About the Author
Mass Effect
Final Fantasy X
Batman:Arkham City
Borderlands Series
Weekly Column
Champions Online
World of Warcraft
DM of the Rings
Good Robot
Project Frontier

OnLive Explained

By Shamus
on Saturday Jan 16, 2010
Filed under:


Back in March of 2009 I took a look at OnLive and said that while it was a great notion, the technology was most likely infeasible, impractical, or only useful to a very narrow audience.

Now here is a talk from Steve Perlman at Columbia University’s Fu Foundation School of Engineering and Applied Science, making the case that the idea works:

Link (YouTube)

As an early critic, I have to make the following observations:

1) While the name Steve Perlman is new to me, I think it’s safe to say that he’s not a snake oil salesman.

2) If he was a snake oil salesman, he’d be giving these talks to investors and staying clear of roomfuls of engineers.

3) He points strike me as fairly persuasive.

Some of the points are still a bit hazy for me, but I am open – and perhaps even eager – to believe that I was wrong and that this is a thing that could be done in our universe.

I’m very curious about how the licensing would work. Initially it sounded like you would just purchase the rights to play a particular game. In this talk it seems more like a GameTap or Netflix type service, where you can just pay a flat fee and eat all the games you like, buffet style. I’d still rather treat games as books, but treating them like cable TV would be preferable to how it works now, where publishers try to sell you the rights to be hassled by a game until they get sick of your whining and shut down the authentication servers.

I still have a lot of objections, questions, and misgivings, but I’m no longer inclined to denounce it as crazy talk. However this plays out, it looks like smart people have an idea they think will work, and it sounds like it would be a welcome refuge for PC gamers.

Whether it ultimately works or not, I’m really curious how it will play out. I’m going to be following this one eagerly.

Comments (50)

  1. randy says:

    OnLive explained: it’s like DIVX (not the codec, see here: http://en.wikipedia.org/wiki/DIVX_%28Digital_Video_Express%29 ), but for games, and completely digital. It should crash and burn. It deserves to crash and burn. It probably will.

    Then again, MW2 should have crashed and burned, and it didn’t. Who knows. People will buy any POS with a good marketing campaign.

  2. Samopsa says:

    I think this can work. If you told people in 2000 that they would be streaming HD-movies from the internet in less then a decade, they would laugh at you. I think the same inflation of internet speeds applies to this.

    In another decade, most people now on DSL/Cable could be on glass-fiber connections, effectively increasing their downstream tenfold, and and even larger increase to their upstream. It’s not impossible.

  3. Eric says:

    I’d rather use gamefly for my rental service, bigger selection of games for more platforms, and if any errors pop up I can send it back and get another copy without having to talk to a techie.

  4. midget0nstilts says:

    There is another aspect, namely the video card companies. This is essentially something that would make getting a new video card pointless. Given the graphics whoring that modern games emphasize, I don’t think I’m too much of a conspiracist when I say they have a stranglehold on the industry. Like the oil companies that benefit from perpetually low mileages, OnLive is not something AMD and Nvidia will stand for.

  5. Primogenitor says:

    It has a number of strong draws, no-hassle upgrades & updates being my favourite, so if it can work for little/no cost over existing games, I think it might be competitive.

    Oh, and I think AMD/NVIDIA/etc would like it; its got to be rendered somewhere first, and bulk-buy modular server farms would still use their kit.

  6. Mark says:

    I’m curious as to how they intend to solve the issue of latency, which probably will not ever be reduced to an acceptable level.

  7. Jason says:

    If you’re that curious, perhaps you should watch the videos?

    I am looking forward to seeing how their roll-out goes. As a games dev, I very much want this to be a success.

  8. swimon says:

    I’m not really a computer science guy so the specifics are a bit vague to me but it seems like this could work.

    I do have some problems with this though, my main problem is that publishers might see this as a green light to spend all they can on graphics technology. Since everyone would have a computer that could run the games (in this scenario onlive penetration is 100%, unrealistic but still). This would in turn lead to even safer games that try to appeal to absolutely everyone and because of this basicly suck. Of cousre that isn’t a new problem but with onlive’s help it might become even more dire.

