more; this means that each of your input commands is 30-60ms behind what you're seeing, which is itself 20-50ms behind what is actually happening on the game server.

Re:Not constrained (1)

hairyfeet (841228) | about a year and a half ago | (#42787403)

I doubt even that would help here in the states as we have seen the ISPs get nastier every year when it comes to caps and throttling. Hell in my area I usually end up paying more in overage charges during the Steam sales than I do for the games themselves, so who is gonna want to use their game service if they end up paying two to three times more than just buying a retail copy?

Lets face it as long as the ISPs care more about bumping their stock prices and giving their CEOs mega-bonuses than actually laying lines shit like this just ain't gonna fly. We saw the same thing during the dotbomb where so many pushed the "thin client" model only to find out bandwidth costs ended up eating any savings they had from running fat clients, OnLive will find the same as users buy their service only to cancel the next month when they get a $200 bandwidth bill from little Johnny playing OnLive.

Hell we saw the ISPs trip over themselves to join the 6 strikes bandwagon because it gave them an excuse to toss anybody that uses even half of what they promise, you honestly think they are gonna go for a service like OnLive?

Re:Not constrained (1)

crazyjj (2598719) | about a year and a half ago | (#42788605)

as long as the ISPs care more about bumping their stock prices and giving their CEOs mega-bonuses than actually laying lines shit like this just ain't gonna fly.

Does laying lines of coke on hookers' backs count? Because with all that money, they're definitely doing that.

Re:Not constrained (1)

LordLimecat (1103839) | about a year and a half ago | (#42791001)

Latency (forget bandwidth, its irrelevant here) has been steadily decreasing over the years. I rather imagine that getting business-class latencies for residential customers would be cost-prohibitive.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42788157)

I miss the days when game companies didn't run servers, that gamers ran their own. Does GameSpy even still exist?

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42789175)

Get informed, please [superuser.com] . The latency on the internet to another continent can easily be significantly less than the latency from moving your mouse until pixels on your screen change as a result of that. You won't want to do the world championship Quake tournament with OnLive, that's true. For most games and most gamers and most internet connections, there is no fundamental obstacle to something like OnLive.

Re:Not constrained (1)

LordLimecat (1103839) | about a year and a half ago | (#42791143)

A transatlantic connection likely is going through one hop. Your measurement completely ignores the fact that MOST latency is caused by routing delays in urban areas, with speed-of-light delays only becoming significant on long hauls.

All of that aside, you completely mis-interpret what that link is saying. The latency for a single pixel to change on an LCD monitor is generally, worst-case, going to be sub-15ms, and usually sub-10ms. The speed of light in fiber crossing the atlantic on its SHORTEST span would take ~13ms, but I challenge you to fire up a BGP looking glass and actually find nodes which get data across the atlantic that fast-- particularly since I dont believe any links take that "shortest path". Its significant that looking at straight "speed of light" usually gets you nowhere close to the actual network latency once you add in all the router delays.

Even ignoring that stupid comparison, every additional bit of latency adds up. When you inject a full 50+ms of latency (which will be VERY common for connections to game servers)-- and not just net latency, but input latency-- you can make a game very frustrating to play. Gamers are used to pressing a button and seeing their character respond immediately, with some network delay for characters around that; with onlive they will press a button, and have a delay (latency x 2, for round trip) before they see their character jump, plus additional delay (latency to onlive + latency from onlive to game server) before they see everyone else move.

I suggest you familiarize yourself with latency, what causes it, and how it effects gaming before you start spouting off again.

Re:Not constrained (1)

harryk (17509) | about a year and a half ago | (#42791103)

I don't know if you're a current gamer/customer of OnLive's, but let me tell you they've done a fantastic job at solving latency issues. In fact, the only time it's noticeable is for driving games (unplayable quite honestly).

For other twitch type games, first person shooters Onlive works surprinsgly well. I've played through Red Faction, Home Front, and a few others and they all play very well. Some of the slower games, Patrician for example, play very well on OnLive.

