SoylentNews is people
SoylentNews
2019 will be the year that sees some of the biggest tech companies in the world diving into the cloud streaming business for games. Google announced its Stadia platform at the Game Developers Conference 2019 for a launch scheduled later this year, Microsoft confirmed plans to publicly test Project xCloud in the coming months and even Amazon is rumored to be readying its own cloud-based streaming service.
With competition often comes strife, and in an interview with the Telegraph (locked behind the publication's paywall), Microsoft's Chief Marketing Officer for Xbox Mike Nichols didn't pull any punches when he discussed the weaknesses of the upcoming cloud-based streaming platform.
While he admitted that Google has the infrastructure (7,500 edge node locations) to pull it off, Nichols pointed out that unlike Microsoft they don't have strong ties to game developers and publishers to deliver the content that fans expect.
[...]Microsoft CFO Mike Nichols also went on to reiterate that regardless of the availability Project xCloud, the 'local' experience users can get on an Xbox console or Windows PC will remain superior to the cloud in terms of quality.
Previously: Google and Microsoft Eyeing Streaming Game Services
Google Announces "Stadia" Streaming Game Service
(Score: 2, Insightful) by DECbot on Saturday April 13 2019, @07:00AM (2 children)
Yeah, the Microsoft guy is correct. Local is better right now. But you know what? Turbo props were once more reliable than jet engines, so there's no chance that air liners would ever switch. Nobody will use dumb terminals again because
the cloudbig iron costs too much. Consumer automobiles will remain as internal combustion engines because they have more range. People will always buy expensive hardware for local gaming because it gets better performance. Nobody will want to use cheap tablets and smart phones to play triple A game titles because they like playing as much as they can to Microsoft and Sony. Google has no idea how to run a mobile platform and guage its limitations, but Microsoft does. Their Windows phone failure makes them more aware of the difficulties of building and sustaining a mobile platform. Nobody wants to play games on a mobile platform because nobody will write games for phones. Just look at these metrics the Windows Mobile app store. In 2019, virtually nobody plays games on Windows Mobile, so why would they play Stadia games on devices they already have? That's just crazy. Consumers want to spend money renting software and buying new, gaming only hardware every 3 years. WTF does Google know about people's devices, how people use their devices, and using gaming platforms? It's not like Google has a browser for every mobile device and general purpose OS.cats~$ sudo chown -R us /home/base
(Score: 3, Interesting) by Bot on Saturday April 13 2019, @09:52AM
I would not spend CPU time wondering on the merits/demerits of streamed games. The only factor is the amount of control you can exercise. If you return to the dumb terminal, you have no control. Technical problems can be overcome. It's the same reasoning that gave you cars where you cannot simply replace a burned light.
Account abandoned.
(Score: 0) by Anonymous Coward on Sunday April 14 2019, @10:21PM
Local is better due to latency, which is a huge, difficult problem.
Still, there's a few options, and they can be used in tandem.
A: move the servers closer -- which is expensive, and you're still fairly heavily at the whims of your users' connections, and you're still nowhere near how local play responds.
Still, Google can totally do this one, which gives them a leg up over other prior cloud gaming services.
B: some kind of GGPO style black magic (which basically involves re-running the game whenever there's a discrepancy between what should have happened when you pressed the buttons and what actually ended up happening). RetroArch's "runahead" feature is very similar, to reduce perceived latency on a local machine.
Very intensive, and the player will still have somewhat of a disconnect, and there's going to be the choppiness of when resyncs happen.
Still, I like this option the most despite the drawbacks (seriously, it's really good at killing latency, I can fight a guy in Korea in KOF98 like he's next to me, even if there's some weirdness on the screen sometimes between syncs), I just don't know if Google's actually going to do anything like this.
But yeah, it's really intensive, and doesn't scale nicely to modern, CPU intensive games that well.
C: design games with extreme input latency in mind.
This means you don't have access to the large, existing library of games that people would want to play, and you'd need to actually go and design those games. It's harder to do this than to just keep designing for locally placed hardware.
D: literally warp space so that your signals reach faster (lol)
ultimately, the problem is that signals take time to travel, and trying to keep latency under two 16ms frames at 60fps (or one 33ms one at 30fps) is hard, let alone dealing with network pings of 100+ms
like, just the signal getting to your TV, getting decoded, and being displayed is a latency issue, what on earth are you going to do about a signal having to bounce back from miles away through the internet on top of that
I've used OnLive and whatever the hell Sony calls their service. It was nice being able to play some games way above my hardware spec on a literal netbook and a Vita, respectively. It wasn't nice how fucking long inputs took to actually reach the server and then for me to see the result.
Assassin's Creed worked a-okay on OnLive, because that game automates a bunch of movement anyway, so you don't need to react that much. Played through most of Brotherhood on the service before it shut down (also, this is one major factor against cloud play, not being able to play the games when the service shuts down).
They also offered Street Fighter 4, which showcases every possible flaw with the model in full glory. Good lord, that was awful. I think they deliberately wanted to show off the worst case scenario.