    Oh well maybe the industry will grow wise ^^.

  9. midget0nstilts says:


    I think it would hurt them more than help. For one thing, a server farm would be much more efficient than the current model. Assuming that they don’t use their own proprietary technology, they’re probably going to go for the high-end models, usually used by industry professionals. There’s also the fact that they would need less. The video card you have stays in your computer all the time, whereas with OnLive, when you’re done playing for the day, that frees up resources for someone else. It’s like a car that transports only one person and gets nine miles to the gallon, as opposed to a bus that can carry 30 people and still get twice the mileage.

    On the other hand, swimon does bring up a good point. It very well could be a boon to AMD and Nvidia.

    However, since I’m cynical, I’m betting on the former. In fact, I bet what you’ll see are hardware (or perhaps APIs) that force you to use a video card. A sort of video card DRM, if you will.

    Ack, I guess either way we’re screwed.

    EDIT: Now that I think about it, one thing the companies could do is go with the DRM and charge OnLive royalties for using their hardware. Er, not to give them any ideas….

  10. Nazdakka says:

    I was intially skeptical when I heard about this, but that talk makes it look a lot more feasible than I had previously thought. 80ms total latency is very fast, but doesn’t sound completely unreasonable given they are using hardware that is solely dedicated to this one task. The vagiaries of modern internet connections are going to hurt them in real-world conditions, but time is in their favour there – connections are only going to get better. I do wonder if their magic real-time compression system has any any other applications – surely Youtube would kill to be able to use it…

  11. Factoid says:

    @randy: I think you’re off base on DivX. We have systems in place that are as restrictive as DivX now, or more so.

    DivX let you rent a movie and keep the disc, so you could instantly rent it again if you wanted by paying the rental fee again via a modem (which would eventually have been upgraded to an internet connection service).

    We have the exact same system now with cable On-Demand video services. Apple TV. The xbox and ps3 video marketplaces.

    In most ways these services are more restrictive than divx ever way. Divx guaranteed a full playthrough regardless of when you rented it so you could take as long as you wanted to watch and then stop half way through and finish it a month later if you wanted.. Most VOD services give you exactly 24 hours from the moment of purchase and if you have to stop playback for some reason you’re SOL if you don’t finish before time runs out.

    Obviously a disk-based delivery service wasn’t that efficient, but it was way ahead of its time, plus those disks had special features and stuff on them usually, which has never been the case with digital VOD stuff.

  12. J Greely says:

    Steve’s previous Big Idea was Moxi, which lost more money each year than most startups ever get (the merged Moxi/Digeo finally died the final death recently). Everyone I know who followed him there from WebTV swears they’ll never work for him again, for both business and personal reasons.

    The only reason I think this could work is that he managed to get Arnold onboard.


  13. Nick says:

    I am still worried about latency as it is. 80 ms sounds like a pipe dream. I mean, unless a server is within my city, I have horrible 100+ latency to anywhere else in our “great” country.

    But I do see a huge positive for the “dying” PC industry. Since these games will be developed for PC’s, that means that this so-called console, which if popular, will encourage game developers to make more PC games, rather than simply consoles.

    But, to turn that positive into a negative, I see companies either making OnLive ONLY version of their games (to optimize it for a single hardware config), or the games will all suddenly come down with intense bouts of consolitis.

  14. rofltehcat says:

    Love the idea. But it will take a few years until this reaches europe. So I hope that they US guinea pigs love it and it becomes a great success :)
    Only watched the first 4 vids for now and I wonder how much this will cost. Could be the steam of the future.

    Watched the fifth vid and he didn’t really answer the question about costs :/ (except that there will be different ways to buy/rent the games)

  15. Matt` says:

    AMD and nVidia could be brought on board with the idea, if OnLive are a good enough customer for hardware to stock their data centres.

    If the average OnLive gamer is effectively running a higher-spec machine than the average other gamer… could be to their advantage (depending on how much saving there is to sharing 1 high-spec server between many gamers).

  16. Nick says:

    Another worry I have is how licensing the games will work. I really hope it’s not another pay-per-item service. If it’s a fixed amount per month, buffet-style, I’d love it. But I highly doubt it.

