Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
DebonaireD
May 7, 2007

This looks really amazing but goons will always insist on being loudmouthed pessimists about everything. There's so many interesting things you could do with streamed games that no one's even thought about yet, especially multiplayer games.
If I understood that digital foundry video, he was talking about how having the same server running the multiplayer session and streaming to each player would open the door to more complicated physics and interactivity in multiplayer, right? Hell, just being completely unbound by the hardware specs would be a very cool thing for any developers trying to make more sim-heavy games.

Adbot
ADBOT LOVES YOU

big nipples big life
May 12, 2014

pixaal posted:

You can reduce input lag the same way we made computer processors faster, predictive execution. This does mean a game would have to fork and attempt to predict player input and start rendering things as if the player had done different possible inputs and only send the correct ones.

This wouldn't make a difference with input lag.

CAPTAIN CAPSLOCK
Sep 11, 2001



DebonaireD posted:

This looks really amazing but goons will always insist on being loudmouthed pessimists about everything. There's so many interesting things you could do with streamed games that no one's even thought about yet, especially multiplayer games.
If I understood that digital foundry video, he was talking about how having the same server running the multiplayer session and streaming to each player would open the door to more complicated physics and interactivity in multiplayer, right? Hell, just being completely unbound by the hardware specs would be a very cool thing for any developers trying to make more sim-heavy games.

lol

This quote also works for the kinect.

e: Just so I'm not making GBS threads on you, there will always be inherit flaws with streaming. Ones that you can never get around like input lag. PSNow has existed for a while and it isn't setting the world on fire. And if the best results they can demo to people at their show is mediocre, just imagine how bad it'll be for the regular user.

CAPTAIN CAPSLOCK fucked around with this message at 20:44 on Mar 19, 2019

Feels Villeneuve
Oct 7, 2007

Setter is Better.

DebonaireD posted:

This looks really amazing but goons will always insist on being loudmouthed pessimists about everything. There's so many interesting things you could do with streamed games that no one's even thought about yet, especially multiplayer games.
If

Yeah. For example you could do cool things like take them off your streaming platform and have them be unavailable forever because it's simply not profitable to keep letting people play your game.

SeANMcBAY
Jun 28, 2006

Look on the bright side.




Wow. I didn’t expect it to be that bad. Imagine trying to play a platformer over that.:lol:

Bardeh
Dec 2, 2004

Fun Shoe
166ms delay even in ideal conditions there on the show floor. fuuuuuck that.

explosivo
May 23, 2004

Fueled by Satan

DebonaireD posted:

This looks really amazing but goons will always insist on being loudmouthed pessimists about everything. There's so many interesting things you could do with streamed games that no one's even thought about yet, especially multiplayer games.
If I understood that digital foundry video, he was talking about how having the same server running the multiplayer session and streaming to each player would open the door to more complicated physics and interactivity in multiplayer, right? Hell, just being completely unbound by the hardware specs would be a very cool thing for any developers trying to make more sim-heavy games.

There's definitely a lot of hyperbole going on here (your old games aren't going anywhere and you can still buy games from Steam and install them locally) but the input lag thing is for sure a red flag. Some of the videos from the past two pages are not exactly promising either.

punk rebel ecks
Dec 11, 2010

A shitty post? This calls for a dance of deduction.

Blotto_Otter posted:

This framing is far, far too generous to Google. In my neck of the woods, Google Fiber fizzled out primarily because they found out that significant municipal infrastructure projects require negotiating with a wide variety of public and private parties, which happens on a timescale measured in years rather than months... and then they just gave up.

Google has no grand vision for what it wants to do and no real business pressure to do anything sensible as long as the ad money keeps flowing, and so it just bounces from one half-assed project to another. Most of Google's non-search business ventures consist of some high-ranking goober getting an idea and convincing other high-ranking goobers to pour all that search-engine ad money into it for a few years, only to then realize that what they've gotten into is either far more complex they anticipated or has an ROI that's short of the once-in-a-generation, impossible-to-replicate ROI of their search engine, and then they drop it and move to the next goober's idea.

It's sad because Google has some GREAT ideas.

Ultra cheap laptops just for streaming, lite gaming, and basic office work?

Sci-fi type glasses where the argumented world and real world become one?

Fiber internet finally coming Stateside?

