Hacker News Reader

Comment by Hamuko | Hacker News Reader

>some kind of proprietary communication stack for latency reasons

It was in fact directly using Wi-Fi because of latency.

Knowing that latency was such a problem that they had to take extreme measures like this to save even the tiniest bit of latency really should have been the end of this project.

"Having engineering problems to solve is proof that you shouldn't try to do anything."

Why can't we just appreciate a creative solution to a problem?

It's the most straightforward solution to the problem, too.

Your local machine doesn't need to know the input for stadia to work, so why redirect it there?

Also seems like a benefit for being able to use "any" screen to game on. The screen doesn't need to know how Stadia works or interface with the game controller, it just needs to receive video and display it.

> "Having engineering problems to solve

Stadia had a fundamental physics problem, not an engineering problem. Their "solution" is little more than a bandaid that highlights the fact that the problem exists.

Between USB polling rates, Bluetooth latency and OS processing, it's no wonder they wanted to avoid that entire stack and connected straight to Stadia. It's not a band-aid, those things together can easily result in 20ms of latency, which is an order of magnitude more than the latency of a fiber connection to Stadia servers.

What fundamental physics problem? Plenty of gamers, even competitive FPS players, found Stadia quite a good experience from a technical standpoint. Best in class, as far as streaming game services go.

Or you know, they could just avoid a whole class of compatibility issues due to not having to connect to a local device (which could be any of the mobile phones, Chromecasts, Android tvs or PCs with any of the four major OSes).

But sure, tell us more with your expertise.

The technology was actually never the problem. It worked great.

Since when Wi-Fi is an extreme measure?

What's extreme is having to connect a peripheral directly to the internet because you need to eke out every last millisecond of latency. The fact that they felt this was necessary is directly related to the reason it failed; because the whole system was inherently susceptible to latency unless you lived very near to one of their data-centers. The product failed mostly for this reason. Everybody responding to me is acting like it was an obvious solution to the problem, but it wasn't a solution to the problem at all. It was an absurd bandaid placed over a gaping wound.

I don’t know why this Bluetooth thing triggers you so much. Only you seem to think that Google spent all of their engineering and marketing towards that one feature.

It added few ms latency to a product that requires low latency. It was easy to avoid it (having controllers talk Wi-Fi on top of Bluetooth is hardly a rocket science). It was likely one of many small projects to improve it.

> The product failed mostly for this reason.

Citation needed. That doesn’t seem to be the case as far as I know. GeForce Now does the same thing and they’re making money.

I guess optimization really is the root of all evil

It worked well for me, very few issues

Reply on news.ycombinator.com