Daydream Believer: OnLive

The OnLive Micro-Console

It's interesting how quickly countervailing opinions become prevailing opinions these days. Don't get me wrong, I'm glad that the Internet is so good at challenging every belief I would care to share (and a few that I should probably keep to myself). It's good to question your beliefs and ideas, and it's good to have voices that tell me not to pin my hopes, dreams and livelihood to a falling star. I'm sure my wife appreciates all the naysayers who've stopped me from chasing daydreams that would lead me away from my fairly secure job with steady paycheck and reliable insurance benefits. But few dreams become reality on accident, and well entrenched cynicism will never stop me from daydreaming.

A lot of ears pricked up when companies started murmuring about streaming high-res games over the internet. Wild, crazy rumors about a world where we wouldn't need to struggle to keep pace in the PC-hardware arms race. A new world where we'd almost automatically keep up with both new system requirements and the Joneses. Next thing, I was thinking about jet-pack commutes and sex-worker AIs that downloaded right into your high-tech pants at the blink of an eye.

But I can count on the Internet to challenge those daydreams. Maybe OnLive's game-streaming service wouldn't be quite as shiny without the studio lighting and lens effects. Maybe that jet pack would turn out to be a gas-guzzler, or the AI sexbot would download and start up while I was in the middle of weaving through jet-pack traffic, and even then it'd be all pixelated and awkward. But a jet pack would still be pretty cool, even if it wouldn't get a reasonable price point and the kinks worked out until iJet 3.0. And maybe, like my purchase of the original Kindle, buying iJet 1.0 could be a small vote of confidence in the potential of the technology.

When the big doubts started popping up about OnLive, I went looking for the promise of iJet 3.0. And boy, does the future look neat.

Looking at OnLive's LinkedIn page and "team" page on their own website, there's a lot of talent both in terms of networking experts and in spreading this technology to broader applications. See that VP of operations? Not only does he have a sweet mustache, but he's also got 16 years of experience delivering programming for NBC. Another VP isn't just the VP of games, he's the "VP of games and media." They're staffed with expertise to expand beyond games. OnLive's service is built around pushing lots of data very quickly. If the massive pipes OnLive needs are able to stay funded, it seems likely that we'll see a network capable of bringing a whole new level of speed and ease to streaming content of all sorts.

Currently, gamers can download games via digital distributors like Steam or GamersGate, from online stores like Amazon, or even from brick-n-mortar stores, but we're responsible for making sure we have the appropriate hardware. We can save money by renting titles that we know we won't be interested in for very long, but our options are restricted to local stores with short supplies or to the often-spotty service from online distributors and their baffling queue logics. I'm daydreaming of a system that combines the best of both worlds. What if game rental were as easy as renting a video over XBL? Drop a few bucks, and the game's ready to play for X hours or for the next Y days. No hoping you worked the right queue mojo or that your local Blockbuster accidentally stocked 100 copies of Gears of Warfare. No upgrading your drivers. It's almost as appealing as robot sex.

But dreams seldom come true on accident--at least not the good ones. Big possibilities don't mean much if the company can't get off the ground in the first place. As gamers, we've all been burned before by some promising new tech that didn't roll out effectively. We've been left with little more than a cute inside joke when friends notice the pristine Atari Lynx mounted on the bookshelf. So it makes sense that we'd be cautious before throwing our money at the folks behind the OnLive booth at GDC.

I'm not as worried about bandwidth as many seem to be. There's already plenty of surplus bandwidth capacity along the main corridors of the US intranet infrastructure, waiting to be used. As "luck" would have it, the three server centers for OnLive at San Francisco, Dallas and Washington D.C. happen to also be at key locations in the broader internet infrastructure. It also happens that OnLive is strongly hinting that AT&T, a major source of capital so far, will be a major partner in the launch.

But it's hard to be certain of what pricing is going to look like. So far, they've announced two price points. The first announced point was a more premium-priced structure built around a $15/month subscription and titled the "OnLive Game Service." That comes with a number of things that might not be hugely appealing (e.g., "Brag Clips™"), some bits that are moderately interesting (multiplayer across multiple platforms and massive spectating), and a few things that are downright appealing, like cloud-based saving that can be accessed from other platforms through your account.