  17. scragar says:

    I have to wonder about the payment methods for this, is it subscription, or do you buy a game that you never actually own?
    If it is a subscription are you paying the fee for each game? A number of games? All games?

    I also need to question the internet speed, my latency to the big cities near me tends to be 80ms or so one way, and I don’t get 2Mib download, ever(despite that being what I paid for). I know I live in the UK, this technology will take a while to appear over here in a reasonable capacity, but I’m still a bit concerned.

    One last thing, W00t, gaming on linux. If it’s platform independent then that means linux is now as good a gaming platform as windows, that is so awesome, in recent years I’ve stopped gaming because I couldn’t stand rebooting into windows to play games(nothing against windows, but by the time I’d rebooted, waiting for windows to load, launched the game, waited for it to update, I’ve already gotten bored of wanting to play a game and rebooted back into linux to play a flash game in my browser).

  18. glassdirigible says:

    If you watch the videos he explains how they essentially do as the crow flies connections to you by cutting deals with all the ISPs. It sounds like it’s technically feasible.

  19. midget0nstilts says:

    Also, I wonder how this will play out with the console makers? Given that they already sell their machines at a loss, this could cut some costs for them.

    Hell, they might not even see a point in having consoles anymore, and maybe, for example, Nintendo will start making all of its games for the PC.

  20. A Gould says:

    It looks like a good talk – they’re showing an awareness (and admitting) to the technical issues that have to be worked around. And I give bonus points for admitting that certain slides are marketing hype.

    In theory, this looks like it should work. In practice, I’m in Canada and likely too far away to try it.

    From a purchasing angle, I can see the logic – I used to subscribe to GameTap, because it let me try out games cheaper than it would cost to buy them. (And I don’t normally go for the high-octane games anyway, so that wasn’t an issue). If this works on a similar subscription model (I pay $X per month, and play whatever I want), and it works as advertised…? I could see myself signing up.

  21. glassdirigible says:

    From watching video four I get the impression that they will have rent or buy as options for games.

    I wouldn’t be terribly surprised if they had a smallish subscription fee on top of that, similar to XBox live.

    Provided the subscription fee isn’t greater than that of an Xbox live subscription and games don’t cost more than the retail variant I’m all for the service.

    I’d love to leave Windows behind, as well as only needing a 13″ laptop for gaming.

    They showed the movie trailer, so I wouldn’t be terribly surprised to see them expand into movie rental as well if this venture is successful.

  22. Koriantor says:

    This looks really interesting. It seems that steam might have a competitor in the PC community.

    I’m curious thought how the PC standards will be affected. I’ve always been a fan of Mods. They allow users to develop their own creative ideas but I wonder if mods are going to be even possible on this system. I mean, essentially this is just streaming video that you can control, so how could you possibly do mods? I don’t see a solution.

    I play a heavily modded Oblivion (you know, hi-res textures, better spell effects, much much better UI, and various little things to make the world a little more immersive, which oblivion is in dire need of). It looks and plays absolutely amazing, but if I play it on ONLive’s servers, how am I supposed to mod it? I won’t have access to the actual game files. Plus, Oblivion was so filled with bugs modders released unofficial patches to supplement and complete what the developers didn’t do. What happens when you get games like that which could be much better post-modification?

    Look at how popular counter-strike and TF2 are. They both started off as mods and it’d be a shame to see the death of mods because of this service.

    For some reason, I feel like the console war might end from this, but I sense a war between Steam and ONLive. Valve will surely find some way to compete. If this happens, I’ll probably stick by Valve unless ONLive gives me a modding solution.

  23. Dys says:

    When it first appeared, the very idea was laughable.
    You can still find posts where people explain why Onlive could never work, because of the way video compression works.

    Now it seems that their compression works entirely differently, and it sounds like they do something similar to mp3 encoding where they cut out huge amounts of ‘imperceptible’ data. He says repeatedly that they exploit the limits of human perception, in terms of both latency and video quality.

    It’s fun to see people talking about PC vs console, linux vs windows… the nature of Onlive, if it works, would make all such divisions irrelevant. The game is run on high end hardware you will never see or know much about, all you get is the output. Hardware of any kind would be archaic.