I won't sit here and tell you that latency is of no importance, it most definitely is. And true they need more data centers to be more responsive. But as a gamer in the midwest (Milwaukee), I can tell you OnLive's game service is quite responsive and a decent alternative to constant workstation upgrades.

Re:Not constrained (5, Insightful)

wolfhead (919963) | about a year and a half ago | (#42786623)

I live in a major city and have a pretty fast connection, I tried OnLive a bit last year and felt the video stream was still way too compressed. Why have real-time rendering in a game if the stream of it is going to be filled with artifacts and a capped frame rate?

Re:Not constrained (1)

sandytaru (1158959) | about a year and a half ago | (#42787695)

Especially in games like MMORPGs, where the players are pretty picky about the graphics of the game since they'll be staring at it for hours on end. It doesn't have to be super realistic (WoW and FFXI are still going strong) but it does have to look pretty and be stylish all the same. I don't know if OnLive can deliver either the crisp realism or good art direction to draw in a true MMO audience. At best, we'll probably see F2P social games. Something like Gaia or Maple Story might work.

Re:Not constrained (1)

Dunge (922521) | about a year and a half ago | (#42788307)

MMORPGs players are picky about graphics? Everyone I know put the graphics at minimum to prevent loading lag.

Re:Not constrained (1)

halltk1983 (855209) | about a year and a half ago | (#42788485)

Farmville, minecraft, animal crossing, roller coaster tycoon, the sims, sim city, chess and other board games, and alpha centauri and other turn based games would all work well. And there is a growing class of causal gamers that play those style games.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42789575)

I would take minecraft off that list, people routinely bitch about multiplayer servers that are lagy. The fighting in the game on hard difficulty actually requires twitch reflexes. Lag kills minecraft dead in the water.

Re:Not constrained (4, Funny)

robthebloke (1308483) | about a year and a half ago | (#42786625)

But if you're not constrained by computing power, you could do all of your 3D rendering using a real time ray-tracer written in Java Script!

Re:Not constrained (3, Insightful)

SpeedBump0619 (324581) | about a year and a half ago | (#42786827)

Don't think so small. You could write a real-time ray-tracer using LOGO and turtle graphics. Talk about geek cred!

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42787461)

Don't stop there. If you already are going to LOGO, you might as well go to a vector path language which has fast (and even hardware) implementations: PostScript!

Imagine real-time ray-tracing on a PostScript printer! You wouldn't be bandwidth or compute constrained so much as constrained by how quickly you could cut down the trees!

Re:Not constrained (1)

X0563511 (793323) | about a year and a half ago | (#42787755)

There's a peak where you start losing points instead. This is well into that part of the function.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42788281)

Did you mean PEEK and POKE?

Re:Not constrained (1)

X0563511 (793323) | about a year and a half ago | (#42789589)

No. Peak, as in local maxima.

Geek points is Y, X is the complicated-geeky-contrived solution factor. Points go up as X does, to a point, but then it swings down in a negative slope and does not return.

(also: probably this is a woosh)

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42789749)

I think that was supposed to be a joke about old programming code.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42787837)

Or you could build the raytracer out of reservoirs and sluicegates in Dwarf Fortress.

Re: realtime raytracer (0)

Anonymous Coward | about a year and a half ago | (#42787959)

A realtime raytracer written in JS would be better, one would guess, than the realtime raytracer that you yourself have written, or indeed than any of the raytracers written by the commenters on this thread, by virtue of it being written at all in anything; had you or they ever written a realtime raytracer, you could be excused for thinking that the fact that the others did not reflects poorly on them. As it does.

Re:Not constrained (2)

gl4ss (559668) | about a year and a half ago | (#42786649)

if he can't make the finances work by renting one pc for a guy.. how the fuck is he planning on making the finances work by renting some guy five pc's worth of hw??

where it would have an unique angle would be on massive real time action, with all pc's meshed on a fast connection between each other up and no chance of cheating.

nobody fucking cares about his pr shilling. get some meat on the story. "You can do CGI essentially realtime." NO SHIT SHERLOCK - I'm doing CGI realtime right now on this pc in background! sure, it's some shitty old gold box rpg game but realtime cgi none the less.

