I could see games being able to signal "there is no input during this cutscene, so it can be buffered at the client." But for everything besides cutscenes, game streaming essentially has to work like a VoIP call (hard-realtime), doesn't it?
They may already be doing something similar (halt rendering on congestion) but when this kind of lag happened to me it felt like the rendering had continued without pause on their end.
It happens occasionally when I drive (bad LTE zones). On bad days (ie, if I'm tethering at hotel to join a conf or using airport wifi), I sometimes hear "robot voice" as the tool attempts to deal with signal attenuation.
This may not work as well in video games, but I would like the option.
I imagine most music, and some during-gameplay dialogue, could be buffered at the client. I'd expect sound effects to be the only audio that needs to be super low latency. Maybe certain dialogue too, if the speaker is visible on screen.
This alone might alleviate a lot of the GP's problems, since most audio wouldn't cut out. That can make a huge perceptual difference.
In essence, you need a fat cable to a datacenter very nearby for shooters/fast action games.
Regular wi-fi is more or less sufficient for everything less (you might get video degradation from time to time).
OnLive was the first and it was just insanely awesome. Too bad it was shattered.
I also tried LiquidSky, and it was insanely awesome. At one point you ended up having a virtual PC where you could install and play any game from your Steam library. Too bad they’ve changed the service and monetisation a few times, and it’s unclear if yhey are alive.
Haven’t tried others.
There's a lot of us out here who are curious how this thing works and the design tradeoffs.
In my experience the game would drop graphics before I would experience input lag. There were a handful of times that I did experience input lag. This was on a wired connection 100mbs down 20mbs up through Xfinity.
I did notice in the Google demo, that when the person was using the game pad, he experienced input lag when he was trying to jump up on top of that steeple on top of the building.
I'm still of the opinion that Google shot themselves in the foot here by having a bunch of wireless controllers in one room. Its like they've never talked to a Super Smash Bros Brawl tournament organizer before: Wii Nunchuks over 2.4GHz Bluetooth have dropped packets / dropped input issues when you get to ~20+ participants.
Don't do mass wireless in one room. It always ends poorly. I'd expect that local wireless in a typical living-room setting would be a better experience actually. After all: the major issue is whether or not the wired-connection / fiber backbone of the typical city is up to spec for this kind of thing. (A typical living room user probably doesn't have to worry about clogged 2.4GHz connections unless they're in an apartment I guess...)
lag is the least of your worries. Encoding artifacts will get to you and completely ruin the experience on anything that is not a simple puzzle game.
It was night and day, both in terms of latency and graphics quality. I don't have high hopes for this service.
I can't imagine trying to play something like a fighting game this way.
Anything more specific than that, i.e. lag would occur every 10 mins on avg?
If this were an entire room of WiFi controllers hitting the same WiFi frequencies all at the same time, it could have been a local problem.
All in all, we can't read too much into the demo-conditions. There are too many variables at play here.
No lag at all on my end. I am in the USA, so I have a solid but not amazing connection.
I should note that I have only played for half an hour because AssCreed was extremely boring. The tech itself was very solid in my experience.
I tried to catch some streaming artifacts, like parts of the screen not updating or a visible lag, but I could not see anything. In a blind test, I doubt I would have been able to differentiate this from running the game on my home console.
In any event I think that if you are a casual gamer and want to get off the graphics card upgrade train every couple of years, then this is a no brainer. It probably would not be a good match for multiplayer shooters, but for solo games, even AAA titles, it's a great option.
My feedback was that if they paired up with Steam they'd have a killer product. They might still have a killer product if they convince all those developers on Steam to also put their games on this service.
> Your approximate location is determined using information from local Wi-Fi networks, and is collected by Location Services in a manner that doesn’t personally identify you.
The idea that one can retune the radio into a different channel for 100 milliseconds or so and not impact user experience is absurd.
As someone who plays through DOOM on Ultra Nightmare, this kills the game. There are times when, if I were to miss even 10 or 15 frames, there would be a good chance that will be the end of my run.
It will instead, expand the audience to a lot of casuals who just want a little demon-slaying power-fantasy on a low difficulty. Nothing wrong with that, and it is a damn big addressable audience. Just like how mobile didn't "kill" any other platform, it just added a whole lot of candy crushers.
That's annoying.
Now, as a technicality specific to ACO, the goals should still appear in the mission's description, no matter whether you hear the NPCs stating those goals or not.
Slow connection speed? Run these tests, etc. Or, switch to a faster more secure browser. Sound familiar?
Google was just today fined 1.49 billion Euros for their ad strategy - this is reality: https://www.bbc.com/news/business-47639228