    I don’t see this as a redemption of pc gaming, if anything it is an entirely new source of competition for both home pcs and consoles.

    We’ll see if it works, when it works.

  24. Calatar says:

    You know, there’s nothing I like more than “owning” things I have no physical access to, getting an “optimal” framerate of 12.5 fps, and having my game’s speed and loadtimes depend entirely upon uncontrollable variables such as server load and ISP latency.

    I like my insular, understandable rig. When things go wrong, I can DO something about it. I can change my game files, I can mess around with the settings. One of the biggest draws that PC games have is their configurability. And you cannot deny that out of necessity, the games they have there will not be anywhere near as configurable.

    Maybe they can get some console gamers to use it since they don’t care about configuration as much, but I doubt that PC gamers who are PC gamers because they like what they can do will enjoy this system very much.

  25. B.J. says:

    Steve Perlman looks like the classic case of the engineer who knows nothing about sales and marketing a product. I mean WebTV? Seriously? He thinks because they make an awesome technology people will buy it, but that’s just not the way it works. Ask Sony.

  26. Shamus says:

    Hey Shamus, I think the phrase, “I think it's save to say that he's not a snake oil salesman.” should actually be “I think it's saFe to say that he's not a snake oil salesman.”

  27. Kdansky says:

    Very interesting. This might just be the future of gaming. And if our internets get a bit faster, latency could easily drop to around 30 according to what he says, since more than 50% of the performance is lost on the last mile. And being able to “run” Crysis on an iPhone is just hilarious. :)
    With a fictional pricetag 20$ a month for unlimited games I would be so on board. But it’ll probably be way more expensive.

    “He points strike me as fairly persuasive.”
    “His points …” ?

  28. J Greely says:

    B.J., WebTV peaked at more than 1.1 million paying customers. When I left in 2005, we still had over 400,000, despite the fact that we were way behind the curve in web technology. The service is still out there, still running, for the small number of people whose needs are satisfied with basic html, email, chat, ebay, and porn.

    (Why were we way behind the curve? Because our box had a custom OS and the servers ran Solaris. Once we were acquired by Microsoft, the bulk of the engineering effort went into starting over from scratch with NT servers and CE clients. This never succeeded, and whatever servers are left are almost certainly still running Solaris x86)


  29. Tiki Tok says:

    But how will this revolutionary new service affect the average pirate?

  30. Chris says:

    I think the biggest problem is that you won’t own the game and cannot resell it on. If you don’t play the game any more and your friend wants a copy, he has to pay full price to the supplier.

  31. Aquin says:

    I actually quite enjoy GameTap and I don’t own those games. I do buy games otherwise, but I appreciate a fast service to play stuff.

    Sure okay, this OnLive is a pipedream for 2010. But 2015?? I dunno, it could be pretty cool!

  32. Brains says:

    I’m part of the beta for OnLive – it’s basically unusable. You can’t play it over Wi-Fi (even 802.11n with a max strength signal), and even over ethernet with a 6 MBps downstream connection, I got booted off for insufficient speed. Repeatedly. The one time I was successful at loading a game (Burnout: Paradise – by the way, only a half hour demo), I got to the first loading screen before it kicked me off again. Unless this changes – a lot – it will be totally impractical for most users.

  33. Zerai says:

    I don’t know if he addresses this in the video (not native english makes it difficult) but there is still something I see that will kill this

    80ms of lag seems little, because we see 80ms with the others movement, that is, you move with zero lag in your screen (to nitpick, while there is, it’s so low we can’t perceive it) but with onlive you would see yourself with a lag of 80ms, this is, you look to the right, and it takes 0.1 seconds to change view

    Unless this is solved, FPS, racing and any game with very fast response times (starcraft, for example) would be unplayable here

    I wonder if it’s possible to create a program that simulates this on the desktop…

    (Edit) Also, IIRC, Shamus talked about this in the other article about OnLive

  34. Another Scott says:

    Hey Shamus, I think the phrase, “He points strike me as fairly persuasive.” should actually be “HIS points strike me as fairly persuasive.”

