* Available for all browsers and comply all web standards.
* Available for all languages across all regions, with accessibility features.
* Promise not to "launch and shutdown in a few years".
* Imply a real future product.
But I'm guessing what Google wants to do is to just prove if "game streaming" is technically / financially viable for mass crowds based on their technology in hand. If not, maybe shelf and wait for a couple of years without dumping a lot of resources and drawing bad sentiment.
What requirements are you thinking which should be applied to such a "tech test"?
In contrast, I made a Steam account back in 2003, right when they launched. In college and grad school, I fell out of gaming, but when I came back in around 2012 with a brand new computer, all my old games were still available in my account, ready to be downloaded.
And the browser-based approach seems like a great solution for the platform issue. Being a Mac user I often miss out on steam games since they are mostly focused on Windows releases.
It also makes sense for Google to push for this sort of usage since it's more data going through pipes. Data they can analyze in some way, make money, etc.
The Google projects that make less sense to me are projects that are largely about data retention, long term stability, and are generally considered solved problems. I.e. email and office suites.
This is hardly true, Nvidia already does this, charges for it, and it works really well on a good connection.
But "Google" means something specific in regards to quality for Billions of people, and you start to have a responsibility to the brand regarding dependability at some point if you want to do things under the Google Banner.
But it's so easy to give these projects their own name, branding, and websites while they are in this tech test phase. If they hit on something great, then they should absolutely fold it in with the rest of the G Suite, but until they know why not sell to people with out the Google connection to see how it does?
And if something goes wrong, I can expect absolutely zero tech support?
And I'll lose access to all my games in 11 months when Google kills it?
And it'll post my Achievements and owned games to my Google+ account automatically and try to make it Social?
And if I get banned from the games stuff for some reason, I'll lose my other Google accounts like Gmail, since they're all linked?
With all that value-add and free extra latency, I basically HAVE TO sign up! What a deal!
I really like video games, and they keep getting worse... not the games themselves, but everything around them (due to "business" concerns), such as my above concerns, and IAP/DLC, and Xbox Live/PSN/Nintendo all charging money to play games on your already-paid-by-you internet connection, where it used to be free (since you can act as a server).
For DLC, in some cases, it's never sold on physical media, so for example you buy a playstation game, and half the game is DLC, and then 5 years later Sony shuts down the service that delivers the DLC, and you're left with half a game, so when your kid turns 10 and you want to show him this cool game, now you only have half of what you paid for.
Games now require mandatory updates to play alone by yourself (GTAV needs gigabytes of updates, no option to play the game offline alone and update later)... so you get home after work and want to play for 30 mins but have 2 hours of updates to download or more.
And a million other little things. Pardon me for not wanting Google to pour more gas on the garbage fire.
Since it is your company, do you have any insight into potential challenges that the average reader wouldn't consider? I understand that a benefit to "game streaming" is that a game originally not networked (like splitscreen only) can be played by multiple people over a streamed connection.
Streaming video at 60 FPS or greater and at 4K is possible. We do it all day at Parsec, but the economic challenges are real due to the requirement that you need to render video in real-time. Rendering in real-time requires each user to have access to a GPU or 1/2 of a cloud GPU until newer technologies allowing for true virtualization appear on the market. Cloud GPUs are generally very expensive to operate, and as many of us know, the cloud providers charge a lot to use them for gaming and/or ML. Companies with scale can lease them at lower rates and have lower energy costs lowering the overall cost to serve, but they are still very expensive to operate. Even if you get the rates down to $0.15 per hour, you're still looking at overall hardware costs at $10 per month. Your subscription price will have to cover the hardware and the content license. The content itself is very expensive to license as well because we consume games very differently from the way we consume TV and media. When people binge on Netflix, they're binging on many shows or many movies. When people binge on games, they're playing 1 game for 100s of hours. You can see this in the Steam Spy data. The median Steam customer only buys 1-2 games each year - https://galyonk.in/steam-in-2017-129c0e6be260. Many cloud gaming companies and game companies are aiming to increase the market for AAA games via streaming from the cloud. They're trying to reach the latent gamers who don't play games any more. In our opinion, that's a lot less exciting than delivering something valuable to current consumers of games. Consumers don't switch to a new distribution technology because it's a cool technology. They switch when it gives them something unique that they couldn't get before the technology existed.
This is why, at Parsec (https://parsecgaming.com), we're focused on delivering unique experiences around games via streaming technology. We make it so you can find other people around you and invite friends to play local co-op games with you online. We're recreating the couch gaming experience we love and making it available online. This experience brings a new element to gaming that gamers today benefit from and enjoy.
It would have been upvoted if it was a reply on one of the top comments that complain about Google, though. Placement.
Great job!
It was all-in-all easy, fast, and worked which is becoming more and more rare imo.
I rather enjoy my Steam Links, although I'm not sure how much traction they really got - not too long back I picked up a couple more on sale for like $10 a pop...
