Hard Light Productions Forums
Off-Topic Discussion => Gaming Discussion => Topic started by: Topgun on March 24, 2009, 08:27:06 pm
-
Full article (http://pc.ign.com/articles/965/965535p1.html)
Before I dive into what OnLive is and how it works, let me start by saying that you should read every word of this article as this service has the potential to completely change the way games are played. If it works and gets proper support from both publishers and gamers, you may never need a high-end PC to play the latest games, or perhaps even ever buy a console again. That is not an exaggeration.
Just announced at this year's GDC, OnLive is an on-demand gaming service. It's essentially the gaming version of cloud computing - everything is computed, rendered and housed online. In its simplest description, your controller inputs are uploaded, a high-end server takes your inputs and plays the game, and then a video stream of the output is sent back to your computer. Think of it as something like Youtube or Hulu for games.
The service works with pretty much any Windows or Mac machine as a small browser plug-in. Optionally, you will also be able to purchase a small device, called the OnLive MicroConsole, that you can hook directly into your TV via HDMI, though if your computer supports video output to your TV, you can just do it that way instead. Of course, you can also just play on your computer's display if you don't want to pipe it out to your living room set.
When you load up the service and choose a game to play (I'll come back to the service's out-of-games features in a bit), it starts immediately. The game is housed and played on one of OnLive's servers, so there's never anything to download. Using an appropriate input device, be it a controller or mouse and keyboard, you'll then play the game as you would if it were installed on your local machine. Your inputs are read by the plugin (or the standalone device if you choose to go that route) and uploaded to the server. The server then plays the game just like it would if you were sitting at the machine, except that instead of outputting the video to a display, it gets compressed and streamed to your computer where you can see the action. Rinse and repeat 60 times per second.
To make this happen, OnLive has worked diligently to overcome lag issues. The first step in this was creating a video compression algorithm that was as quick as possible. The current solution only introduces one millisecond of lag to encode the video, which alone is completely unnoticeable to you. Obviously, a fast internet connection is required on your end to stream the gameplay video. A 1.5 mbps connection (which is usually what base-level DSL is rated at) is required for standard-definition video (480p), while a 5.0 mbps connection is required for HD (720p). The actual necessary speed is a tad less than advertised, so as long as your provider says you have these speeds, you should be OK.
The OnLive controller and MicroConsole.
The cool thing here is that your only requirement is a capable internet connection and some sort of computer. In theory, you should be able to play Crysis on a netbook. A handful of us have played the game, at its highest settings, on a MacBook Air with the service. Not only is the game not normally available on the Mac (outside of running Boot Camp), but the MacBook Air is hardly a gaming device, and yet we were able to hop in and play it as smoothly as a nicely-specced machine. We also played Burnout Paradise on a similarly-equipped PC laptop, and despite how quick that game is, it ran and played fine as well.
Do the games run at 60fps? Technically, yes, but the video stream makes it feel less so. They're still smooth, but Burnout wasn't as brisk as it is on a PS3, for instance. But make no mistake - everything we tried was completely playable (and most importantly, quite responsive), and being that you're able to play these games without any dedicated hardware, that's a huge, huge thing.
Another article (http://pc.ign.com/articles/965/965542p1.html)
-
I'll believe that when I see it. As other commenters elsewhere (http://games.slashdot.org/games/09/03/24/135212.shtml) have noted, the potential lag between your input and the result appearing on your screen, may be a real problem. Also, I kinda like having Games (and the Hardware needed to play them) here, at my home, where I can play whenever I want without worrying about network outages or the publisher or OnLive deciding to drop support for that game. Although I would really like to know what kind of hardware they have running in order to accommodate several HD Video streams to their clients.
-
I also didn't see support for resolutions greater than 720. That means I wouldn't be able to play Freespace 2 at 1680x1050 like I can now if I used OnLive. Plus the fact that the whole thing requires the Internet to be able to play is not something I like.
-
And of course, modding would be out of the question.
