OnLive Game Service Shutdown, Sony Acquirer | Hacker News Reader

OnLive Game Service Shutdown, Sony Acquirer

support.onlive.com

37 pointsshurcooL32 comments

OnLive had some of the best tech for cloud gaming that I've experienced, but it seems like they could not get it to take off (I'd say due to very limited game selection, nothing new coming).

Sony is clearly going to use the tech to improve their PlayStation Now remote play. And hopefully have better luck with it, since they have access to many new and old games.

The idea behind cloud gaming has many advantages, for things like instantly trying demos, to playing old school games that aren't even available on your platform, gaming on MacBook Airs, etc.

I was a huge fan of OnLive in the beginning, but given that the service was semi-dead the last few years (no new games), I hope its tech will be better utilized by Sony.

I gave OnLive a try and was bewildered that I could play cutting-edge games on my Macbook Air.

unfortunately Time Warner Cable disagreed…

I would agree their tech was AMAZING. Even on my rather not so hot charter cable and very outdated 6 year old laptop i was able to play their latest games. The problem was the connection. I could feel the lag. I really tried to make it work but the lag was just to much. It wasn't onlive's fault. the ping time was just to much.

> I could feel the lag

That's exactly the problem. If the "ping" to the datacenter is too high, the gaming experience will be not great (lags). As the mouse/keyboard/gamepad inputs have to do a round-trip to the datacenter that is more problematic than traditional multiplayer games. You need several data enters around the world, near the gamers. Even than it will be unsuitable for certain games like high speed multiplayer ego shooters (Counter Strike, Quake 3, Unreal Tournament '99) where even a minimal lag of a wireless mouse/keyboard is noticeable and wired gaming input devices are used.

So what I always found interesting was that they didn't use the ping to their advantage.

Most online games have the ping problem anyways. Imagine everything in a [] is in the same datacenter or house.

[client] ------ [game server] ------ [other_clients]

[client] ------ [onlive server - game server ] ------ [other_clients]

And it doesn't really matter how the other clients are connected, they could be using OnLive as well. This would work amazingly well for online shooters or MMOs.

yes, shooters with >70ms latency BEFORE data even starts going out of your box into the display, plus additional 20-50ms before keypresses are even registered on the game code running hardware is the way of the future

I actually wrote an article for Gamasutra about this when they were first launching: http://www.gamasutra.com/blogs/DaleBeermann/20090327/83618/G.... Still, when I first tried it in their demo booth I was really impressed. But the data center was only 60 miles away. As soon as you get too far from the data center, there are actual physical limitations of being able to accomplish this effectively.

> The idea behind cloud gaming has many advantages

Some disadvantages also, as outlined by the FAQ answers in the linked article. Your ability to play games is at the mercy of the company continuing to exist, and if it decides to shut down or merge for any reason, your games and saved games might just be deleted with no ability to export.

>OnLive had some of the best tech for cloud gaming that I've experienced

and it was still NOT GOOD ENOUGH for current state of the internet (latency)

>The idea behind cloud gaming has many advantages

like being a slave to content owners, borrowing for the rest of your life

If you can't afford to buy, renting is better than nothing.

It's interesting to me that about 5 years ago John Carmack was talking up game streaming quite a bit in his annual QuakeCon talk. From what I remember he was pretty confident it would be the future of gaming at the time. However I'd love to know what went wrong and why game streaming as a whole has pretty much completely fizzled. Was it just too much hype and optimism that the technology would improve and latency wasn't an issue, or perhaps bad business models that doomed it from the start?

IMHO game streaming is pretty much doomed if the future of gaming is ultra low latency and high framerates for virtual reality. There's just no way a streamed game could work with Occulus Rift, etc. and have low enough input and display latency to be believable.

edit: I think it was his 2012 QuakeCon talk: https://youtu.be/wt-iVFxgFWk?t=3522

I'm thinking that in a few years there will still be games with intense graphics that aren't VR - these will just be a new type of casual game. In the last months the barrier to use high-end engines has been significantly lowered, which should make these tools available even to such lower budget titles (but sometimes with a huge following). VR helmets will probably stay a 'nerdy thing' for a while, taking over the gaming world from the top end and (hopefully) through arcades.