Re:Not constrained (3, Insightful)

vlm (69642) | about a year and a half ago | (#42786901)

if he can't make the finances work by renting one pc for a guy.

This is the best question about their whole business model. The stereotypical gamer is supposed to be a lumbering herd animal, right? Everyone plays the same game at the same time together online? So you can't make money off over subscription. So instead of the user directly financing a gaming PC, they'll intermediate themselves in between that transaction by providing .... Um...

I can't see the health club model working either, where you get people to sign up for new years resolution and then never see them again.

So when you strip away the tech angle, what is their business model exactly?

Re:Not constrained (1)

Luckyo (1726890) | about a year and a half ago | (#42788149)

I suspect it was "selling ultimate DRM to publishers".

Then publishers noticed what kind of turd it was and pulled the financing.

Re:Not constrained (2)

CastrTroy (595695) | about a year and a half ago | (#42786929)

Yeah, I never really thought about it like that, but it's probably one of the reasons it won't work that well. Most people willing to pay monthly to play games probably will want to play multiple hours a day. Which means that they'll be usually a large portion of the machine's resources. Very few people are going to play a monthly fee and then only play games for a couple hours a week.

Re:Not constrained (1)

gstoddart (321705) | about a year and a half ago | (#42787077)

I think it would almost have to be ...

1) Make streaming video game service
2) ???
3) Profit

For me, a cloud based gaming service is a non-starter. I play video games only infrequently, and have recently disconnected my XBox from my network because I was starting to see ads in games.

Combine that with paying for your network bandwidth, and I can see a lot of people deciding they're not interested.

Re:Not constrained (1)

poetmatt (793785) | about a year and a half ago | (#42787117)

exactly. the compromises for all the network bandwidth costs are enormous. I don't think people realize that if you really wanted a quality onlive-equivalent experience it's probably going to be $50+ a month, which could easily translate to a high end computer every 2 years.

Re:Not constrained (1)

crazyjj (2598719) | about a year and a half ago | (#42788667)

I knew things were bad when a dragon in Skyrim asked me to bring him a cool, refreshing Coca Cola.

SecondLife with Post Crysis graphics?? (1)

RobertLTux (260313) | about a year and a half ago | (#42787211)

of course that would post likely bring half the Internet to its knees but hey a guy can dream.

(challenge for an animator create a tummy rub animation that can be used with the WereHouse Dire Wolf)

Re:Not constrained (1)

cream wobbly (1102689) | about a year and a half ago | (#42787903)

You can do CGI essentially realtime.

NO SHIT SHERLOCK - I'm doing CGI realtime right now on this pc in background! sure, it's some shitty old gold box rpg game but realtime cgi none the less.

It's at times like these the depressing conclusion dawns that the (decidedly) non-technical press will redefine CGI to mean Cinematic-quality Photorealistic Computer-generated Image[ry|s].

Well they can keep it. CG was good enough for me as a two-letter abbreviation, and it'll never sound right to add that redundant "I" on the end.

Re:Not constrained (1)

CastrTroy (595695) | about a year and a half ago | (#42786673)

Not just bandwidth, but ping time. Ping times are a big enough problem when the controls show the results in real time on your screen but it gets updated a little later on the screen of the guy you're playing against. But with the way OnLive works, your controls have to to the server and then back to you before the image updates on your screen. Certain types of high-twitch games will completely not work in this environment. But the kinds of games that won't work on this environment are exactly the type of game where fast reflexes matter. It doesn't matter how long it takes the graphics to render for chess, but then again the graphics for chess aren't that complicated (or don't need to be). However, a game like a fast paced first person shooter or driving game needs the graphics rendered right away, but these are the games where the graphics are usually more complicated.

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42787017)

That's why I block ICMPs. My ping times go to infinity!