All of these seem like incredible concepts that were unfortunately either abandoned due to fear of significant commitment or were too ahead of their time. What companies like Apple and to a lesser extent Nintendo show is that if you really have to fully commit to your product and know exactly what to sacrifice and what to not. The iPhone and Switch were/are so successful because their respected companies knew what vision they wanted for those products and did what needed to be done to execute said vision. In contrast, Google just doesn't do the same with their projects. And I call them projects because that's what they feel like, more like alpha (or beta at best) releases going public.

pixaal posted:

Everyone saying they don't like not owning a game fine, but think about how much you'd save if you didn't need a gaming computer? A $500 computer will work just as well as a $2,500 gaming computer.

The tech is not ready, the compression is all over the place and properly solving it is probably going to require coding games specially with this in mind. Delay can be accounted for. Nintendo did syncing the display on the controller with the TV over wireless for the WiiU. There was some trickery there if I remember the technical details behind it where it should be impossible but it works.

I think this is doable but we are currently seeing the Virtual Boy implementation of VR. We're probably decades away from this being a viable replacement to modern gaming. It will be a massive cost savings though.

As I've said before, new hardware is needed less and less.

I just ran Devil May Cry 5, a "graphics king" contender, maxed out at 60fps with an almost 5 year old GPU and an 10 year old CPU. Neither of which were at the absolute top of the line when purchased. Things will only stay viable longer from here on out.

Cabbit
Jul 19, 2001

Is that everything you have?


From 0 to Google Plus in about 60 minutes, nice.

jokes
Dec 20, 2012

Uh... Kupo?

DebonaireD posted:

This looks really amazing but goons will always insist on being loudmouthed pessimists about everything. There's so many interesting things you could do with streamed games that no one's even thought about yet, especially multiplayer games.
If I understood that digital foundry video, he was talking about how having the same server running the multiplayer session and streaming to each player would open the door to more complicated physics and interactivity in multiplayer, right? Hell, just being completely unbound by the hardware specs would be a very cool thing for any developers trying to make more sim-heavy games.

There's a difference between the concept, and the actual implementation of it. This is not a new concept, it's already been done. But all other implementations of this are awful because of the same reason this is awful: input delay, frame drops, unstable/insufficient network connections.

Feels Villeneuve
Oct 7, 2007

Setter is Better.

jokes posted:

There's a difference between the concept, and the actual implementation of it. This is not a new concept, it's already been done. But all other implementations of this are awful because of the same reason this is awful: input delay, frame drops, unstable/insufficient network connections.

Yes and the cause of these problems are not things Google nor the end user are in a good position to fix, mainly that the last mile between Google's big GPU box and your PC is out of their hands.

RPATDO_LAMD
Mar 22, 2013

🐘🪠🍆

pixaal posted:


You can reduce input lag the same way we made computer processors faster, predictive execution. This does mean a game would have to fork and attempt to predict player input and start rendering things as if the player had done different possible inputs and only send the correct ones.


There are games that already do this to reduce the feeling of lag in multiplayer. There's a popular library for predictive execution netcode aka "rollback" netcode called GGPO that's used in a few fighting games. The local player acts with no lag but it uses predictive execution stuff for other players, and rolls back a few ms to a previous state if their actual input was different enough to cause different results.

The problem is, you still have to graphically render the predicted state locally while you wait for the network to bring updated data in. With Google's setup where video is just streamed to the client device, it would be impossible to implement.

RPATDO_LAMD fucked around with this message at 20:54 on Mar 19, 2019

big nipples big life
May 12, 2014

Feels Villeneuve posted:

Yes and the cause of these problems are not things Google nor the end user are in a good position to fix, mainly that the last mile between Google's big GPU box and your PC is out of their hands.

that and the whole speed of light thing

DLC Inc
Jun 1, 2011

the main thing here is: what loving games can we play on it. we saw two. I'd want to see poo poo like DMCV, Rage 2, Cyberpunk, Witcher 3, big boy poo poo that would never run on my outdated laptop but could be played at 60fps streaming at ultramax HD. a lot of the success isn't just on stream and input connection but the amount of devs supporting this. right now all I have to go on is "it might work better than other stream services" and maybe not even that if input lag is bad.

Zephro
Nov 23, 2000

I suppose I could part with one and still be feared...

pixaal posted:

Everyone saying they don't like not owning a game fine, but think about how much you'd save if you didn't need a gaming computer? A $500 computer will work just as well as a $2,500 gaming computer.