-
NONONONONONONONONO
This is stupid. This is IDIOTIC. You cannot STREAM BACK VIDEO of any kind of decent resolution at any kind of high quality whatsoever, and the servers required to make this work would be completely insane. Rendering games on a computer allows them to be specifically optimized for its hardware and to render things with pixel-perfect clarity. Now way back in 1999 when we were playing starfox 64, this might have had some merit, except that it wasn't possible. I mean, holy crap, is this article a JOKE? Seriously, it has to be - you can feel input lag of 2 FRAMES on 60 FPS, and you expect people to live with input lag that would stretch up to 5 frames? It's STUPID and it's not going to go ANYWHERE. Any graphics programmer, any game programmer, anyone with HALF A ****ING IQ POINT will tell you this is the most idiotic idea anyone has ever come up with.
Seriously, this has got to be an April fool's joke. Please.
...(some google searches later
This is dumb. This is just incredibly, unbelievably dumb. It's just like subscription games - another way to yank more money out of gamer's pockets. Ladies and gentlemen, I bring to you the company that will go bankrupt faster then Enron after it's scam was exposed. This is pathetic.
And I can't help but wonder - what kind of ungodly servers would you NEED to get this to work for MILLIONS of people?!
EDIT: I want to point out that the entire reasons companies are so giddy over this is piracy. The whole ****ing thing is just more DRM.
-
Say goodbye to your network bandwidth...
-
Well, I can see one application for this: Hotels. Imagine having a server for such a system in the basement, with the hotel rooms each having a client for it. That would eliminate most bandwidth issues, and may actually be feasible. But something like this for the mass market? Never.
-
This is an overly retarded idea and there's porn for hotels. If you're with parents, then i guess you'll be watching cinemax porns instead.
-
...blackhole, did you even read one paragraph of the ****ing article? People, actual human beings, have sat down with this thing and tried it out. And according to them, under the conditions that they were presented with, the system was utterly and completely playable. Does this mean that an actual final implementation of the product would work just as well? Of course not. But throwing some completely asinine *****fest over a product that you've never so much as seen in action, much less sat in front of and played around with yourself, makes no sense whatsoever.
Now as for me, I honestly don't really know what to think about this concept. My sole machine is woefully underpowered even by SCP standards, so I haven't been able to play any higher-end PC titles made in the last few years or so. (And I don't own a 360, so I can't use that as an alternative for those games that exist on both.) A service like this would allow me to get a taste of all of the graphical/physics progress that I've been shut out from. And I've been a subscriber to GameTap for quite some time, so I'm already accustomed to the whole subscription-based model. What I don't know is whether paying the price of the service would really be more valuable than putting the money towards upgrading my system, or even building a new one. Maxing out this machine's RAM and giving it a half-decent video card wouldn't just help me gaming-wise...it would make the everyday tasks I perform far less frustrating. And that's nothing compared to the benefits a brand-new system would give me, which with the way hardware prices are falling nowadays would probably be far cheaper than I'd expect. I think I'd definitely have to see far more concrete details, as well as a final pricing scheme, before thinking any further about this.
-
...blackhole, did you even read one paragraph of the ****ing article? People, actual human beings, have sat down with this thing and tried it out. And according to them, under the conditions that they were presented with, the system was utterly and completely playable. Does this mean that an actual final implementation of the product would work just as well? Of course not. But throwing some completely asinine *****fest over a product that you've never so much as seen in action, much less sat in front of and played around with yourself, makes no sense whatsoever.
I read every stinking word of that article, and each subsequent word made my jaw drop lower, and lower, until I was practically gagging over my seat out of sheer disbelief at the stupidity seeping out of it's pores. THEIR version of playable is most definately not what MY version of playable is. Think about this - I spend weeks on end optimizing my graphics engine to ensure it only has 1 single frame of lag, and not only are they transfering this over the internet, by necessity it is a 30 FPS video feed which inherently introduces another frame of lag, and the internet connection itself will introduce yet another frame of lag, in optimal conditions. This thing ain't gonna operate at optimal conditions all the time, and so you will have at bare minimum 3 frames of lag possibly going up to 5 or 6 frames of lag, which is just nasty. You are asking me to take, on face-value, the experiences of a bunch of media crankjobs who are hailing this technology as something that will replace consoles completely? I think I'll start taking them more seriously when they stop preaching armegeddon like gleeful schoolchildren searching for candy.