Re:Not constrained (0)

Anonymous Coward | about a year and a half ago | (#42786939)

Who says they aren't constrained by computing power? Larger companies have issues with scaling up for demand on their MMO servers when they are busy as is and they probably have way more computing power at their disposal. Adding in the additional stress of having to render everything for all users AT THE SAME TIME in realtime would take enormous amounts of power beyond what OnLive likely has.

Re:Not constrained (1)

GameboyRMH (1153867) | about a year and a half ago | (#42787087)

Don't forget network latency...lag will become far more furstrating, and affect your single-player games as well!

Re:Not constrained (1)

poetmatt (793785) | about a year and a half ago | (#42787099)

actually they're constrained by relevance - mostly a lack of.

Please remind me again (1)

Dishwasha (125561) | about a year and a half ago | (#42786537)

Why would I want a cloud game streaming service?

Re:Please remind me again (0)

Anonymous Coward | about a year and a half ago | (#42786557)

Its like Aarons Rent To Own, or really any of those.

If you're too poor to afford a good console and all the games, then pay us a monthly fee for it!

Re:Please remind me again (0)

Anonymous Coward | about a year and a half ago | (#42786627)

Also, its ultimately portable. As long as you have a 8mb+ connection.

No need to pull around consoles everywhere you go, just plug your laptop in and play. Although I would be curious if you could successfully play a competitive FPS like quake 3 on it. (not that you ever would since it runs on just about any device now-a-days).

Re:Please remind me again (1)

vlm (69642) | about a year and a half ago | (#42786923)

As long as you have a 8mb+ connection

And in America or other 3rd world locales, that means you're easily wealthy enough to afford multiple consoles and laptops...

So you can't afford a $200 one time xbox but can afford $150/month to the cable company, plus your access device which probably costs more than the xbox anyway. Hmm.

Re:Please remind me again (0)

Anonymous Coward | about a year and a half ago | (#42787873)

I get 16 Mbps for 35 bucks a month (after taxes, etc). 22 Mbps for the first few MB of a connection when bandwidth allows. I live in Hickville, AR. Well Bentonville-Rogers-Fayetteville, but it's all the same.

Re:Please remind me again (1)

Mister_Stoopid (1222674) | about a year and a half ago | (#42788305)

Congratulations on winning the internet access lottery. For 95% of Americans, anything better than 5Mbps is only available as a top-of-the-line business connection and is priced accordingly.

Re:Please remind me again (0)

Anonymous Coward | about a year and a half ago | (#42788921)

You do realize your 95% thing is complete and utter horeshit right?

I don't mean to call you out as a liar, but you are a fucking liar.

Re:Please remind me again (1)

LordLimecat (1103839) | about a year and a half ago | (#42791291)

http://www.akamai.com/dl/akamai/q3_2012_soti_infographic.pdf [akamai.com]
http://www.akamai.com/stateoftheinternet/ [akamai.com]

Average connection speed in the US is 7.2mbps. The absolute top is South Korea, with 14; I believe we're somewhere around 10th place, but you have to knock one or two places off since theyre counting Hong Kong as a country, which it really isnt.

Re:Please remind me again (4, Interesting)

Adrian Lopez (2615) | about a year and a half ago | (#42786999)

No need to pull around consoles everywhere you go, just plug your laptop in and play.

Or you could just, you know, get the PC version of the game and install it on your laptop?

Re:Please remind me again (0)

Anonymous Coward | about a year and a half ago | (#42788493)

No need to pull around consoles everywhere you go, just plug your laptop in and play.

Or you could just, you know, get the PC version of the game and install it on your laptop?

Majority of gamers still choose consoles in spite of your claim that laptop gaming is the way to go.

*jeopardy theme*

Why does PC gaming suck?

Re:Please remind me again (1)

Anonymous Coward | about a year and a half ago | (#42786783)

Because it's got cloud in it... that's where it's at, mate... And by a stroke of luck, rainy days, when there are more clouds, is when people game the most...

Re:Please remind me again (1)

Adrian Lopez (2615) | about a year and a half ago | (#42786973)

Why would I want a cloud game streaming service?