  35. Bailey says:

    Another Scott: If you know what he means ;)

  36. Nick says:

    The biggest things we need to look at are:
    – Ownership of games. We need to either get a game we can later ask them to ship us one (for resale), pay a fraction of the cost to only own it in their service, or pay nothing except a monthly fee for infinite use of all their games.
    – Ensure low pings. I still cannot believe, even with backdoor deals with ISPs that they can optimize their traffic to make it 80 or less.
    – What about developers use of this? Will we see too many games released ONLY for OnLive (laziness of developers to only make sure it works on the one platform). If the controls are with a OnLive controller (versus a KB and mouse), consolitis creep. Will we lose the ability to mod these games, since OnLive obviously won’t allow this at all?
    – What will this do to the DLC model? If we don’t own the game, do we own the DLC? personally, I’d like to see the vast majority of DLC to DIAF (horse armor, or actually anything LESS than what we considered an expansion 5 years ago).

  37. B.J. says:


    You basically proved my point. 1 million people is a drop in the bucket compared to the standard PC internet users even then. And then the technology fell behind and only appealed to a tiny niche. That is hardly a success story. Keep in mind these OnLive guys are bragging about the “future of gaming” just like TV web browsing was the “future of the internet.”

    The problem with both of these things is that their base assumption, that hardware is an obstacle for consumers, is wrong. What’s keeping people away from gaming isn’t hardware, it’s disinterest.

  38. Nathan says:

    Alright, let’s pair this idea up with other forthcoming “future” sorts of gaming technology. A few questions:

    1.)How will this service handle procedurally generated games? For a cutting edge game, procedural generation could easily mean a hell of a lot of strain on both the CPU and the GPU. Will OnLive be able to support this?

    2.)What does buying a license give me as a consumer? Can I use mods? Am I still allowed to cheat on single-player games? I understand that the cloud technologies will allow us, as gamers to support good games with more consumer demand, but what about creative input from fans beside simple market forces? If I played Fallout 2 on OnLive, could I still use the Mega Mod (for instance) and have the freedom to tell a different story with the same old game? There doesn’t seem to be any incentive for publishers to add that functionality, as they can’t monetize it, so does that mean I’ll become a less creative gamer, trapped inside a less creative industry?

    3.) Working toward the holodeck: Can we get Wizards of the Coast to please please please, for the love of God, put their character generation, dungeon design, and GM software up on a service like this? The best thing about playing online is that I get to play with other people, the best thing about playing with people is that we can do stuff machines can’t. How much do I want to play a game of D&D using this. I know you can already sort of do that with a PC, but it’s a bit cumbersome.

  39. J Greely says:

    B.J., you were trying to make a point about Perlman; unfortunately, he cashed out soon after the Microsoft acquisition, so he had nothing to do with the eventual failure of the product. We tripled our subscriber base in the first few years with MS, and stopped growing because nobody was engineering an upgraded product.

    (and if you get an average of 400,000 people paying you $20/month for 10 years, that’s a nice-sized drop in the bucket…)


  40. Josh says:

    I agree with Zerai. 80 ms of latency, both ways, will not be acceptable for fast-paced FPS games. It’s fine for a lot of other kinds of games. I doubt it would be a problem for strategy or puzzle games.

    As an example of why it wouldn’t work – I play Team Fortress 2 on a pretty souped-up rig. Sometimes I have 40 ms ping times to the server I’m playing on. And there is still a problem of lag, where I have to lead people extra when shooting, or when I thought the person shot behind me… but the server says they hit me. Now, double the latency because my input has to travel to the game server first. It would be painful.

    Of course, the network has a lot of room for improvement (still), so sometime in the future even a game like TF2 will probably work fine in a cloud.

  41. Stormcaller says:

    i hate some of you people in the states sometimes…

    A F%*king awesome ping rate for me is ~150 usually…

    Generally it is around ~300, and in WoW its currently hovering at around ~450/500 on a good day… So seeing people complain about 40 is kinda weird… (also could explain why i suck at multiplayer FPS’s

  42. Stormcaller: The network lag in WoW isn’t in between your controls and your output devices, that’s the difference. It might make your enemies act jittery, but the sort of lag OnLive has will make *you* act jittery.