I don't think streaming is dead. I think it's mostly suited to casual gamers though. And the casual gamer market should be much bigger compared to the hardcore gamer market anyway.

I did like OnLive on my MacBook Air quite a lot, but their game selection was quite limited.

The problem with streaming will always be that computers get better faster than Comcast (or w/e internet service you have) does. Just going off a quick search for GPUs, pretty much every metric has doubled in the last 5 years. My internet speed over the same period has literally not changed. Not to say that the GPU is directly responsible, but that if memory usage in games is growing at even a linear rate (which is conservative), it's still impossible for internet speeds to match that improvement.

It's probably things like resolution. I mean, we now have 5K screens in homes. Gaming on a high res (not 5k, but at least full hd) has become the norm, and internet speeds did not grow as quickly.

I think at some point the timing will be better because the growth speeds will reverse. i.e., the human eye cannot see the difference between refresh rates and pixel per inch density once it becomes too large. I doubt after 5k we really need more pixels.

Of course by then we'll come up with new stuff to use bandwidth capacity for instead so it's hard to tell. But with everything moving to the cloud & wireless, the notion that we'll wirelessly be outputting video to a monitor, or running graphics software in the cloud, it's not far fetched.

I kinda believe in streaming app. Onlive's problem is coming out too early.

music streaming, for example, came out long time ago. only recent attempts succeeded, such as spotify.

the internet needs to be improved to support streamed gaming.

I think that there's pretty little that can be done in terms of infrastructure improvements to the internet that will really help. We have more than enough bandwidth to stream these things at full resolution, but latency will always be the killer, at least among the less casual gamers. Short of cutting down the speed of light, what we can do to help there is pretty limited beyond building out datacenters in physical proximity to anyone who might want to use it.

Right. The speed of light is a hard limit here. The only way to cut latency is to move closer to the user. The closest you can possibly be to the user is on premise.

Which to be honest is not a bad idea at all...

Imagine playing your PS4 games (rendered and streamed from your PS4) on your MacBook, or any screen connected to your local network?

I've wanted this for years. Infact in university I even did all the theoretical design to build myself this server. As is normal, I learnt a lot, couldn't spend enough time to make it work though :(

Steam has in-home sharing, which is the same idea but PC games. It's actually pretty nice letting the desktop in the other room do the heavy lifting/run windows and playing some of the games on a laptop. I haven't tried any latency sensitive games though, just things like Civilization.

You can do this but only to other Sony-approved devices. Xperia and Vita I think. There is an unofficial app to enable this on more Android devices: http://forum.xda-developers.com/android/apps-games/ps4-remot...

Internet latency is around 50% slower than it could be, though: http://web.engr.illinois.edu/~singla2/papers/hotnets14.pdf

"The speed of light sucks." - John Carmack

http://archive.wired.com/wired/archive/4.08/id_pr.html

Sad to see them go, but it's been a long time coming.

I played through just cause 2 entirely on onlive, and wouldn't have been able to play it at all otherwise. And being able to demo any game instantly was cool too.

Ultimately, it was poor leadership and poor title selection that killed them. I would have used them a lot more, otherwise.

Some really cool tech - i expect we'll see it become more mainstream over time.

Sony acquired Gaikai time ago, and now OnLive

https://www.gaikai.com/

Too bad they didn't offer a migration path for existing games to either Steam saves or PSN.

I guess Sony is interested in tech, not in keeping OnLive's (few) users happy. It wouldn't work for most games anyway (like ones that are not available on Steam or PSN), and be too much effort, so I'm not too surprised.

They often didn't offer a migration path for their own service, as they'd frequently remove games. OK, contracts expire, but they were also flaky and didn't look after their existing customer base.

Thr least they could do is hold a patch up for a year that would allow people to general purpose the hardware they have.

This is one of the reasons why I am reluctant to go with technology that has almost no hope for an alternate backend when things shut down.

boy I'd hate to have paid full sticker price to manipulate a streaming video on the screen than buy the actual game itself.

especially when the company (and your investment in the game they are housing) goes poof.

that's also a gripe of mine pertaining to any walled-garden, though.

I have similar worries about Steam but at least we still get the games. I miss having the physical case and the cd like the good old 90s and early 2000s.

Reply on news.ycombinator.com