I can see the microconsole and a hard drive (as well as a mouse, keyboard, controllers, and a headset) to be all you'll ever need for hardware again. And even the hard drive is just in case you have a computer capable of running these games (in which case you'd need to be connected to OnLive in some form, while playing the game, to prevent piracy.)

I mean now power outages and internet slow downs would kill. Ouch, there's no band aid big enough.

Indeed. As I blogged about as well today (blog.terravision.no); to capture a significant piece of the more casual market, it needs to have a favorable range of titles, so the sign-up, code-modification and distribution process has to be as smooth as possible. And to capture more hardcore audiences, they'll need to be able to offer an experience that is no less than what gamers can get from their current setups.

And an offering in Europe, Asia, as well as in the US.

It comes down to cost and usability on the consumer side, and ease-of-development and profit margin on the developer side.

What happens if their network goes down? We all know that would never happen. /sarcasm

What happens when I hit my monthly Comcast bandwidth limit for the third time and they shut me off forever? Other internet companies offer much less of a limit than Comcast too.

What happens if my internet connection goes down? We also all know this never happens. /sarcasm

What happens if they take down an older game that I still like to play and want to play again? They can't tell me that that they will keep every game up and active forever, right?

"The OnLive model will appeal immensely to publishers, who will likely derive greater revenue per sale than is derived through conventional retail distribution," says Pachter."

Oh yes, taking the power of true ownership away from the consumer forever. This seems totally fair. Good for publishers(notice they don't mention the devs) and bad for the consumer.

I predict this system will crash and burn.

@Alan: The bandwidth limit is mostly a US "invention", as I can understand it. It has been briefly tried here in Europe, but was abandoned just as quickly. Hopefully, if OnLive gains momentum, it will either get Comcast and others on board by placing servers at their data centers, and/or offering "gaming broadband packages" with included free use of OnLive services.

As to your other concerns, I too feel that the ability to go back and play any old game has intrinsic value (heck, I even had to find both a diskette station and a DOS emulator a month or so back, to play my old Space Quest IV). If they don't get these things right, they may very well, as you so diplomatically put it, "crash and burn". ;-)

@Erik Harg - Australian and Canadian companies both have very strict bandwidth limits in place though. More so than most American companies. I am glad to hear the Europe has rebelled against this trend. I like your idea of "gaming broadband packages" from companies like Comcast and Cox. I wonder if they would do this and what the pricing structure would be like.

Can you imagine the amount of hardware need to render all these games running at once?

They would have to have entire datacenter full of hardware probably more. A 360 is pretty much maxed out playing alot of the new titles in HD. They will basically need the equivalent hardware as a 360 for each user that is subscribed to the service.

Why is nobody bringing this major issue up?

@joe

I think because we all presume that they have a solution. If they're running server farms then they it's not so much of a stretch. In the near future we'll be looking at upward of 16 cores on a single server CPU, not to mention triple and quad SLI set-ups, extreme dynamic memory allocation, petabytes of storage, etc. etc. They already use special petaflop and petabyte super-server configurations in the information systems world. It's not an impossible dream on the hardware side. The real limitations are going to be bandwidth (as mentioned above) and latency. If they conquer these demons, provide a varied library, and keep the cost down they could find a decent market share. I'm not ready to write off physical media, game consoles, and gaming PCs just yet; but this tech could have a major future.

The concept of a having the bandwidth for a server to send over an image to a thin client I think is what is truly interesting here. If the execution of this concept makes it more of a mainstay in the consumer dictionary, then the concept itself could become a market. Instead of publishers you'll have channels that offer different content that can be streamed to thin clients.

I keep telling to myself that I hate it from a gamer's point of view (I am from Argentina, this system may never hit here because the fastest speeds we get are in the 10mb for around USD 200 per month). But from a programmer's point of view, it is the solution to all your problems. Of course, there are a lot of ifs there (can they provide the horsepower? can they encode that fast? can they take work out latency? can they handle the bandwidth?) but if those are all addressed, men, here is the future.