But here, just to make you happy, lets just totally ignore the entire gameplay experience. How the f*ck do you sustain enough server power to run Crysis at near full power for tens of thousands of people playing distinct games of it at once? Do they discuss THAT in the article? No, they just vaguely mention "gaming servers." I think they have a few *****INGLY HUGE infrastructure problems to solve. You can't use cloud computing if the game is sucking up all available resources.
Now, let's go ahead, for arguments sake, ignore that problem too. We still have the fact that the demographic this is appealing to is not as large as these people think, and that it is only going to serve to get a group of people who don't normally play high-end games playing high-end games - It's not going to destroy the console market entirely and it never will unless by some miracle it DEFIES THE LAWS OF PHYSICS. So in a best-case scenario, this technology is going to give the poor man access to games like crysis. And it won't do a goddamn thing more.
-
Thing is, a lot of people go through a gateway system, putting several computers Internet access on the same line. Hows this thing going to perform when there's 3 computers in the same house, all trying to play different games?
I can see possibilities for this as an option, but as a replacement, I just don't see it happening right now, most Countries Broadband/DSL capacity just isn't up to the job, even with excellent compression routines.
-
I can see possibilities for this as an option, but as a replacement, I just don't see it happening right now, most Countries Broadband/DSL capacity just isn't up to the job, even with excellent compression routines.
PRECISELY
-
As an option, yeah keep it, but DEFINITELY not as a replacement
I like the idea of having everything rendered here and have it on MY screen and save my bandwith for whatever else I'm downloading in the background. It's just going to cause too many problems.
It's like trying to replace a human pilot with an AI pilot, there are just TOO many variables in the cockpit, no matter what, you'll always need a human pilot there to take over if something goes wrong. There's just too many variables with that idea, and if the host comp crashes, well guess what, f*ck THAT!
-
It's not gonna fly for flight sims or FPS games. There might be some credibility for games that don't rely on twitch but instead tactics and strategy (like Total War, Red Alert or something like that).
I'm still not going to pay for it.
-
Pay for it? As well as the massive over-the-limit fee from your internet provider? I think I will stay with my own hardware. I would also like to own my private copy of the game; I'm not really a fan of "internet computing". It's too vulnerable.
-
This doesn't sound impressive to me. Anything going over the internet like that will easily have at least a couple frames of latency and will cause horrible input lag. The small resolutions and highly compressed video means a heavy image quality reduction too.
THEIR version of playable is most definately not what MY version of playable is.
Exactly. Reviewers these days often have a very different idea of "playable" and "responsive" than gamers do. :p
-
Ye gods...even assuming an optimal download/upload speed (in my case, 300 kbps and 100 kbps, respectively) The resolution would need to be shrunk to the point of fugliness. It would be worse on big-screen monitors.
Unless they have a phenomenally well-created lossless video codec that allows them to stream video at a low bitrate and high resolution--which I rather doubt--then this thing's going to go the way of the Spruce Goose: large, expensive, frustrating to create, and outdated by the time it makes a major public appearance.
-
I loved the concept basicly but I would hate it become a reality. Damn, I hate Steam too. I want a goddamn DVD in my hand that can be played as often as I want, without internet, DRM and ****.
Would be awesome for causal games still.
But I doubt you could ever play high-speed action games. You just can't stream in that quality.
You press keys, those info is send, the server calculates the action, recalculates the world, sends it back.
Would need lightspeed to work.
-
You are asking me to take, on face-value, the experiences of a bunch of media crankjobs who are hailing this technology as something that will replace consoles completely? I think I'll start taking them more seriously when they stop preaching armegeddon like gleeful schoolchildren searching for candy.