I've heard both sides on this pricing scheme, and I figure it could work out well for some. At $15 each month for the subscription, it could work out well for some. Say you spend $1,000 every 4 years on a new gaming rig? That's $720 in OnLive subscriptions during the same period. But being charged a monthly fee while still having to pay for rentals and game purchases seems off for a lot of gamers. We can easily understand paying for hardware and then paying for software, but paying for a subscription and then for games doesn't partition as easily.

And I was starting to wonder: Why aren't they leading into the OnLive service with enough of a discount to make sure they gain market traction (i.e., as a loss-leader)? Maybe I've been reading too much Chris Anderson, but between that and being used to the loss-leading strategies of console companies, I found OnLive's angle a bit puzzling.

Which is why it was a relief to see the announcement of the free, a la carte model during GDC last week. It doesn't look like it'll offer cloud saves, and this "OnLive Game Portal" wouldn't offer as broad a selection of games or new releases as quickly as the subscription service, but it lacks the up-front price that makes me suspicious before I even know how much game purchases and rentals are going to cost. If I can sign up for this service free of charge, I'm more than happy to help OnLive brag about a larger user base next time they're talking to investors. It sure costs less than investing in the Kindle or preordering an iJet.

OnLive is titrating information carefully between now and the June 17 (E3) launch date, so it's too early to know exactly what we'll be asked to sign up for this summer. There's plenty of time to tweak the announced pricing models, add details, or announce entirely new price options. But my sense is that OnLive's products are likely going to fall into a cozy hybrid of GameFly's subscription model and the immediate satisfaction of a brick-n-mortar rental shop like Blockbuster. And if they can get a critical mass of subscribers, I can see great things coming for online games and entertainment.

Comments

Seems like you'd still need a decent amount of processing power to decompress 720p30. If their compression algorithm is good enough to maintain a reasonable quality at a few Mbs, then I would guess it's going to take something more than a min. spec PC to decompress it in real time.

Switchbreak wrote:
Quintin_Stone wrote:
TheCounselor wrote:

Correct me if I'm wrong, but On-Live is supposed to be relatively light on the bandwidth usage side of things. Thus, it's not like you'll be downloading whole games. You're just getting compressed video.

Just? HD video takes up a lot of bandwidth and to start the games on OnLive will only be available in HD. They say it'll be possible because of their revolutionary compression scheme. But light on the bandwidth side it will not be. And that bandwidth usage will be continual and steady for as long as you're playing the game.

Anyone signing up better hope their ISP doesn't have one of those secret data usage caps.

This is where I don't buy what they're selling. Real-time 720P video at 30FPS with no buffering and with playably low latency doesn't seem even remotely doable on current internet connections. The fact that they've done nothing yet to demonstrate their supposed fix for this dilemma (and no, demos running over a LAN at a convention don't count) makes me doubt that they actually possess these deep magics. I'm thinking that the promise of 720P/30FPS will hold for people who live right next door to their servers, and in the average use case the "compression algorithm" will be a fancy word for delivering video at a dramatically lower resolution and probably lower frame rate.

What if it does come down to a massive LAN, with, say, local nodes tying into that private network throughout this map?

IMAGE(http://olwriter.files.wordpress.com/2010/01/ol_coverage1.jpg)

Gravey wrote:

Ubisoft DRM: bad, always-on requirement is nonsensical
OnLive: I gotta file this with the Ubi DRM for now.

I once again agree with Gravey 100%.

I'm thinking of a city in Poland and an even number. Do your worst, mind thief.

Heard some good things about it. But I will join the herd once this product has proven itself.

Its interesting how OnLive is being compared to PCs when in reality it is more of a console. From what I have read it will not support keyboard and mouse for controls but rather use a controller. So if you are comparing a subscription of $720 per year for a console, that is a $200 XBox and 8 games every year.

I have been hearing rumors that the $15 a month is only for the service and you still have to pay to play the games. If it is a service where you can play any game that you like for $15 a month, then it would be appealing since I would be able to try every game that is released and only play the ones I want. I buy at least 5 games a year that I only play 2 hours worth.

wordsmythe wrote:

OnLive is titrating information carefully between now and the June 17 (E3) launch date, so it's too early to know exactly what we'll be asked to sign up for this summer

"Tit-rating?", my brain said, "Now they're talking my language".

kazar wrote:

I have been hearing rumors that the $15 a month is only for the service and you still have to pay to play the games.