The tech is not ready, the compression is all over the place and properly solving it is probably going to require coding games specially with this in mind. Delay can be accounted for. Nintendo did syncing the display on the controller with the TV over wireless for the WiiU. There was some trickery there if I remember the technical details behind it where it should be impossible but it works.

I think this is doable but we are currently seeing the Virtual Boy implementation of VR. We're probably decades away from this being a viable replacement to modern gaming. It will be a massive cost savings though.
Yeah I think this is pretty much right. You can design around a lot of the input-lag problems if you know that's going to be the use-case. In many ways smartphones are objectively terrible devices for playing games because touchscreens suck for precise input and your thumbs obscure half the playing area. There's still a huge market for smartphone games. Streaming will limit what you can do, but that's not necessarily fatal.

OTOH it also gives you a chicken-and-egg problem. Devs aren't going to bother designing games for streaming platforms first until they have a load of users. Platforms aren't going to have a load of users unless the game experience is good, which it might well not be because devs aren't designing for it, etc.

Also that post about Google having a tiny attention span and just lurching from one half-arsed project to another is also completely correct.


edit: from that point of view launching your product with Doom Eternal, a twitch FPS from the franchise that founded the twitch FPS, seems to be a really bad idea

Zephro fucked around with this message at 20:55 on Mar 19, 2019

Cicero
Dec 17, 2003

Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.
I'm surprised at the 166ms number, given how many gamers were pleasantly surprised at the input lag when they did the Project Stream beta last year. It sounded like it was considerably lower then.

If nothing else, this seems like it'd be really dope for instant demos/trials of games.

How are u posted:

I am really surprised to see this thing being panned across the internet. It is absolutely the future of playing games.

For me, a guy with a 3 year old laptop that can play non-graphically intense games pretty OK, this service looks loving cool as hell. I'd love to be able to subscribe to a service and get to try all the big flashy games without having to keep up with hardware and buying expensive new titles.

I will definitely be giving it a shot.
Part of it is that this event was clearly targeted at developers and 'creators' more than consumers. They want to get those parties on board first, I'm sure they'll do a consumer-oriented event with actual details like pricing and game availability later.

I said come in! posted:

The biggest misstep Google took today was not discussing pricing.
Nah. When they're ready to announce an actual launch date, I'm sure they'll do that then too.

jokes
Dec 20, 2012

Uh... Kupo?

Feels Villeneuve posted:

Yes and the cause of these problems are not things Google nor the end user are in a good position to fix, mainly that the last mile between Google's big GPU box and your PC is out of their hands.

Right, so this project is poo poo and is DoA.

e: Also, important thing everyone needs to note is that this technology isn't new. The established games you can get away with severe input delay are turn-based games, strategy games, and single-player games that have a lot of leniency as far as reactivity goes.

Feels Villeneuve
Oct 7, 2007

Setter is Better.

big nipples big life posted:

that and the whole speed of light thing

Yeah it's kind of like the whole "replace everything with autonomous taxis" thing where eventually you run into the problem of "there is literally not enough space on the road"

DebonaireD
May 7, 2007

explosivo posted:

There's definitely a lot of hyperbole going on here (your old games aren't going anywhere and you can still buy games from Steam and install them locally) but the input lag thing is for sure a red flag. Some of the videos from the past two pages are not exactly promising either.

Maybe it'll get better and maybe it won't. It's not like anyone's saying anything new by going "but the input lag!" So yeah wait and see until it actually comes out but I am optimistic. I actually used onlive when it was around and thought it was such a cool idea, and maybe this is a tech whose time has finally come. I've got a 4k tv and would love to play games at 60fps without buying a new $500 card. Not to mention all the neat side effects, like take over someone's stream, no load times, more multiplayer physics, etc.

Feels Villeneuve
Oct 7, 2007

Setter is Better.

DebonaireD posted:

Maybe it'll get better and maybe it won't. It's not like anyone's saying anything new by going "but the input lag!" So yeah wait and see until it actually comes out but I am optimistic. I actually used onlive when it was around and thought it was such a cool idea, and maybe this is a tech whose time has finally come. I've got a 4k tv and would love to play games at 60fps without buying a new $500 card. Not to mention all the neat side effects, like take over someone's stream, no load times, more multiplayer physics, etc.

The input lag is based on an inherent limitation that Google is not able to fix, unless you are lucky enough to live in a city with Google Fiber.