Precisely. We have such powerful GPUs inside our consoles and computers, but they want us to have things rendered remotely along with the added latency? I'm sure it's great for developers who don't want customers illegally uploading copies of their games, but how exactly does it benefit consumers?

What a pointless waste of good bandwidth.

This sounds cool. (0)

Anonymous Coward | about a year and a half ago | (#42786543)

But very hard to pull off. Kind of sounds like an idealistic future of computing paradigm theory that I hear bandied about from time to time.

The GAMING SINGULARITY... tada... bump, psh...

Re:This sounds cool. (0)

Anonymous Coward | about a year and a half ago | (#42786759)

You spelled BA DUM TSS wrong.

Re:This sounds cool. (0)

Anonymous Coward | about a year and a half ago | (#42789799)

Thanks, I really need to go to special sound effect spelling school. Either way I really appreciate this lesson as I struggled for at least 1 minute and thirty seconds trying to figure out how to spell out the BA DUM TSS sound phonetically. As you can see I still got it horribly wrong.

History (4, Funny)

WilyCoder (736280) | about a year and a half ago | (#42786545)

That's funny, I read the title as "OnLive's Epic Plan For a New Way To Screw Its Employees"

I had forgotten about them. (2)

medv4380 (1604309) | about a year and a half ago | (#42786631)

This is defiantly going to be Epic. Probably not in the way they are thinking though. We're in the middle of the Next Great Video game crash, and all we're missing is an Epic Fail like ET. Someone, unexpected, needs to roll snake eyes already.

Re:I had forgotten about them. (1)

hairyfeet (841228) | about a year and a half ago | (#42788019)

I think the next ET will come from EA, you look at how much money they are shoveling to their triple A titles like Dead Space 3 and they are quickly getting to the point that they will have to sell more games than there are players of a specific genre just to break even. My guess is EA is gonna crank out some $150 million plus triple A title and have it so spread out, trying to cover every possible demographic for "mass appeal" that it'll appeal to nobody and go down in a giant ball of flames, just as Atari bet the farm on ET and had it cripple the company when it went tits up.

Re:I had forgotten about them. (1)

medv4380 (1604309) | about a year and a half ago | (#42788749)

If I were to bet I'd have two picks. First, if Windows 8 and Office 365 panic investors, Ballmer could be removed, and his replacement may have a mandate to remove any losing divisions and focus on getting Windows and Office back in line. That is unlikely, but if it were to occur MS backing out of the Console market rapidly could trigger a crash. Activision/Blizzard is my number 2 pick. They already have a bunch of Short sellers circling them. Too much revenue is bundled into WoW and CoD, and the moment ether fails will be a big day for the vultures. It's already headed down, and some estimates having it at 6 by 2014, but that's a slow decline. If something happens too rapidly and investors get spooked that correction can happen in an instant. So many things can go bad now I'm just waiting for who goes first.

Latency (0)

Anonymous Coward | about a year and a half ago | (#42786647)

Who cares if a game looks like a CGI movie if the latency is over a few hundred milliseconds. These guys consistently miss the entire point. This kind of platform will always be useless for first-person shooters and any kind of action simulator like space and racing games.

MMOs could work, yes, but I still don't see the advantage. There is just no way to ever solve the latency problem. Yeah, yeah, I know, there are tons of people screaming that the latency is not a big deal, but you know what? They're wrong. They clearly haven't played space sims and racing games with low latency, which are actually good!

Re:Latency (2)

0123456 (636235) | about a year and a half ago | (#42786723)

There is just no way to ever solve the latency problem.

You could put a server in every house with a direct connection to a monitor, then you only get one frame latency.

Re:Latency (1)

HaZardman27 (1521119) | about a year and a half ago | (#42787011)

Quick, market this idea to Intel and Nvidia before someone else figures it out!