That's not a rumor.

kazar wrote:

Its interesting how OnLive is being compared to PCs when in reality it is more of a console. From what I have read it will not support keyboard and mouse for controls but rather use a controller. So if you are comparing a subscription of $720 per year for a console, that is a $200 XBox and 8 games every year.

I think you were reading about the "microconsole" version. That's what you use if you don't even have a PC. The reason OnLive in general is being compared to PCs is because the service runs on PCs and plays PC games. And uses PC controls.

I'm not a fan of renting, so I hope any game on onlive has the usual Steam counterpart (as long as it's not Ubi).

Otherwise, I hope this onlive crashes and burns. Which it will probably do.

kazar wrote:

So if you are comparing a subscription of $720 per year for a console, that is a $200 XBox and 8 games every year.

It's not $720/year it's $720/4 years or $180/year or 3 full priced console games/year.

Now you're right you will have to add on rental fees, but assuming they're reasonable, it should be cheaper than buying new console games, and certainly cheaper than buying hardware and new PC games.

I don't know that I'm sold on the product, but I like the concept. Particularly the idea of "renting" PC games and also play from anywhere kinds of stuff.

paketep wrote:

I'm not a fan of renting, so I hope any game on onlive has the usual Steam counterpart (as long as it's not Ubi).

Otherwise, I hope this onlive crashes and burns. Which it will probably do.

Just because something doesn't interest you doesn't mean you should want it to crash and burn. You can not like something or even dislike something without wanting to see it fail.

kaostheory wrote:
paketep wrote:

I'm not a fan of renting, so I hope any game on onlive has the usual Steam counterpart (as long as it's not Ubi).

Otherwise, I hope this onlive crashes and burns. Which it will probably do.

Just because something doesn't interest you doesn't mean you should want it to crash and burn. You can not like something or even dislike something without wanting to see it fail.

No I agree, I hope it crashes and burns. I think the whole "Cloud can include personal hardware" thing is a horrible direction to move towards. If this takes off, I see the entire future of personal computing to be one where everything works like a console and you rent your hardware requirements. Renting is always a bad investment.

Maybe this is for a pc gamer who doesnt mind Ubisofts latest DRM solution. In some ways, this takes it to the next level. Ubisofts DRM solution allows you to install the bulk of the content locally, and you play by streaming tiny amounts of authorization content. On live has no locally installed content and you are now fully dependent on a live connection for the content / input streaming.

One business model includes a one-time all you can play type (rental) fee in the case of AC2, where this will be subscription based.

It seems quite similar in spirit. It gets more complex than that client -> server versus cloud, where is content stored etc, but.... they seem to be different points along the same line of evolution.

Switchbreak wrote:

This is where I don't buy what they're selling. Real-time 720P video at 30FPS with no buffering and with playably low latency doesn't seem even remotely doable on current internet connections. The fact that they've done nothing yet to demonstrate their supposed fix for this dilemma (and no, demos running over a LAN at a convention don't count) makes me doubt that they actually possess these deep magics. I'm thinking that the promise of 720P/30FPS will hold for people who live right next door to their servers, and in the average use case the "compression algorithm" will be a fancy word for delivering video at a dramatically lower resolution and probably lower frame rate.

I will buy the idea of actually playing an action oriented game on this kind of thing when I actually try it - from afar it seems like a pretty dubious prospect.

I feel the same way. It's not that I think this stuff isn't possible, it's just that it's not possible right now. It's a bit of a paradox since OnLive might be the stepping stone to something greater (and therefore necessary for that greater thing's development - See Law of Accelerating Returns, et al.) but everything that OnLive has been suggesting just isn't going to fly. I don't want it to crash and burn, but when I put on my let's-be-realistic hat, that's all I can see it doing.

Nevin73 wrote:
kaostheory wrote:
paketep wrote:

I'm not a fan of renting, so I hope any game on onlive has the usual Steam counterpart (as long as it's not Ubi).

Otherwise, I hope this onlive crashes and burns. Which it will probably do.