I'm asking you not to throw some sort of ****ing hissy fit over a technology that probably won't affect you in the least...and that's even if it gets off the ground. I can understand disagreeing with the concept, but why in God's name did you come out with such an irrationally angry outburst? Please, help me understand that, because I'm at a complete loss.
And of course they're over-sensationalizing the potential of such a technology. They're part of the media. It's what they do. But the fact remains that they're the only ones in question who have sat down in front of an implementation of the service and played on it with their own two hands, which automatically lends them nearly infinitely more credibility than any of us sitting here discussing the topic. Considering the fact that these guys have based their entire careers on the gaming industry, I'm going to give them the slight benefit of the doubt and assume that they know how games are supposed to play.
But here, just to make you happy, lets just totally ignore the entire gameplay experience. How the f*ck do you sustain enough server power to run Crysis at near full power for tens of thousands of people playing distinct games of it at once? Do they discuss THAT in the article? No, they just vaguely mention "gaming servers." I think they have a few *****INGLY HUGE infrastructure problems to solve. You can't use cloud computing if the game is sucking up all available resources.
So in what strange fantasy world have you figured out this hurdle while the people who are developing said system haven't? Could it possibly be the case that, just maybe, the developers trying to market this as a viable product have taken this completely into account and know how to deal with it? Last time I checked, the system specs required to run Crysis adequately aren't some form of arcane knowledge, revealed only to the Illuminati.
Now, let's go ahead, for arguments sake, ignore that problem too. We still have the fact that the demographic this is appealing to is not as large as these people think, and that it is only going to serve to get a group of people who don't normally play high-end games playing high-end games - It's not going to destroy the console market entirely and it never will unless by some miracle it DEFIES THE LAWS OF PHYSICS. So in a best-case scenario, this technology is going to give the poor man access to games like crysis. And it won't do a goddamn thing more.
No, it won't "destroy the console market," not any time soon, anyway; that's just marketing hype. And it probably won't destroy the relatively small hardcore-PC-building market, since those types will keep blowing obscene amounts of money on the latest and greatest until the day they die. But where it does have a chance to shine is in the medium market, the market of people like me who don't have super-systems lying around yet wish to play newer titles that we'd never be able to otherwise. Or those people who would love to be able to rent PC games for a short period of time, yet have no means to do so. That's the niche they're trying to fill, and if they can get the technology to work as well in full-scale execution as it seemingly has in the current setup, I could easily see them getting a significant amount of business.
Would need lightspeed to work.
...anyone else wanna point out the irony here? :p
-
No. Unneccessary, really. The valid concerns here are:
1. Can OnLive maintain a sufficiently powerful array of servers to handle the datastream?
2. What about latency inherent in the internet? I mean, with the current infrastructure, a system like this will be hard to use with anything requiring split-second reactions. Turn-based games sound rather feasible, by contrast. Now, the developers have tested it, apparently, but without knowing their testing setup, I'll take anything they say with a grain of salt. As I said earlier, I can see this working over LAN, but I'm kinda dubious if it works over normal internet connections.
-
...anyone else wanna point out the irony here?
:lol:
Well, nearly lightspeed, it's actually a fraction under lightspeed in copper cable, and slightly slower still in Optical cable ;)
Thing is, when dealing with stuff like IP, you're heavily reliant on the quality of service, since IP carries no QoS protection whatsoever. A set-up might work just fine in a country with a good Internet service, and in a limited context, but there's a slight problem here that occurs to me.
The fact is, this sort of system will appeal to the range of people who either have the lowest access to new technology, or are in a wage bracket that wouldn't support maintaining a high-end system. The problem I see here is that, particularly in the first case, the quality of the Internet service also drops with the lack of access to technology, this sort of thing would be great, for example, in less affluent countries, but the Internet service in those countries is also going to be worse, thus removing one technological bottleneck and introducing another.
-
Well, my first thought was that OnLive will be a good service, after all, this does mean you can "play" Crysis on a low-end machine with the gameplay video streaming back to you...