The idea of Parsec is that a machine you own is rendering and uploading a game that you also own. Google seems to be playing around with the idea of games as a service, where they render a game you don't own on a machine you don't own, and let you play it for presumably a monthly subscription.
Unless google undercuts competing small businesses in the same space, they're dead on arrival. They do not offer anything that's not on the market today.
From my own experience the extra delay can be tolerable in a few game genres, especially on console-optimised games where, to my knowledge, developers optimise controls for higher latency due to a possibility of a slower tv. Obviously PC FPSes and the like suffer the worst. It does work really well on the same LAN, though: I decided to use a tiny SteamLink for my living room gaming needs instead of messing around with long HDMI cables and the added fiddliness of having to directly use the desktop launch games.
I wouldn't really sign up and pay for a service operated by Google, tho.
By the way, while a future affordable game streaming service that works well might lower the bar of entry to the hobby, I have a serious fear of such a thing enabling the worst parts of the video game industry to take over. The software-as-a-service model and renouncing the last vestiges of actually owning a copy of a game seems like a terribly tempting way to turn the entire industry off of actual creativity and onto even more "whale-chasing".
[0]: https://news.ycombinator.com/item?id=18115445
[1]: https://projectstream.google.com/aco/signup
[2]: https://blog.google/technology/developers/pushing-limits-str...
Sidenote: can't stand how much space the header and footer (related articles) takes up on that page on desktop. The scroll-up-to-view-the-header pattern is disgusting with an header that big.
https://kotaku.com/in-japan-the-nintendo-switch-is-streaming...
Instead of ActiveX I wonder what would be an appropriate way to describe it.
it is not on liquidating (at least not yet)
OnLive came out in 2010. I assume that there has been a significant improvement in average consumer Internet connection speed in the last ~8 years.
Google partnering with Ubisoft is promising here, but the question will be whether they could get the likes of Activision, EA, Square Enix, Capcom, Konami, Take Two, etc to work with this service as well. If they can't, they'll end up with the streaming equivalent of Origin or what not, while whoever does get everyone on board will take the market.
Seriously, if what's the goal? Being a supporting network for gaming infrastructure?
HN loves to point fingers when company stands still. Heck, we made a teaching moment out of Kodak and others who failed to innovate and relied too much on their cashcows.
On the other hand, we proclaim nice-sounding ideas such as being aware of sunk cost fallacy and how everybody should say no more. We talk about being data-driven and dont go by what gut says. Beware of the vocal minority. Dont be afraid to pivot or let go. We tell others dont wait for your products to be perfect to launch. Get it out there, test the market and iterate.
HN has their biases (I can build that in a weekend!) but the recent discourse of "OMG! they are going to desecrate it" is very silly.
Joke? There's no need for the false dichotomy.
Google in particular has the organizational attention span of a goldfish on meth, and many of us have been burned one, two, or three times by their recurrent habits of setting something awesome up and then killing it off a short time later with no sane alternatives in place.
This reputation is entirely earned[1]. The safe thing to do if you don't want to sign up for a sudden migration headache in the future is assume that any new Google product is another symptom of their "throw everything at the wall and see what sticks" mentality, and that it won't last for more than a year or two.
[1]: https://en.wikipedia.org/wiki/List_of_Google_products#Discon...
(For bonus points, do a mental exercise and note how many of the "absorptions" lack the features of their standalone counterparts)
This is a significant distinction. It's also a massive deterrent. Who builds on experimental? Google isn't incentivized to do this. So, everything new Google puts out is hence, experimental.
It's not an old joke. You're just getting tired of it.
You're basically invalidating all of these engineers who believe in a thing called "trust," and undervaluing it with a statement like this.
You say that like it's a bad thing, but isn't that effectively how Silicon Valley as a whole operates? Most startups putter along for a few years before running out of money and getting shuttered. A couple become surprise successes and take off.
Google is effectively just a single corporation acting like a hive of startups.
Honestly there's a really good blog post/article waiting to be written grouping and classifying shuttered google projects into groups based on purposes, because for a lot of them, they aren't created to be successes or make money, but for other business reasons, legal precedents, creating competition or disrupting a particular business segment, etc.
Admittedly, there are other people in this space, but I figure they'd find a way to make the incentives pretty hard to pass up.
How could it possibly fail?
Their blog post has more information: https://blog.google/technology/developers/pushing-limits-str...
As well as the help and FAQ support articles: https://support.google.com/projectstream/
EDIT: For people who are interested in this, I would recommend studying the history of AT&T, especially the 1956 consent decree.
Is there other providers of similar services?
“Project” also make it sound like a ongoing effort of sorts, and not a (finished) product.
The gap between a "test" and a product is huge. I don't think it's wise to dump resources into building a product based on unproven technology.
Also, at least for now it really isn't a finished product. Perhaps they'll re-brand if the test pans out and they actually start selling a service, or maybe they'll just keep the moniker like Project Fi has.