Just because something doesn't interest you doesn't mean you should want it to crash and burn. You can not like something or even dislike something without wanting to see it fail.

No I agree, I hope it crashes and burns. I think the whole "Cloud can include personal hardware" thing is a horrible direction to move towards. If this takes off, I see the entire future of personal computing to be one where everything works like a console and you rent your hardware requirements. Renting is always a bad investment.

I'm not sure I dig what you're shoveling here. You think buying a hard copy of a PC game is an investment? You're threatened enough by the model that it makes you "hope it crashes and burns"? Maybe I'm wearing the Pollyanna hat here, but I just don't get it.

No I'm looking longer term. I'm thinking about the future implications if a service like this should work.

I don't see buying a PC game or even a PC as an investment. But I do believe that over the long haul, owning my own hardware makes more sense, and costs less, than renting someone else's computer. Microsoft already wants us to do this with "Metered Pay-As-You-Go Computing"http://www.faqs.org/patents/app/20080319910. With Apple's closed systems being so popular, it seems like the next logical step for them as well. OnLive is simply the game version of this.

So say I go with OnLive. I now have a computer to do computer-like things on at home (browsing, finances, office work, etc.). Because I don't like laptops for regular computing, let's say I spend $500 on a cheap system. I'm also now paying $180/year for a separate game system (not including anything extra I would need to use it like display or controls). Assuming the useful life a computer for me is three years, I'm at $1040 for my tool and my hobby (ignoring the cost of games). Could I get three years of good gaming life out of a $1000 computer that I put together? Possibly, probably for those who really tweak their systems. Would it be worth even more to me if I could own the damn thing? Definitely.

And this is, of course, ignoring the death of modding, of which many of us enjoy using to give legs to a game.

I fail to see this really taking off unless they figure out how to transmit data at faster than light speeds, otherwise it will really only be usable by people who are close enough to get the data in less than 60-70ms. Unless they plan on building enough servers around the country, but that would be a crazy enormous investment (maybe that's what AT&T is for?). And beyond that, 720p/30fps is TERRIBLE for modern games, especially if they're thinking they will get PC gamers interested. Any high speed FPS like MW2 just would not be playable at less than 60fps, especially if there's noticeable lag, which, if you take Internet latency + display lag + controller lag, there almost certainly would be. And from what I've seen, the 720p through OnLive is not as sharp as 720p locally.

So yeah, I'm skeptical.

Chairman_Mao wrote:

I fail to see this really taking off unless they figure out how to transmit data at faster than light speeds, otherwise it will really only be usable by people who are close enough to get the data in less than 60-70ms. Unless they plan on building enough servers around the country, but that would be a crazy enormous investment (maybe that's what AT&T is for?). And beyond that, 720p/30fps is TERRIBLE for modern games {snip}

So yeah, I'm skeptical.

Me too. I just don't see this being an acceptable substitute for PC gaming. So few games (other than MMOs) are PC-only these days. Not to mention that fewer and fewer console games are console-exclusive. I guess if you really wanted to play GoW3 but cant afford to pay $300 for a PS3 + $60 for the game, it would be nice to pay $15 for one month of their service to play it on your laptop or desktop PC.

Plus, I have no interest in playing games at 720p (1280×720) on my 21" monitor.

Dysplastic wrote:

I really see this as a proof of concept that they're trying to sell to a major cable company. OnLive would work best as a feature you could get along with your regular cable bill, where you're already paying a monthly fee, for on-demand games that you can play without having to own a console or go to the game store/rental place. In this regard, it could be a huge success.

I would be more likely to go for this type of model over what they're actually offering. In fact, I hope this is what OnLive eventualy becomes, if it survives long enough to, or what another company does after/if OnLive fails.

I love the concept of Onlive and I can dream but success/failure aside I really do like that there is companies willing to take the jump to different distribution/consumption models. I do wonder about audience though. Will Onlive be more successful with the hardcore or casual crowd? Is this the Netflicks of gaming? (Canada still doesn't have Netflicks but thats a different story) The more casual audience isn't as much into this connected device ecosystem. Pitching the idea of Onlive to this subsection of consumers will have to be simple: Turn on and play, with minimal bandwidth issues etc. Most just wont take to this in place of consoles that aren't dependent 24 hour access to the internet (I bet Ubisoft is loving Onlive). For perspective, a couple friends and my wife and I were just discussing with how they were debating getting a laptop computer, they are only 5 or 6 years older then us... my word do we ever live in different worlds... my home has 2 or 3 computers on at all times plus who knows how many connected devices...