Re:Latency (1)

gmueckl (950314) | about a year and a half ago | (#42788735)

One frame of latency? Only on CRTs. Modern games want to run at 60fps, that is 16ms per frame. Modern LCD screens do so much postprocessing on the video signal/image that it typically takes them at least 30 to 40ms to display that image on screen. Or as Mr. Carmack put it: "It's faster to send data packets across the atlantic than to display a frame on the screen" (QuakeCon 2012 keynote, quoted from memory)

Bridge for Sale (0)

Anonymous Coward | about a year and a half ago | (#42786671)

"you can do CGI essentially realtime".

Statements like this does not exactly convince me their platform has a future. Anyone thinking movie CGI can soon be done realtime are plain kidding themselves.

Re:Bridge for Sale (1)

hedwards (940851) | about a year and a half ago | (#42786705)

I'm sorry, but that's BS, what you're talking about is completely different from what he's talking about, and you're being deliberately obtuse. What takes the most time is designing the things to be rendered. And yes, if you throw enough computational power behind it, I'm pretty sure you can get it to essentially real time. Whether or not that's cost effective is a completely different matter.

Also, you're a douche for taking it out of context.

Re:Bridge for Sale (1)

Kielistic (1273232) | about a year and a half ago | (#42786991)

Throwing more computer power at it was always possible. The context of the quote in the article indicated the speaker thought movie grade CGI can be done in real-time because of advancements made in the last few years. In any reasonable interpretation of the word "possible" it is not possible. In other words "essentially" you can't do CGI real-time and anyone who says you can is either lying to sell you something or doesn't understand the technologies they're talking about.

Why are you so defensive about this?

Re:Bridge for Sale (3, Insightful)

hairyfeet (841228) | about a year and a half ago | (#42788107)

Uhhh...what EXACTLY is there to understand? The guy said that movie quality CGI is possible in real time and looking at the football stadium full of computers required to render "The Lord Of The Rings" which is over 16 years old shows that to be total bullshit.

I mean sure if you had infinite money to build a high rise filled with nothing but tesla cards sure it would be POSSIBLE, but it sure as hell ain't gonna be done by this company or any other and actually make a cent in profit as the cost of all those machines (plus power and cooling) will be more than they could ever make selling the service.

So sorry but he is full of shit unless you count "movie CGI" to be on the level of the first Tron from 1982, anything made in the early 90s or later is just gonna take more horse than is possible for this company to muster.

Dont think it will work. (0)

Anonymous Coward | about a year and a half ago | (#42786701)

Who would write a game that only works on one platform? A proprietary software as a service like platform that they don't have control over? OnLive will have to make their own game, its too much a risk for anyone else. Good luck getting investors for that after that whole rename to dump the old stock business.

The future, not yet, but soon maybe. (1)

Zeromous (668365) | about a year and a half ago | (#42786765)

It's too bad Onlive will be the "prodigy" of online gaming services. By that I mean, early to the party, home in bed before it gets started.

The delta between quality of graphics, render times and bandwidth to send completed frames is getting smaller ever day. I just don't think it's where it need to be for anything beyond local broadcast.

What we will see is road gaming using a user's local console first. Cloud needs to be everywhere before we can outsource the frame generation to a laggy internet-shared based service. For Chrissakes, youtube can't even deliver me a super bowl commercial without buffering on my 30Mbit connection.

constraints (0)

Anonymous Coward | about a year and a half ago | (#42786769)

It's funny to hear this guy talk about computing being unconstrained, but neglect to mention that's the least of their problems. What about financial contraints? They can't even afford to cut paychecks to their employees, so they had to let them all go, but they want to build a real time CG system with unlimited compute power? And they are going to pay for that how?

Onlive should do it (0)

Anonymous Coward | about a year and a half ago | (#42786799)

They should bring their services to developing countries where it makes more sense (consoles are very expensive).
It could also help improving the internet infrastructure there.

Pricing failure (1)

Anonymous Coward | about a year and a half ago | (#42786809)

I never purchased anything from onlive because the pricing wasn't competitive with steam/amazon/greenmangaming/etc.