Yes, the system goes down, your games go down too. The same as if Apple ever decides to close iTunes, or if Amazon kills Kindle, or if Microsoft kills Hotmail, or if Netflix kills their video delivery. A lot of teaching must be done (you don't "own" content, you "license" it). But once that is accepted (and the mass market needs to accept it, not the hardcore people who will never like that), it is game over for the retail system.

Roberto Alfonso - "A lot of teaching must be done (you don't "own" content, you "license" it). But once that is accepted (and the mass market needs to accept it, not the hardcore people who will never like that), it is game over for the retail system."

I would be so pissed off if they stop selling games at retail. In fact, I would beyond pissed off. "License" games? Forget that. Next they will want me to "License" movies too. I will literally stop playing games if that happens.

Go Sony and Microsoft. Please help kill this insane idea.

I would too. But think as a publisher and not as a gamer. What makes sense from a publisher point of view? Imagine this: developers don't need to scale down productions because the necessary equipment isn't still adopted widely. They just code and tell this company "Look, you need a IBM mainframe with a tera RAM", they get it, and blam, end users can play it. And publishers don't need to worry about the second hand market (hey, I am one of those that think why the second market for games should be different from the one for books or CDs or videos, the IP owners don't get retribution for resales, nor the gaming industry should). No piracy: you don't even need to use DRM solutions, you deliver the gold copy to OnLive, they install it in the necessary computers, and done. They don't need registering, they don't need key checking, nothing. If the gold copy appears in a BitTorrent site, sue OnLive.

In technology, almost everything is done in a server/client paradigm. Banks are just a lot of dumb terminals connected to a server, supermarkets have several dumb terminals connected to a server, ATMs are just dumb terminals connected to a server, World of Warcraft are just dumb terminals connected to a server, MUDs are dumb terminals connected to a server, internet are basically dumb terminals connected to a cloud... why gaming should be different?

By the way, the bandwidth cap is silly. Not even down here there is one. Telecom, one of the Tel companies here, tried to implement it but got rejected by users and they dismissed that. Are you telling me users in a third world country can force a company not to implement it, and users in first world countries just accept it?

Roberto Alfonso - From the Publisher point of view I agree that all your points make sense. But as you agree that does not make any of it right.

As for bandwidth caps, all most every company that provides broadband internet service in America, Canada, and Australia has one. Comcast (my provider) has a 250 GB per month limit. If I go over that limit three times I get cut off for life, or till they tell me I can come back.

That bandwidth limit is very generous as well. Other companies like Cox and others have limits as low as 20 GB's per month. I can tell you that the subject of caps is not silly at all. It is a new fact of life for us.

Even if this never takes off in America (though after playing a demo earlier, I don't see why it wouldn't), this instantly gives publishers a business model in Asia. No more easily pirated physical disc--force them to subscribe. I think it's huge.

This does have potential - perhaps not right now, but in 5-10 years time, as the technology matures.

The best way forward for OnLive is almost certainly to get major triple-play media companies (i.e. phone/broadband/video) to licence this technology and incorporate it into their settop boxes while also installing the requisite "mainframe" to service said boxes.

We'd then be in a world where one single box can play virtually anything - and there's no way to pirate the games or trade them in. Absolute heaven for the publishers and developers (in theory at least - no-one has ever been able to quantify the positive aspects of piracy/second hand trade in terms of free-publicity/upsell terms), though there are several issues for the consumer as various people have outlined above.

Still: no need for hardware upgrades. No need to configure anything. No more need for proprietary hardware - Sony, Nintendo and Microsoft had better beware! The ability to log in and play your games anywhere - think Steam, but without the need to download or install anything.

How likely this is, I don't know: it's all based on the premise of free, low-latency, high bandwidth connectivity - and this requirement has scuppered a great many previous thin-client attempts. There's also a question floating around on how performant the server needs to be - as far as I'm aware, there isn't currently a way to virtualise GPU technology (though Microsoft's work on a software DX engine may help to a degree there), so there's likely to be a hard physical limit on how many simultaneous sessions a single machine can handle.

Give it 5 years. Let the average broadband speed rise. Get GPU technology integrated into the CPU. Get OneView integrated into the media companies. Then things will start to look a lot more interesting...