Cloud-based gaming

http://www.gamasutra.com/php-bin/news_index.php?story=22875

If this works as they say it will, it could potentially usher in the end of gaming consoles, Steam, etc. as we know them.


It's a bit scary, frankly.

I'll believe it when I see it - what they're offering (eliminating latency and bypassing hardware requirements) sounds too good to be true. You'll still have to buy/rent access to the games you want to play, which should translate into a steady income stream for the publishers while reducing the volume of stock they'd have to carry and bandwidth requirements for downloads - a win/win situation for them. Steam - as we know it - could soon be history, though I'm sure they'd set about re-purposing their servers if OnLive started to look like serious competition.

I know a lot of people would rather have physical copies of their games instead playing them online.

I think this will be a nice service for people who don't have very good computers, but want to play the newest computer games.

We have developed an incredible new technology... Let's use it for games.

One word: Mods?

It's definitely an interesting development, but I don't think it's going to kill off physical games any time soon. That said, it's a logical evolution considering where hardware trends seem to be going. I no longer own any computers with a dedicated GPU, so honestly this seems interesting to me, but probably something I'd pass on for the long run, especially if games are "rented" or there's a "subscription." However a Steam-like system for "owning" games, I might buy into. Ultimately they'll do whatever nets them a bigger profit.

Another concern I have about this is the limited nature of modding possible. Everyone here knows what I mean. Independent, small scale mods would be much less likely through this system as they'd have to go through the parent company, which is a huge drawback. But honestly, I'd love to play Valve games on this little 1.6ghz laptop. It would make LANing much more feasible and inject a lot of fun into my late night drunken video game sessions with friends, as we all have relatively low power laptops, and our options with two 360s are pretty limited.

Yep, you guys beat me to the mod thing. Just look at Quake Live.

I don't see why it has to be an either/or situation. I really see room for both ideas on the market. You've got people who have slower computers who might want to play the latest games, and you've got people who want to own their games and do modding and stuff.

Personally, I'm excited if it actually works - I'm sick of missing out on new games because my computer can't handle them.

Hmmm, I wonder how this will work on my internet connection where I can't even stream a couple of minutes of YouTube video without it stopping half a dozen times to buffer.

Enjay said:

Hmmm, I wonder how this will work on my internet connection where I can't even stream a couple of minutes of YouTube video without it stopping half a dozen times to buffer.

That's something I'm still wondering about myself. Now the article says they've worked to eliminate flaws in internet connections by working with the cable and phone companies directly, but isn't a shoddy connection still a shoddy connection? Perhaps they can't cater to everyone who doesn't have a solid internet service....

I have a shitty computer and always have, but I still don't see this as exciting. Like people were saying before... First of all, you wouldn't be able to pull off mods very well if at all. Secondly, I like physical copies of all my games TYVM. Third, losing your internet connection will suddenly render you incapable of playing any games if you go fully to this model. Whoops.

My thoughts? Anyone ever heard of netflix?

The way it works is that it uses some hyper-advanced (arrrgh, sarcasm) technology to determine your "video quality" based on the speed of your internet connection, in order to find the best way to stream content withput interruption. I imagine something like this would have to be implemented. Not all internet connections are created equal. I see great potential for this.

I am curious.

There will always be a place for physical copies of games, I think, or at least for natively-run games bought through services like Steam, as there will always be computers that can run games via those methods but not this one and vice versa. Publishers will have to sell games in multiple ways to reach the largest possible market, so we don't need to worry about this obsoleting native gaming and taking away the ability to make and run mods. This is an interesting development and a positive one.

I probably wouldn't use this service (even if it's made available up here) but if I did, I think I'd actually like it best if games were rented. There are some games I would want to play often and play mods for, and then there are games where I'd just like to play through the official campaign once over a weekend without paying $60 for the privilege.

Kirby said:

That's something I'm still wondering about myself. Now the article says they've worked to eliminate flaws in internet connections by working with the cable and phone companies directly, but isn't a shoddy connection still a shoddy connection? Perhaps they can't cater to everyone who doesn't have a solid internet service....

Ten bucks says the cable and phone companies are all American, so it wouldn't help the other 6.4 billion people in the world. ;)

Anyway, performance would cause a lot of different kinds of issues. How are they going to be able to serve all the users without latency at the same time? They are apparently dreaming to have millions of users, probably at the same time too. Not to mention all the burden that would cause on the network infrastructure, working with the providers or not. Throw in everyone "soon" watching IPTV, listening to streaming music and watching video on demand and you've got one dead Internet (with the current infrastructure). :P

Patrick Pineda said:

The way it works is that it uses some hyper-advanced (arrrgh, sarcasm) technology to determine your "video quality" based on the speed of your internet connection, in order to find the best way to stream content withput interruption.

Those kind of sites have a poor record of success with me so far.

Jodwin said:

Ten bucks says the cable and phone companies are all American, so it wouldn't help the other 6.4 billion people in the world. ;)

Agreed. That would explain why they're only aiming "to hit 200 million people that Steam can't sell to".

The systems likely to be dependant on customers being able to afford connections that can stream low-definition video or better - which would rule out most of Australia's online community unless they're prepared to accept seriously degraded video.

ISP's are going to have a raging fit over this if it takes off (which I doubt).

My internet connection sucks too hard for this to be feasible for me. I just imagine that running a game physically on one's computer is a lot more stable and less subject to server errors, loss of connection, etc. then just having the game would be. Both of my computers are fairly old (6 and 14 years) and can't play most games, but in all honesty I don't usually like modern games that much anyway, and the games I do like I can usually get on our 360.