I would have expected the pricing to be even cheaper because I don't have a copy of the files locally in case they go belly up.

constrained by latency and video compression (1)

MaxToTheMax (1389399) | about a year and a half ago | (#42786813)

I must confess, I still don't get OnLive. Sure, you can have powerful hardware rendering immaculate frames, but then you have to use lossy compression to get those frames to the screen, so you probably end up with inferior visual quality (and a subpar framerate.) Not to mention the latency concerns others have mentioned here-- OnLive might be fine for a single-player game, where there's only one round-trip, but a multiplayer game? Not a chance.

Let's play a FUN game! (0)

Anonymous Coward | about a year and a half ago | (#42786909)

Let's play something like Super Mario Bros! That was so much fun! Much more fun than playing Onquack Lifailure.

Haha, did you see what I did there?

Just in case you didn't pick up on the joke, I changed "OnLive" to resemble Barack Obama's name. I am the funniest person. Mod me up for telling a funny, bitches.

Re:Let's play a FUN game! (0)

Anonymous Coward | about a year and a half ago | (#42786985)

You constantly tell non-funny jokes and then laugh at yourself at work while everyone else gives blank stares, don't you?

interactive cinema (0)

Anonymous Coward | about a year and a half ago | (#42787013)

imagine toy story where you control an incidental character like Rex and follow the story from Rex's perspective then jump to slinky etc etc. the interactivity would be increased as you could effectively become the director/editor of any CGI film you're watching

Solving a problem that doesn't exist (1)

Alphadecay27 (1277022) | about a year and a half ago | (#42787019)

Mid-range hardware is insanely cheap these days and will play all but the most high end games. Even tablets and smartphones can handle some pretty intense gfx. The next gen of consoles looks like it won't even be trying to push the envelope on performance because it is already good enough. My gaming rig is about 4+ years old and I'm pretty happy with it. Why exactly would I want to push rendering into "the cloud"?

If they can produce a kick-ass game that cranks everything to 11 with no lag, it might generate some interest. Which publisher is going to push out something like that for a service that seems to be tanking?

Distributed physics (1)

Dan East (318230) | about a year and a half ago | (#42787131)

The interesting part of this, to me, is the potential to have both a larger and more intricate physics simulation. Essentially you would be distributing the physics across many processors, then player interactions would be fed into that. Thus there would be a single physics simulation occurring for everyone, instead of the more typical method where each client is performing its own simulation on local objects and simply reporting back to the server the raw position of various affected entities.

Whether the actual rendering also takes place at the server level or not doesn't matter - the position of the objects would managed by the servers. This would allow vastly more realistic scenarios, especially for MMOs. You could get into things like erosion, plant growth, branches falling off of trees, etc. IMO, visually games are pretty good, and the problem is now the kinematics, nuances and ambiance of the virtual world itself, and not just the more superficial eye-candy (pixels).

AI would be better (1)

elucido (870205) | about a year and a half ago | (#42787429)

It could also translate into much better AI. For MMO AI is very important.

Re:AI would be better (0)

Anonymous Coward | about a year and a half ago | (#42788017)

Not really. AI isn't computationally expensive unless you're creating some sort of expert system (learning) or a simulation-style game with many, many agents. The AI you'd find in any WoW-style MMO is trivial.

Re:AI would be better (1)

Luckyo (1726890) | about a year and a half ago | (#42788197)

MMOs don't need "better AI". They need CONSISTENT AI.

In many cases, AI being too good is actually a very bad thing for a MMO.

Re:Distributed physics (0)

Anonymous Coward | about a year and a half ago | (#42787983)

There's a niche MMO that already does this. Forgot the name, but physics calcs are shared across servers, allowing for impressive effects.

Genius (1)

BaronAaron (658646) | about a year and a half ago | (#42787263)

I think this is a smart move by OnLive.

My experience with the service showed the technology worked, but the pricing model sucked. There was no way I was going to pay almost full price for a game I don't actually own. It would be like Netflix asking $19.99 for every movie you wanted to stream. A monthly fee for all you can eat would have been much better. Of course OnLive was at the mercy of the game publishers who, I'm guessing, didn't make a Netflix-like business model possible.

