f13.net

This news broke earlier in the day but I waited to post about it since OnLive denied they were shutting down but the evidence says otherwise :oh_i_see:

Brian Fargo tweeted about this earlier in the day having received an email from a supposedly ex-OnLive employee with the termination email he/she received:

https://twitter.com/BrianFargo

http://venturebeat.com/2012/08/17/breaking-employee-email-says-onlive-is-closing-its-doors-today/

Wow. I remember when that launched and that is about all I can recall about it.

The messed up part is that over half the company got laid off with no severance. That really sucks.

You beat me. I was about to necro the OnLive thread, but waited conflicting statements out.

Either way, a service no one needs with an infrastructure expensive as fuck. Since neither Xen nor KVM got anywhere close enough to stably virtualizing PEG devices to date, especially the primary one, they pretty much had to run a server per game instance. Not really cost effective. Especially since no one's wanting to subscribe it.

That's too bad, it seemed like some fun tech. I just couldn't see how they were going to win enough people over to have a sustainable business model. Even services like Spotify have been tough to get off the ground and they require simpler tech, have a broader appeal, and distribute a cheaper product. OnLive seemed bogged with complexity, and I never really saw any advertising or promotion beyond a few vloggers noting how cool it was.

Is this actually confirmed now? Or is this still based on that once anonymous internal source?

I wouldn't be surprised of course. All the fun talk aside, I always felt they were better off showing games that didn't require such low latency. Light speed and internet routing and all that fun stuff that to date I still don't think they really got over (nor could without insert-TBD-tech-here).

Perfect device for casual games on TVs though.

Someone care to dig up my comments in the original thread about this obvious shitservice?

Nevermind, I did it:

They lasted longer than I thought they would.


This still holds true. However, I NAILED the expiration date on the company in this post. NAILED IT.


Yup.


Still true.

Assets sold to a new yet unnamed company:

http://venturebeat.com/2012/08/17/confirmed-onlives-assets-sold-to-another-company/

Some people will be getting their jobs back (if they want them).

Ah, there it is. Thanks (not that you did it for me :-) ).

Reports are coming out from laid-off staff that OnLive was running at a $5 million operating cost per month; that's even more than 38 Studios was burning through. With triple the staff of OnLive, no less! Oh, and OnLive only had about 1800 average concurrent users while the company trumpeted the 2 million accounts figure.

It seems like the guy running this outfit has just been trying to shop OnLive around to angel investors, get a multi-million dollar payday by selling the pretty much worthless company and quarter-assed service, and then hightail it outa there.

Oh, and apparently it's not a company that bought OnLive, but just a single individual who was "impressed" with the company and tech. I'd guess some clueless new angel investor who got hoodwinked.

Only knowing of its existence when it's dead is a good sign of something I shouldn't really pay much attention to.
Cloud gaming - hahaha , good luck with that. Max Payne 30gb bitches.

I'm curious how investors will react to the selling of the company to someone else yet unnamed. Especially if they are cut out or put at a substantial disadvantage.

Also makes the announcement of OnLive on the Ouya just a week or so ago even funnier.

The Ouya announcement doesn't shock me. Current owners probably wanted to put more value into their shit product for the sale.

Cloud gaming is the future of gaming.

Keyword there being future.

Man this is shady shit.  Its seemingly like the Onlive owners fired everyone so whoever was going to buy the assets had zero employment liabilities.

Serek Dmart just posted this on his Facebook feed with the comment, "Yet another case of egos getting in the way of common sense."  :grin:  :grin:  :grin:

DSmart makes the worst facebook posts.

http://www.joystiq.com/2012/08/20/onlives-mysterious-benefactor-revealed-almost-half-of-origin/

At least he's a good urbane liberal, I give him points for that.

I liked how the OnLive employees who weren't rehired were offered the opportunity to work as OnLive consultants in exchange for stock options. Because OnLive were so good about honouring the original stock they'd given employees.  :oh_i_see:

That is a primo dick move.

Recently there was a new onlive-like cloud gaming service launching in South Korea called "C-Games" (inspirational name right??) http://www.onlivespot.net/2012/07/lg-u-launched-c-games-cloud-gaming.html .

I'm pretty sure this is going to take off in South Korea.. the infrastructure is there and LG-U+ has a *ton* of capital (enough to get them over the early-adopter chasm). Pinging 1-5ms with an 100 mbps connection is standard across most of SK, so as long as they manage to keep their server costs down, the actual gaming experience will be near-identical to what you get if the hardware is actually in your box.