American McGay
Feb 28, 2010

by sebmojo
Here's the thing: game companies aren't going to bother implementing the graphical improvements that won't just get chewed up by the streaming encoding (physics stuff, other weirdly intensive compute stuff that I can't think of because I'm not a game dev) because there's no reason for them to. I doubt anything that comes out on this will be anything other than a slightly optimized PC port. Without a solid user base there's no reason for them to develop games for the feature set of server hardware, and without a decent play experience (less than 100ms of lag) there won't be a user base.

e;fb

Zephro posted:

OTOH it also gives you a chicken-and-egg problem. Devs aren't going to bother designing games for streaming platforms first until they have a load of users. Platforms aren't going to have a load of users unless the game experience is good, which it might well not be because devs aren't designing for it, etc.

The 7th Guest
Dec 17, 2003

Cicero posted:

I'm surprised at the 166ms number, given how many gamers were pleasantly surprised at the input lag when they did the Project Stream beta last year. It sounded like it was considerably lower then.
It's because Assassin's Creed Odyssey already has over 100ms of built in input lag at 30fps, and according to digitalfoundry, 145ms on xbox one. so you already are used to it with that game

If project stream tried to do a preview event with, say, Cuphead, or Tekken, or Rock Band, or Beat Saber, or basically anything that isn't an AAA 'cinematic' game, the feedback would be much different

pixaal
Jan 8, 2004

All ice cream is now for all beings, no matter how many legs.


punk rebel ecks posted:

Ultra cheap laptops just for streaming, lite gaming, and basic office work?
Chromebooks still exist, and are actively used in classrooms, it's what poor districts use while more wealthy ones use iPads. I think the poor districts are getting a better experience.

punk rebel ecks posted:

Sci-fi type glasses where the argumented world and real world become one?
These were abandonded for a good reason. It caused people to glance where the screen is even when not wearing them. It gave people lazy eye and other rare eye conditions. Microsoft also had a glasses project that they quietly abandoned.

punk rebel ecks posted:

Fiber internet finally coming Stateside?
I had fiber long before Google even stepped foot into it FiOS has been around for awhile. They did actually drop this one without a good reason though.

They also had project loon the really cool weather balloon wifi idea that they came out with after they killed their fiber as an alternative so they didn't need any permits.

They have tons of other projects that failed miserably.

They also had Project Ara a modular cellphone, and even patched support for this into Android OS

RPATDO_LAMD
Mar 22, 2013

🐘🪠🍆

Feels Villeneuve posted:

The input lag is based on an inherent limitation that Google is not able to fix, unless you are lucky enough to live in a city with Google Fiber.

166ms sounds a lot higher than the speed of light.

I've played online games with 5-15ms pings before, though the servers were probably pretty close to me geographically.
If google can get it down in that area then it will at least be a decent option for people who live near a datacenter.

jokes
Dec 20, 2012

Uh... Kupo?

American McGay posted:

Here's the thing: game companies aren't going to bother implementing the graphical improvements that won't just get chewed up by the streaming encoding (physics stuff, other weirdly intensive compute stuff that I can't think of because I'm not a game dev) because there's no reason for them to. I doubt anything that comes out on this will be anything other than a slightly optimized PC port. Without a solid user base there's no reason for them to develop games for the feature set of server hardware, and without a decent play experience (less than 100ms of lag) there won't be a user base.

That's true. Hell, even with a massive userbase streaming tends to poo poo all over graphical fidelity. Like how streams that claim to be 720p/1080p/4k are rarely actually that resolution because of whatever reason.

Feels Villeneuve
Oct 7, 2007

Setter is Better.

RPATDO_LAMD posted:

166ms sounds a lot higher than the speed of light.

I've played online games with 5-15ms pings before, though the servers were probably pretty close to me geographically.
If google can get it down in that area then it will at least be a decent option for people who live near a datacenter.

The inherent limitation I'm talking about is not SOL it's "the last mile infrastructure is godawful for huge parts of the country".

Zephro
Nov 23, 2000

I suppose I could part with one and still be feared...

RPATDO_LAMD posted:

There are games that already do this to reduce the feeling of lag in multiplayer. There's a popular library for predictive execution netcode aka "rollback" netcode called GGPO that's used in a few fighting games. The local player acts with no lag but it uses predictive execution stuff for other players, and rolls back a few ms to a previous state if their actual input was different enough to cause different results.