  43. felblood says:

    This is clearly the same slide show and speech he gives to investors and publishers, with some badly kludged add-ons for the engineers, like the feelies, and such-like.

    I have some serious doubts that the range on his gadget is as good as he thinks/says it is. There’s no way those kind of ping times are going to show up here in rural Idaho, let alone any part of Montana.

    I’d have to see the pricing scheme (which he avoided talking about) to know if I would use his service (if it could work here). Not that I have any money to give him anyway.

    It’s too bad that Dwarf Fortress will never be available through this service in this universe. Long after you need an emulator to play Crysis, DF will be making CPUs beg for mercy.

  44. Lalaland says:

    Digital Foundry “In Theory: Is this how OnLive works?”

    A very comprehensive look at this from someone with better knowledge of this than me but they distel it quite well. Basically it seems they’re delivering gameplay but 720p native with max settings on Crysis it is not . Lossy encoding and all the latency concerns make this a miss for me.

  45. eri says:

    I feel as if a lot of this presentation is smoke and mirrors; it looks very nice, has a cute little interface to show the students, nice toys to pass around, and even a hands-on demo where one of the students pwns some noobs. But it’s deceptive, and it’s intended to be deceptive; facts have been played with in order to hide problems and provide answers which sound good, but leave much to be desired. OnLive is, as it always has been, a neat idea, but there are just too many concerns for it to be as good as it’s made out to be by both its creators and the industry. Early adopters have been singing its praises, but we need to bring things back to earth and assess them in a reasonable and realistic way.

    The major issue, as has been stated, is latency. 80 ms doesn’t sound that bad, especially since a lot of offline games have latency of around 130 ms just for their basic input/output (this varies with framerate). However, when you realise that 80 ms is under optimal conditions, and comes in addition to the response time added by the actual rendering of the game, it doesn’t sound all that great. Even if you’re going very fast, try playing an online game and it looks like pings of 120 ms are going to be about as low as you can reasonably get. This is acceptable, but not ideal, especially when you have input lag as well. The presenter is quick to say 80 ms is “outside a human’s perception threshold”, but I don’t think that 200+ ms or more (depending on conditions) falls within that threshold.

    Secondly, there is the issue of bandwidth and accessibility. Internet connection speeds are, on average, simply not good enough across the continent to make OnLive work in a reliable and efficient way. While you might boast a “blazing fast” 10 Mbps connection living in downtown New York, many people make do with much, much lower speeds, and even 56 Kbps is still quite common. Recently, broadband has been defined by the US government as 768 Kbps, which is still not adequate for streaming video, and certainly not adequate for streaming videogames. Unless you live in a place with a fast, stable Internet connection, with an OnLive server relatively nearby, you will not have the same sort of experience as what was shown in that video. Let’s not even get started on what to do if you live outside of the US and/or North America, because I doubt a plan even exists to accommodate that. Same goes for language and accessibility options.

    Thirdly, there is the issue of ownership and control. Whether or not you feel as if you “own” your PC games or not by possessing a license to play them, the fact is that having the code installed on your computer means that you can modify it as you see fit. This means a number of things, but chief among them are mods and customisable game settings. Being able to use mods seems like an impossibility with OnLive, and due to the limits of server infrastructures, are players going to change in-game graphics and audio options to meet their needs? My guess is no, they won’t be able to. I also assume your save files are stored in the cloud, which means that if you stop paying, your progress is gone. The fact is that this is all like an elabourate rental procedure, and it has all the disadvantages of it.

    Lastly, there are some outstanding technical issues. How do they plan on providing the server infrastructure to support thousands if not millions of gamers at a time? This is an extremely expensive proposition, and I doubt that it will be paid for by $15 monthly subscriptions alone. Furthermore, the way games are programmed, they are not too flexible when it comes to using virtual machines. This means that intensive titles like Crysis will require dedicated systems, which means that potentially you will need even more servers with the highest-end hardware. How do we pay for this? The startup costs are huge; expect to see in-game advertising of some sort to make up for it.