I do hope this works out, and that 2.0, 3.0, etc really bring something to the table.

wordsmythe wrote:
Switchbreak wrote:

This is where I don't buy what they're selling. Real-time 720P video at 30FPS with no buffering and with playably low latency doesn't seem even remotely doable on current internet connections. The fact that they've done nothing yet to demonstrate their supposed fix for this dilemma (and no, demos running over a LAN at a convention don't count) makes me doubt that they actually possess these deep magics. I'm thinking that the promise of 720P/30FPS will hold for people who live right next door to their servers, and in the average use case the "compression algorithm" will be a fancy word for delivering video at a dramatically lower resolution and probably lower frame rate.

What if it does come down to a massive LAN, with, say, local nodes tying into that private network throughout this map?

IMAGE(http://olwriter.files.wordpress.com/2010/01/ol_coverage1.jpg)

This won't fix switchbreak's issue, unless OnLive is planning on laying fiber out to each customer. The "LAN" here is basically a marketing ploy-consumers will still be connecting through their existing ISP's to the OnLive servers, wherever they are located.

Also, it's not just the speed of light that's an issue here. It's also the routers and switches in the way between the consumer and the OnLive servers, each of which adds a non-trivial amount of latency to the connection. Even then, as it is currently, users in Chicago will most likely see a .1 second input delay to either the DC or Dallas servers(assuming that my normal ~50 ms TF2 ping is correct for a Dallas server) BEFORE rendering time is taken into account. That's unacceptable for any action game.

I'd hope there would be minimal lag.

I believe it was on Ars Technica that a user brought up the single point which brings Onlive into perspective: They say they have an amazing compression algorithm better than anything currently available, but the application they will use it with is gaming?!

When said like that the entire thing seems infeasible.

Spaz wrote:

I think this will be immensely useful for people with, like, actual jobs who only dabble in games because they can't commit.

Perfect example: [stuff]

Something like OnLive would be a godsend for him, as he's only got a passing interest in a handful of games. I think, ultimately, this is what the bulk of the market will be.

I sort of doubt that someone who's not all that into games will be willing to commit to a monthly bill to play games. Particularly a monthly bill that doesn't actually allow him to play games, but just allows him access to the service that will allow him to buy games.

In my experience, the sort of gamer you're describing finds one game that they're interested in every five years or so, buys that game (which is to say, pays a one-time fee and can then play as much as they want) and plays the heck out of it for a while, then forgets about gaming for another five years until they happen to notice something else that interests them. Maybe they go back to their favored game once or twice in that "down" period.

But what I don't see them doing is continuing to drop fifteen bucks every month just in case they feel like playing their one game that month. I see them paying their one-time price and then knowing that their game will always be there for them when they want it.

While I think that OnLive's concept could have potential when broadband catches up to it, the business model they're describing finds itself between a rock and a hard place: the sort of light gamer that values simplicity above all else isn't likely to go in for a subscription model, and the sort of hardcore gamer that frequents a site like this is alread well-served by their consoles, PCs, or both.

hbi2k wrote:
Spaz wrote:

I think this will be immensely useful for people with, like, actual jobs who only dabble in games because they can't commit.

Perfect example: [stuff]

Something like OnLive would be a godsend for him, as he's only got a passing interest in a handful of games. I think, ultimately, this is what the bulk of the market will be.

I sort of doubt that someone who's not all that into games will be willing to commit to a monthly bill to play games. Particularly a monthly bill that doesn't actually allow him to play games, but just allows him access to the service that will allow him to buy games.

Shut up before Elysium starts talking about WoW again!

Quintin_Stone wrote:

Anyone signing up better hope their ISP doesn't have one of those secret data usage caps.

Comcast's cap isn't even secret. I get 250 gigs per month. At this point, I don't even go near the cap but if I started using a service like On Live I'm sure I would blow through it easily.