I'm a bit worried though, as this might have the potential to destroy the PC Gaming market from the inside out. What if a game is only available for OnLive and not Steam or GoG? What if OnLive or something similar becomes the standard?
And my second worry, I just tried something similar to game streaming... Remote Access on Linux. And it was choppy as *HECK*. I had to wait at least FOUR SECONDS before I got the desktop cube zoomout on the machine that I was accessing my big PC with. Which is a lot longer than I should have had to wait.
-
Naa, I don't think this will be a big competitor to standard game releases. There are still those gamers who'd rather own the game and the machine it's running on themselves. It could be a danger to the console market, but even that depends on OnLive's ability to deliver the promised performance to everyones satisfaction.
-
You are asking me to take, on face-value, the experiences of a bunch of media crankjobs who are hailing this technology as something that will replace consoles completely? I think I'll start taking them more seriously when they stop preaching armegeddon like gleeful schoolchildren searching for candy.
I'm asking you not to throw some sort of ****ing hissy fit over a technology that probably won't affect you in the least...and that's even if it gets off the ground. I can understand disagreeing with the concept, but why in God's name did you come out with such an irrationally angry outburst? Please, help me understand that, because I'm at a complete loss.
...because it really does affect me in a huge way? I'd try explaining but it involves things like "business" which I'd rather keep out of this. Now of course this doesn't validate the fact that i had a hissy fit over it. I had a hissy fit over it because I felt like having a hissy fit over it and I thought it was totally ridiculous. If you don't like that then you are free to throw your own hissy fit over my hissy fit but it's not liable to make much of a difference. :P
So in what strange fantasy world have you figured out this hurdle while the people who are developing said system haven't? Could it possibly be the case that, just maybe, the developers trying to market this as a viable product have taken this completely into account and know how to deal with it? Last time I checked, the system specs required to run Crysis adequately aren't some form of arcane knowledge, revealed only to the Illuminati.
I'm not saying they haven't figured out a way to solve it, but I question the feasibility of buying that many systems without charging ridiculous amounts of money. Alternatively, the issue is that they won't let you play the games to their fullest graphical potential, which really kills a lot of the appeal because a large number of people would only be playing it on slightly higher graphics settings and would be better off just playing it on their computer
That's the niche they're trying to fill, and if they can get the technology to work as well in full-scale execution as it seemingly has in the current setup, I could easily see them getting a significant amount of business.
I agree
anyone else wanna point out the irony here?
He has a point. HTTP is not designed for live streaming of video, and even though they have obviously spent time working around these issues, there is a lot of lag that is inherent in the physical infrastructure of the internet itself. As my friend pointed out, the absolute minimum lag time is 2 times your latent ping to the server, which on a T1 line from his college would be something like 20 ms, times 2 is 40 ms, + 1 for encoding, + 16 ms for rendering the frame on the server in the first place, + a good 5 or so ms for the computer to process the incoming video feed, which means the minimum lag is about 57 ms, or almost 4 frames per second.
Current broadband is excellent for transferring extremely large amounts of data, but it isn't well known for being particularly reliable. Again, they can compensate for this in the technology, but you are hosed if your internet cuts out or if your ISP gets clogged up, which is going to be a huge problem if you have hundreds of people in a concentrated area all with constant live streams of video feeds sucking up their entire internet connection. ISPs are NOT designed for people to be using 90% of their potential internet connection all the time.
-
I loved the concept basicly but I would hate it become a reality. Damn, I hate Steam too. I want a goddamn DVD in my hand that can be played as often as I want, without internet, DRM and ****.
/me agrees wholeheartedly with Vidmaster.
-
Like or not, a new game system that dont allow some level of piracy is condement to fail.
Onlive is at last a new idea, but an idea that will only work on US, and at some level.
-
Like or not, a new game system that dont allow some level of piracy is condement to fail.
Onlive is at last a new idea, but an idea that will only work on US, and at some level.
Oh yeah!
Just likely to create -more- pirates.
-
(http://cad-comic.com/comics/20090327.jpg) (http://cad-comic.com/comic.php?d=20090327)
How quaint. ;)