    There’s also the issue of image quality. While seeing Crysis running in motion up on-screen seems all well and good, the fact is that an analysis of the game shows it is definitely not running at highest settings, and that its framerate is not as high as it could be. This is compounded by the issues of video and audio compression. I’m sorry, but while 720p video will look okay on a smaller HDTV, compression artifacts and the like will mean that next to the real thing, and on a higher-resolution display, it won’t stand up.

    On top of this, most PC monitors sold these days are capable of far more than 720p, and the lack of post-processing on the image means that a PC delivers a more accurate, and thus worse-looking picture. DVD video on my 1680×1050 display looks awful when played full screen, and even “HD 720p” looks pretty poor too. What about video framerate? Will I be able to play DiRT 2 at 60 frames per second in DX11, or will I have to settle for 30fps and DX9 due to hardware limits and video compression? I don’t know if their 5.1 audio solution will stand up either. If you compress things too much, you’ll start to get muddy and distorted sound, and I doubt they can do 44.1 Khz in “real-time”.

    Is OnLive functional? Definitely, and frankly it’s quite a technical feat. For more casual PC gamers and those who might want to get their feet wet, or just play around a little bit without much hassle, it works pretty well. The problem is that it is definitely not a replacement for current PC gaming standards; in fact, it really does not even come close in any important way: playability and performance, image and audio quality, latency, and content control are all totally inferior. The great thing about the PC market is that it scales vertically; if you have the know-how, you can always build a capable system for a reasonable price. The cost argument sounds initially compelling, but I think it’s moot. If you have no money, you probably aren’t buying or renting many games at all, and I doubt you have enough to spend on monthly fees.

    Despite all these problems, I don’t want to see OnLive fail miserably. It’s very cool and will probably fill a certain niche of the market. It probably isn’t the future of gaming – maybe it will achieve some level of mass market acceptance, but I think it will exist in parallel to our existing setup, serving a different set of desires. Of course, never underestimate how well corporations can convince people that less is more – just look at our decreasing standards of quality when it comes to audio distribution (i.e. lossy formats and poor-quality speakers/headphones dominating the industry).

    And yet, even if it is promising in some respects, I think OnLive stands a good chance of crashing and burning. Word spreads fast, and any, yes, any significant problems in the system will be felt instantly. It needs to be nearly flawless right from the beginning, and I’m just not sure, given all the above issues, that it will be.

  46. midget0nstilts says:


    Especially in regards to the virtualization question, I was wondering the same things myself. They don’t give much in the way of specifics, and I never really thought about the application of servers in the context of gaming.

    For example, how are Xbox games handled? Do they use off-the-shelf Xboxes with elaborate modifications? Did they cut a deal with Microsoft to make special Xboxes? Do they use servers with emulators? If they use actual Xboxes, wouldn’t that mean they’d need thousands of them?

    The same sort of thing applies to PC games. There’s no way they could use dedicated systems; they’d need thousands upon thousands to accommodate users. They’d have to use virtualization tech. I’ve never played Crysis, but I’m confident that servers could handle multiple instances, excluding anything to do with graphics. I don’t know a lot about graphics programming (but Shamus sure does!), but I understand that graphics cards are optimized to do many tasks in parallel, which would at least allow for the theoretical possibility of sharing a card or multiple cards in the same system for multiple instances of Crysis, provided you have cards powerful enough. (Something tells me that they wouldn’t be using GeForce’s from Best Buy.) I’ve seen games played in virtualized environments a couple of time, but there were always glitches.

    I also wish I knew what OS they’re using and what virtualization software they’re using (if any at all). If anybody has any information on any of this, I’d be curious to hear.

  47. Henebry says:

    As a Mac user, I am excited at the prospect of playing first-run games here on my Mac.

  48. Tesh says:

    I prefer buying my games and not needing permission or an internet connection to play them. (The “book” method.) This OnLive thing might be technologically possible, but it’s not useful to me at all.

  49. Willy Jewel says:

    Notto disu shittu agen

Leave a Reply

Comments are moderated and may not be posted immediately. Required fields are marked *


Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun.

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="http://en.wikipedia.org/wiki/Darth_Vader">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>