Don't get me wrong, I think On Live is a great idea. It's going to get killed by ISPs who say they are putting in the caps because they are afraid of network load. In reality, network streaming from sites like netflix have pissed them off. They want you to get your streaming content from them and not netflix.

I feel really bad for On Live. I know that they have been working on this project for a while now. When they started it there were no caps. Isp greed is going to kill them. The best that they can hope for is to recoup their losses and sell the tech to one of the cable companies.

On Live seems like it would have even greater immediate potential for commercial channels like hotels as opposed to consumers. They could offer cutting edge games on a pay for play basis from the guest rooms. (Though maybe longer games don't make sense.... unless they offered hotel exclusive versions / experiences).

Irongut wrote:

On Live seems like it would have even greater immediate potential for commercial channels like hotels as opposed to consumers. They could offer cutting edge games on a pay for play basis from the guest rooms. (Though maybe longer games don't make sense.... unless they offered hotel exclusive versions / experiences).

I certainly hope they're negotiating that.

Indignant wrote:

I feel really bad for On Live. I know that they have been working on this project for a while now. When they started it there were no caps. Isp greed is going to kill them. The best that they can hope for is to recoup their losses and sell the tech to one of the cable companies.

I don't feel bad for them at all. They've decided to charge forward with a business plan chock full of holes big enough to fly a 747 through it. I understand the theoretical appeal of this sort of thing, but there's nothing to even vaguely suggest it will work from either a technical or a business standpoint. It's just the Phantom with a different box and shinier press releases.

Excellent article Wordsmythe. I've tried to keep myself on the optimistic side of this technology even though it doesn't really apply to me - being a hardcore techie/gamer who loves building and tweaking the latest and greatest hardware. The internet is rife with naysayers on this technology - some with very eloquent reasons why it will fail and why it shouldn't even be attempted. Perhaps they should be prepared to eat crow.

I've seen many things accomplished in the technology field that were said to have been impossible. Now I look forward to seeing it once again. Certainly the technology will not be perfect at first, but we are looking at the pioneering effort that may drastically change how we play some, if not all games in the near future.

Give them a chance - let them fail if the service isn't up to par. People will vote with their wallets. But at least give them the chance I say.

hbi2k wrote:

I sort of doubt that someone who's not all that into games will be willing to commit to a monthly bill to play games. Particularly a monthly bill that doesn't actually allow him to play games, but just allows him access to the service that will allow him to buy games.

The monthly bill thing is a major pill, but consider that (much like a GameFly account) there should be the ability to drop the subscription when the game-crazy wears off. I agree that there's absolutely no reason for the kind of person I described to carry a year-long sub. to OnLive, and even less of a chance that he'll venture out and try GoW3 or something.

But for someone that's REALLY just interested in one game? I think it could work.
There's also the slightly-gimped free option, which is nebulous but would liberate this kind of gamer from the burden of subscription costs.

In my experience, the sort of gamer you're describing finds one game that they're interested in every five years or so, buys that game (which is to say, pays a one-time fee and can then play as much as they want) and plays the heck out of it for a while, then forgets about gaming for another five years until they happen to notice something else that interests them.

This assumes that the person in question has sufficient hardware to be able to go out and buy the game in question. A lot of people don't. They're on Lonerware from work, have old PSUs or Motherboards, or can't be assed to buy a whole new PC on the basis of Mass Effect 2 coming out. They're even less inclined to buy a gaming console, since all you can really do with that is play games (which they're not that interested in) or play DVDs.

I really don't see such a disinterested gamer plunking down the $300 cost of hardware every 5ish years because one game interested them.

I see them paying their one-time price and then knowing that their game will always be there for them when they want it.

Well, that's what the person I was speaking of thought. He bought the PC version of the game and completely ignored DX requirements, the age of his hardware, etc. He just thought it would work, and as a result, he's got a $40 coaster that won't play in any of the PCs he owns.

So he can either pony up the cash for new gear, or he could try this OnLive thing and see if that's an adequate solution to his HW problem.

Just to make it clear, I'm advocating for OnLive as a kind of glorified Netflix (subscribe when you want one or two games, drop it when you're done) whose main strength is liberating players from hardware.

*in a perfect, lag-free world, btw.