If OnLive can get games produced directly for their service they'll have more control over the pricing model. MMO games are a good choice because their players are already accustomed to paying a monthly charge. WoW is popular, in part, because it sacrifices high-end graphics so it can run on low-end hardware that your average casual gamer has. Imagine a MMO with Crysis level graphics that can run on a netbook. That's an advantage game developers might be interested in. Another advantage for developers is near zero effort spent on anti-cheat mechanisms. There are no local files to hack, and the network stream is just a video feed and input controls.

Re:Genius (0)

Anonymous Coward | about a year and a half ago | (#42787769)

Imagine a MMO with Crysis level graphics that can run on a netbook.

Now imagine it compressed with 1Mbps H.264 so it can actually get over the Intertubes to the netbook.

Worst case scenario (1)

Schnapple (262314) | about a year and a half ago | (#42787789)

To me this is like the worst case scenario. Bad enough that OnLive might make an otherwise good looking game look and play like shit, but now they're going the rest of the mile and saying that games should be changed and designed for the service.

No, OnLive, go fuck yourself. Your idea will never work technically or logistically and you need to hurry up and die.

Still streaming (1)

Dunge (922521) | about a year and a half ago | (#42788203)

Amazing native image is useless when streamed with compression to 60% of the quality.

NVIDIA's Grid (2)

killdashnine (651759) | about a year and a half ago | (#42788207)

With NVIDIA entering this market, how relevant is OnLive?

Re:NVIDIA's Grid (0)

Anonymous Coward | about a year and a half ago | (#42790037)

OnLive still owns patents regarding this market so until someone bigger gobbles them up, they are very relevant.

realtime CGI? (2)

hierophanta (1345511) | about a year and a half ago | (#42788731)

you can do CGI essentially realtime

So, wth have video games been doing thus far?

Re:realtime CGI? (0)

Anonymous Coward | about a year and a half ago | (#42789249)

There really is no major difference between the rendering you see in movies and what you see in a game. Just different amounts of source data, and that is rapidly changing as us graphics coders develop new tricks to push even more detail with less work.

An example of one trick that provides a huge jump in quality. When Shader Model 3.0 hit consumer graphics cards a lot of graphics coders shifted towards developing full screen effects that weren't just extra niceties. One which I rolled with was rendering lakes, oceans, and rivers in a single full screen pass after every opaque object (with a second pass between layers of translucent objects). I no longer had to perform complex calculations for a water mesh on the processor, I could simply represent the water and it's boundaries in textures, and have the "physics" done in a pixel shader which is highly specialized just for the kind of math required. Not only did it free up precious processor time, it allowed for pixel perfect water, usage of color extinction curves, as well as opened up the possibility to use even more complex lighting on the surface of the water. It's relatively trivial to create a photo realistic lake or ocean in pixel shaders on 5 year old gaming hardware, in fact I have to intentionally make the water look worse because it would make everything else look cheap.

Can I Buy It On Physical Media? (0)

Anonymous Coward | about a year and a half ago | (#42789347)

Being able to buy a game on physical media, insert it in my game PC which is not connected to the Internet, and play it is my only unchangable criteria for whether or not I buy a game.

A proper gravity model would be a start... (0)

Anonymous Coward | about a year and a half ago | (#42789685)

Anybody else sick of 'CGI' having totally unrealistic inertia and gravity models, rendering (no pun intended) all the hard work useless, since it's always obvious that the CGI isn't real? Just watch Gollum in Lord of the Rings jump off anything, or just his general movement, or the recent Planet of the Apes movie, or ANY movie that uses CGI - why did the company (or companies) that wrote the programs for this, deliberately set up the inertia and gravity models incorrectly? It's not as if it's difficult to do - the basic equations of motion will always give realistic results, so what did they do to change them, and why on earth would anybody do that?

Stationery, modern CGI looks incredible, but as soon as it moves, you know it isn't real. Why does nobody on these teams do something about this, when they put in so much hard work on the models themselves?

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?