Google Stadia: Leaded Documents Explain Its Failure (Porting to Linux a Factor) | Hacker News Reader

Google Stadia: Leaded Documents Explain Its Failure (Porting to Linux a Factor)

boilingsteam.com

14 pointsekianjo10 comments

Yeah, I mean, Google was no Valve. The secret was to avoid porting. The secret was to just be awesome at running DirectX on Linux.

It does seem like Vulkan is slowly slowly taking over. The bitter bitter switching cost isn't going to go away, but it seems like there's finally strong momentum back to there being more than one choice. That wasn't really available in Stadia's time, but slowly this is happening.

I think Apple has a chance to make a dent with Metal and AAA games with the hundreds of millions of Apple Silicon SoCs with a powerful GPU sold each year.

You have a point. How’s gaming on the Apple TV going? And Apple Arcade.

Apple has a chance to stay at the table as relevant because they stopped mandating you target Mac & pulled a Valve & invested effort into WINE. Apple rides in on the coattails of success. https://www.macstories.net/linked/apples-game-porting-toolki...

Until recently Apple hasn't really had a chance, even if their hardware was 5x better.

There's attempts like Meta's too, to become the intermediary layer, IGL. https://github.com/facebook/igl

You had to pay for the monthly service, as well as PURCHASE the games. I was an early adopter and when I realized this was the business model I knew it was not going to succeed. $10/mo + $60 for a game, and as soon as I stop paying for the service I can't access my purchased game anymore.

> You had to pay for the monthly service, as well as PURCHASE the games

This was never true. You could do _either_ to access the platform.

Just like you pay extra for Netflix, high bandwidth 4K and HDR content was part of the $10/mo subscription.

If you bought a game, you owned that game, no monthly fee required to play it at FHD anytime and anywhere (with Internet, of course).

There was the additional benefit of the $10/mo subscription also giving you new games monthly, without buying them at full price (as you were saying, spending $60 per game). You would have access to these games so long as you maintained a membership.

You could have only ever paid $10/mo and had access to a wide array of games over time.

Conversely you could never pay for the subscription and just enjoy PS4-level gaming through your phone, laptop, or Chromecast for the games you bought a la carte. Pay $60 for one game, one time, and put in as many hours as you'd like, no console purchase necessary.

The failure of their marketing is pretty obvious, since as an early adopter you're still saying things like NEEDING to do both.

I bought my games and then it was free to stream.

I loved it

They showed really really cool features and never delivered.

It could have had a huge pull effect.

Imagine being able to play were your streamer is or were a YouTube video is showing you something.

Create a snapshot and share your game state.

Perfect for tutorials, demos, indepth videos.

Then pivot to more.

We talk about Google after all!!! Unreal engine could have enabled it by default, unity too.

I think by far the biggest factor was a lack of actual commitment. As the article says, Google has extremely deep pockets but, when evaluating how effective they are, you must also consider that they do not open them widely or for long for anything. Breaking into this space, even if games required no porting whatsoever, would require a long period of putting up with large losses on the project before it became widely used (see epic game store, which had 10x the commitment from epic that google had on stadia, had basically zero porting effort required, and is still yet to make a profit), and google is just not capable of that, it seems (the perception that google does this is of course a partly self-fulfilling prophecy).

> Windows licensing fees were cost prohibitive for the long term viability of the service

Good that nothing threatens that long term viability

Btw, why not offer both? Windows to grow the initial catalogue faster since breadth and depth are critical, and than phase it out to save on licensing costs when enough is ported to Linux?

Reply on news.ycombinator.com