Honestly, the largest hurdle is game input on devices like tablets/whatnot. But, putting that aside, imagine being able to buy an ultra-cheap (compared to what you need to run games at ultra-res) piece of hardware, and run any game without even having to store it on a drive. The only limit on graphics being the size of your display...

That's seriously awesome.

Also, would probably be useful for lightweight VR hardware directly piping in from the cloud. Not to mention that multiplayer games could be optimized for this tech since the actual data processing happens in very few locations.

Have to agree with Goreschach that cloud gaming is the future and it looks to me that it's going to start in SK.

1. u cant mod a game.
2. u dont own the game, ur just renting it

cloud gaming is bullshit.
after GFWL,  ea origins and d3 authentication hoopla, why would i even support this sort of venture?

Just because a game might reside on the cloud doesn't mean these cloud gaming services couldn't offer the option, or even just change their business model, to actually store the game on their cloud somewhere with you having complete rights/access to it. If they went in that direction, you could install whatever mods/customizations you wanted on it (maybe they would give you a max 4GB additional modding space or something for each game? who knows) and would be able to pull the data directly down to your own box if you wanted to.

I'm not saying they are smart enough to actually give ownership like this in the short term, but eventually, someone will do it.

I saw 5ms ping and my pants fit a little tighter.  The younger twitch gamer in me who loved CS on his parents' 56k connection  wants less ping for his CS 1.5!

They actually had a quite promising product for running Office apps (http://desktop.onlive.com) from the cloud. Looked like a practical way of running full-scale desktop apps on, say, an iPad. They did run into some problems with Microsoft, though, around licensing.

In reference to what?

Screen scraping tech is not new. Cisco Citrix has been doing that for a decade+.

The new bit is the "from the cloud" part. OnLive runs the servers that you're connecting to. Effectively, they're providing a VM that you can run from your iPad (or whatever.)

And their screen scraping tech is of course the best in the business since it was developed for running games.

Anyway, my only point was that the OnLive tech is applicable well beyond gaming.

IMO client-server stuff is often a case of grass looking greener on the other side.

"This is easier to manage, we only have to maintain a thin client. It's so convenient! We can get rid of most of our IT department!"

"Oh man this is laggy and looks bad. Their servers went down and now we can't do anything! We don't have precise control over the software and versioning. We need to take this stuff in house!"

"Oh man, managing this stuff in-house is a nightmare. We have to do all this IT! To the cloud!"

For games though, given that control and graphic fidelity are so important, streaming stuff just doesn't make much sense to me. On one hand people are jazzed about Retina displays and big-screen TVs and XBox 720 or whatever, on the other hand they're ok with laggy games with bad video artifacting? If anything it makes a lot more sense for business software or barely interactive stuff like movies.

I still think that game streaming has its niche. But OnLive's business model was all wrong. I can imagine people paying a $10-15/month sub that allows them to play a bunch of games on their TV. But having to purchase each and every game at full price? Pretty obvious that wasn't going to work.

Heh.  It's the truth.

All I can do is give the pros and cons for each and threaten to walk when they want to change it up yet again.

This. From day one. They wanted to get people to believe their tech could work for the very audience that just had to walk by it to know it was not going to work that way. Then layer in the business model.

I'm ambivalent about cloud vs client based. There's nothing I do that requires deep access to anything beyond what the publisher is providing. And if modding is worth it as a business, they'll support it in some vague form that is more promising on paper than reality, until it's really worth it as a business (as in, the best modders don't just get scooped up by the developers). I'm not tweaking my BIOS to eke out an extra 5% boost. And I'm paying others to develop games for me because they're better at it than I am.

As long as a fast reaction actually-using-that-3D-world game requires some type of input that's both fast and accurate, just gimme the keyboard and mouse and I'll be fine.

If you use any DD service and don't have all the titles you've bought from there saved on your HD, you're vulnerable to similar issues. Plus the proportion of players who mod most games is very much the minority.

But it's the way things are going. I don't trust the cloud either, but a lot of people seem very taken with it.

In the proper high bandwidth low ping environment, the cloud offers better graphics for a lower price without having to care about anything except the keyboard and mouse.