Plus the modding issue, which would make the more hardcore gamers who like making and playing mods avoid this system and stick to owning the games.

I dont think this thing is going to become too popular. (hopefully)

r_rr said:

I dont think this thing is going to become too popular. (hopefully)

Yeah, it won't be too popular UNTIL GRID connection system will be used massively (read: it's 10000 times faster than US connections, you can bake eggs on it's cables!). HEY LET'S WAIT 10-15 YEARS FOR THAT

So practically it's the old dumb terminal and super-server architecture reproposed, something which has already been beaten to death with "distributed computing", "internet applications" and then some.

I really don't know how they are hoping to "eliminate" latency: at most it will change its nature.

big mouthed words"Not only have we solved the problem of compressing the video games, we've solved the latency problem," Perlman said to Gamasutra.

"Solved" is just too much of a preposterous, big word here. If they "solved" it so effectively, then that should be their primary product. For it will mean that they must have some revolutionary, highly advanced communications technology is at least 20 years more advanced than anything else.

"We knew, in order to make this thing work, we'd have to figure out a way to get video to run compressed over consumer connections with effectively no latency."

Video can already be streamed "over consumer connections", if you consider dedicated analog/digital cable TV and TCP/IP based protocols.

Of all these, only the analog TV signal can claim to have anyhing close to "zero latency", since it displays signals as soon as they arrive with no buffering, and displaying a full TV frame in PAL requires 1/25th of a second or 20ms.

To come even close to that performance, any digital/TCP/IP based method must be able to transmit and decompress a whole frame in less than 20ms, which is close to the internal lag of a large LAN network not using exactly top-notch equipment, and can only become worse when going WAN and through telcos.

Our video compression technology has one millisecond in latency -- basically no latency at all. All the latency is just for the transport, and we've also addressed that."

Anyone having used digital TV in either satellite or terrestrial/cable digital form, knows that before any content is even shown, the set-top box must buffer several seconds worth of content, and that it's practically impossible to sync two separate digital receivers to the same frame (forget store windows with 50 TVs tuned to the same channel), or have any sort of hard real-time display on it.

Sure, you could start displaying stuff with close to no buffering if the transmission speed grossly exceeded (by two orders of magnitude, let's say) all actual technologies, so that a retry would likely be successful in under 20ms.

Anyway, the whole thing sounds like the next Daiskatana.

P.S.:

Maes said:For it will mean that they must have some revolutionary, highly advanced communications technology is at least 20 years more advanced than anything else.

How could I forget...this wonderful technology already exists! Skeptics and "close minded" alike, don't read it, for you have been "brainwashed" by "official science" and the such.

I'm sorry, I look at the title of the article and I can think of is this.

While this does indeed sound revolutionary and dare-I-say-it interesting, I'm with GreyGhost - I'll believe it when I see it.

What I see as a drawback is more centralization or concentration in who provides such gaming services.

Jodwin said:
One word: Mods?

If this sort of thing really takes off and accumulates potential I'm guessing they'll introduce modifiable elements in the games so people can mod them in some respects, to compete with that advantage in traditional games.

TheDarkArchon said:
ISP's are going to have a raging fit over this if it takes off (which I doubt).

Why? Online gaming already taxes their services. It happens precisely because bandwidth (among other technology) allows it. It might well attract users and even allow some mutually beneficial deals between big game companies and ISP services.

I can't wait until everything I do with my computer is output through a highly compressed audio/video stream.

Starting up your computer to read something you left in a text file, you will get the following error:

SYSTEM FAILURE: Windows was unable to detect an Internet connection. You cannot use your computer without an Internet connection. Press ENTER to restore your network settings, or press ESC to turn off your computer.

AgentSpork said:

I can't wait until everything I do with my computer is output through a highly compressed audio/video stream.

Which is just like using a computer remotely with extremely long video/audio/controller cables, only with the complication of a network protocol on top (even with analog signals, there would be propagation delay). Anyone else seeing the lameness factor in this?

Maybe one day computers will be so fast that even exponential-time algorithms will be acceptably fast for everyday use. Following the broadband pimp's logic, that would be viewed as a "breakthrough in programming techniques" (which are already going down the drain as it is...)

myk said:

If this sort of thing really takes off and accumulates potential I'm guessing they'll introduce modifiable elements in the games so people can mod them in some respects, to compete with that advantage in traditional games.

And maybe one day, when this sort of "gaming" will be the standard, somebody will propose a "revolutionary" new gaming platform that does most of the processing and data storage locally "without the burden of a network", "faster" and being "able to operate just about everywhere". Not to mention that such a platform will offer "high quality video and audio without latency, breakups or compression artifacts"!

myk said:

Why? Online gaming already taxes their services.

Essentially, the service is streaming video over the network (5MB/s at full pelt), which is a lot more intensive than online gaming.

If it gets big, I'm sure ISPs (or their company groups) would look for a way to share a piece of the pie. It looks like the type of service they could actively get into.

If this service succeeds, that would not be a bad thing. That would not be the harbinger of the end of all localized computing. It would just be an alternate way to play games, and options are good. It's a little bit silly to not just expect but to hope for this to fail, or to weave paranoid scenarios about this service's implications.

"I think there is a world market for maybe five computers." -- Thomas Watson, chairman of IBM, 1943.