Metanet Forums

http://www.onlive.com/

Due for release at the end of this year, OnLive is set to be the Youtube of the gaming world. By grabbing games straight from the Internet, and funneling them directly to your computer or television in less than a couple of milliseconds using some crazy efficient algorithms, there's no need to download games, or clog up your house with discs. In fact, apart from a 1.5Mb broadband connection (or 5Mb for Hi-Def), you only need a palmtop to get going. Alternatively, the service is provided in a console version, which is the size of two PSPs stacked on top of one another.

Personally, I'm psyched, this is such a brilliant idea. But there's no question that games will be just a little more crisp if you have a copy at your end of the cloud, but graphics aren't a big bother to me, my only current-gen console is a Wii. Another worry is the possible lag with hundreds and thousands of people crowding the gargantuan servers.
But what are your opinions? Many reckon it'll be the end of consoles, and Steam alike, but I'm not so sure.

A few links from IGN's tests:
http://pc.ign.com/articles/965/965535p1.html
http://pc.ign.com/articles/965/965542p1.html (This one brings up many potential flaws)

I'd love to play Crysis on my XPS m5130 at anything over 15 fps, so I'm a little chuffed about the potential here. The potential flaws seem reasonable predictions but the company seem confident (but then, we've heard that all before). The lose of visual quality is a little bit of a worry (I like my games crisp and clean, no matter the quality of graphics (hyper real to stylized flash), and my current ISP applies a hefty cap for a few hours if I really hammer my connection, so there are issues and I'd be interested to read/experience how and if they can be overcome.

I can imagine purists making a noise, and perhaps rightly so. It's fun and satisfying to earn yourself a system capable of running the latest software at blistering speeds. Tweaking and troubleshooting and overclocking and upgrading, for some of us, is all part of that fun. It'll have to be a pretty special and very hassle free service to trump that.

For may though, Hi-End gaming without dedicated hardware will be a very, very exciting prospect. I guess the truth will be in the cake :)

The prospect is nice. There is just one giant problem.
It won't actually work

For about 95% of the population, they can't even get enough bandwidth to support this. Think about how fast youtube videos load and play for you. Now think about trying to load at least 4 of those at the same time. That would still give you relatively shitty quality even at just 1024x768 (think lots of high-compression artifacting).

Once you have got quality of the image out of the way, there is still the lag time between you pushing the button on the controller, the signal getting to the server, it processing what you want to do, and then sending the current frame render back to you. This will incur a good quarter second lag between you pushing buttons and the game responding on all but the highest bandwidth connections.

Now imagine trying to play online with this, where not only does your controller input get sent to the server that is essentially taking the place of your computer, that signal is then being forwarded to the game server of whatever game your running, then the reaction is being sent back to the render server, and then the video stream of the render is being sent to you. I have played TF2 on a british server where my latency was ~150ms, which equated to .3 seconds between me doing something, and the game server sending me back the reaction (ie: I headshot someone, .3 seconds later they die). Now try doing that, but with closer to a second of lag.

The whole idea is just current technologically infeasible, we would all need true gigabyte ethernet and the server would have to be next door.

I agree with everyone. I'd love to see it happen but I don't think I will.

I agree: YET.

Perhaps they're being all nintendo about it and Sony or someone will take what they've done and improve it later on?

(I mean come on, Six Axis versus Remote and Chuck?)

Pretty neat idea. Too bad I don't have a good internet connection...

I'll believe it when I see it. You know, when their data transfer algorithm is good enough to actually do what they claim. As in never.