Most of the time I don't have to screw around with my comp or it's hardware, but when I do, it's a pain in the ass (most of the time). Instead of that just push everything to the cloud and let someone else take care of it way easier due to efficiency of scale.

Better 720p graphics in a tiny radius around expensive servers for a higher price than owning the product.

Computing power has never been cheaper, the idea of renting it makes no sense to me. Outside of a couple of niche cases which are insufficient to build a business on.

I'm referring to the South Korean situation in regards to availability of bandwidth/computing power.... though I do see what you mean about not having to "rent" computing power... i'm sure that with current tech you could get the cost of "renting" down to much lower over time then what it would cost to run your own computing power (electricity and upgrades from time to time)...

According to various articles OnLive was running 8000 servers even though they had a peak of 2k on at one time.  The problem with the whole model is it doesn't scale very well, and the problem with games is that 1 user has to take up a whole GPU for the duration of the game, so at a minimum you need 1 GPU per expected user at peak time.  Combine that with the CPU performance (You'd need at least 1 maybe 2 CPUs dedicated per use as well) plus a TON of ram per game, and it just isn't efficient.

For computing power, this is one instance where I see virtualization being a huge boon.  Of course the rest of the business doesn't make sense, but if they were really doing a 1:1 server farm, that's pretty crazy.

You still need a 1:1 relationship for GPU, CPU and RAM per user to have acceptable performance (even before lag is taken into consideration).  Virtualization doesn't buy you much in these regards.

Unless someone builds a game engine that allows a single machine to do scene rendering for multiple camera actors simultaneously to take advantage of being able to share memory allocated to primitives and calculate things that aren't point of view specific with massive performance gains scaling with the amount of players per map.  Presumably that would be a single physical machine, because I imagine large amounts of latency in retrieving primitives for rendering is unacceptable.

Of course, everything currently is programmed with the assumption that one process will be rendered to one display, because that's a sane thing to assume.

Well that was fast.

http://www.joystiq.com/2012/08/27/onlive-founder-ceo-steve-perlman-exiting-company/


Amusingly, Perlman also has another startup going (since 1999 apparently, that he's now going back to), named Rearden. Yes, that Rearden (http://en.wikipedia.org/wiki/Atlas_Shrugged).  :oh_i_see:

OnLive lost: how the paradise of streaming games was undone by one man's ego (http://www.theverge.com/2012/8/28/3274739/onlive-report)

The one man being Steve Perlman obviously.

Haha, holy shit. No wonder they died.

Eh. The article says they were trying to develop virtual machines to take the load, but it seems like Perlman managed the company into the ground.

VMs don't really do GPU stuff well at all, I'm guessing that was the hold up there.

NVidia tried to give them a solution and Perlman told them to stuff it.

Yeah, that pretty much is exactly NOT what "cloud" and server-based gaming solutions are about. If you have to have one machine per user, what the fuck is the point of the service?

Given that not everyone plays every day, there is still big opportunity for savings. Only one person can sit in a seat in a taxi cab, but that doesn't mean that you need a taxi for every person in your city...

But yeah, dunce hats all round at OnLive apparently.

I think we are going to have to implement a counter for whenever anyone says "the cloud", then dole out punishment on intervals.

I bet it's even worse than that.  It's not like server blades are built for GPU's.  They probably had to come up with some sort of custom solution that took up craploads of space compared to a typical server farm.

Make up a drinking game for that and web 2.0.

Apparently they weren't totally dead but now they really are.

OnLive shuts down streaming games service, sells patents to Sony (http://arstechnica.com/gaming/2015/04/onlive-shuts-down-streaming-games-service-sells-patents-to-sony-embargoed-7pm-eastern/)

Good. Fuck them.

I wonder what kind of patents they hold.  They can't be that valuable if NVIDIA is able to have the exact same service.

Check the claims vs. filing dates if you're bored:

http://www.onlive.com/legal/patents

Here's a bunch off Google Patent Search. (https://www.google.com/search?q=or+inassignee:%22Onlive,+Inc%22+|+inassignee:%22OL2,+Inc%22&hl=en&tbs=ptss:g,sbd:1&tbm=pts&ei=SzAgVfWuH8qmgwTwlYSACg#hl=en&tbs=ptss:g%2Csbd:1&tbm=pts&q=inassignee:%22Onlive%2C+Inc%22+|+inassignee:%22OL2%2C+Inc%22) Since their patent page doesn't link through.