The problem is, you still have to graphically render the predicted state locally while you wait for the network to bring updated data in. With Google's setup where video is just streamed to the client device, it would be impossible to implement.
All the fundamental ways to attack this problem have been well-known at least since John Carmack wrote QuakeWorld back in the day. Any modern multiplayer game that is in any way twitch-based will be using those sorts of techniques already. I don't think there's much more you can squeeze out of them than what we already have.

Cicero
Dec 17, 2003

Jumpjet, melta, jumpjet. Repeat for ten minutes or until victory is assured.

The 7th Guest posted:

If project stream tried to do a preview event with, say, Cuphead, or Tekken, or Rock Band, or Beat Saber, or basically anything that isn't an AAA 'cinematic' game, the feedback would be much different
A rhythm game would actually probably be the best case for this, those already come with lag compensation tools. Just shifting the window for what counts as a hit is easy.

edit: not Beat Saber obviously since it's VR

Cabbit
Jul 19, 2001

Is that everything you have?

pixaal posted:

I had fiber long before Google even stepped foot into it FiOS has been around for awhile. They did actually drop this one without a good reason though.

That's cool. I live in a major metropolitan area on the west coast of the United States and I can't get fiber. The only available internet is either DSL or cable, and both have data caps that will gently caress you up if you cross them. It gets worse the further inland you go.

Data infrastructure in the US is hosed.

American McGay
Feb 28, 2010

by sebmojo
My dad has fiber to the curb in my small hometown and it sucks rear end. I've got 200mb internet which is good but it costs $160 a month and I've got a 1TB cap. There are very few people in the US who are happy with their internet service.

RPATDO_LAMD
Mar 22, 2013

🐘🪠🍆

Cabbit posted:

That's cool. I live in a major metropolitan area on the west coast of the United States and I can't get fiber. The only available internet is either DSL or cable, and both have data caps that will gently caress you up if you cross them. It gets worse the further inland you go.

Data infrastructure in the US is hosed.

Data caps aren't an infrastructure problem they're a lovely-company-with-monopoly-doing-whatever-it-wants-because-you-can't-leave problem.

Feels Villeneuve
Oct 7, 2007

Setter is Better.

RPATDO_LAMD posted:

Data caps aren't an infrastructure problem they're a lovely-company-with-monopoly-doing-whatever-it-wants-because-you-can't-leave problem.

The good news for Google is that this is an easily solved problem, and won't run into giant companies with billions of lobbying dollars at all

Cabbit
Jul 19, 2001

Is that everything you have?

RPATDO_LAMD posted:

Data caps aren't an infrastructure problem they're a lovely-company-with-monopoly-doing-whatever-it-wants-because-you-can't-leave problem.

I consider "who owns the infrastructure" to be an infrastructure problem, but fine.

Andrast
Apr 21, 2010


Hearing about how lovely american Internet is will never cease to amaze me

Andrast fucked around with this message at 21:23 on Mar 19, 2019

Steve Yun
Aug 7, 2003
I'm a parasitic landlord that needs to get a job instead of stealing worker's money. Make sure to remind me when I post.
Soiled Meat
I wonder if Stadia would do well in say, Korea, and potentially destroy the PC cafe business.

I’m sure Street Fighter V will feel the same, lol

RPATDO_LAMD
Mar 22, 2013

🐘🪠🍆

Cabbit posted:

I consider "who owns the infrastructure" to be an infrastructure problem, but fine.

It's much cheaper to build a new prison than new infrastructure. Just lock all the telecom CEOs up :black101:

punk rebel ecks
Dec 11, 2010

A shitty post? This calls for a dance of deduction.

pixaal posted:

Chromebooks still exist, and are actively used in classrooms, it's what poor districts use while more wealthy ones use iPads. I think the poor districts are getting a better experience.

Oh Chromebooks are still used, but I feel that they aren't the breakaway success that Google wanted them to be.

pixaal posted:

These were abandonded for a good reason. It caused people to glance where the screen is even when not wearing them. It gave people lazy eye and other rare eye conditions. Microsoft also had a glasses project that they quietly abandoned.

Yeah, the tech wasn't there for the vision.

The Wii U of Google products.

The 7th Guest
Dec 17, 2003

https://www.youtube.com/watch?v=WJSn1WmzHaQ

Blotto_Otter
Aug 16, 2013


Cabbit posted:

I consider "who owns the infrastructure" to be an infrastructure problem, but fine.

This is your periodic reminder that 20 states in the US either restrict or outright ban municipally-owned or operated internet services.

Adbot
ADBOT LOVES YOU

big nipples big life
May 12, 2014

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply