I would say this is a win for everyone. It sucks we lost Stadia. But the way Google has handled shutting it down should be seen as a role model for any cloud gaming service. They have refunded me for every game I bought and now this.
For anyone curious, the Stadia controller is one of the better controllers released in awhile. I would say its on par with the Xbox Series controller.
I remember when The Verge was on a campaign to shit on Stadia, and few weeks later were ready promoting "XCloud" and spreading messages like "cloud gaming is the future of gaming and it is inevitable", totally ignoring Geforce Now and Amazon's Luna, that's great journalism
Stadia still has the best tech, and its latency is unmatched, you only needed an URL to start playing game, they had a web interface meanwhile Microsoft was complaining because apple blocked their mobile app and they didn't have a web interface to counter apple's move, that was funny to witness, Tom Warren definitely has some conflict of interest going on
Also had pretty awful support, but Nvidia SHIELD wireless controller was also a wifi controller. It also had audio. No USB port though, iirc! I forget who, but I believe the whole thing was some 3rd party chip that did almost all the work on the chip, Nvidia was just like, the only name to ship it.
I'd love to see wifi start to make some kind of incursion into device/peripheral space. But there's not really a starting place is there? Bluetooth is typically point-to-point (albeit it has "mesh" capabilities too), wifi is typically point-to-network (even wifi-p2p is actually just negotiating an ad-hoc group, but functionally can be seen as point-to-point). But more so than the link layer, there's just not the support, the idea of how we'd peer/pair.
The last thing I'd want to see is more low end PAN clients piling into being Wi-Fi, it's already congested enough! The best part about bluetooth is it at least made attempts to minimize the impact of having a bunch of low end devices using the same air space.
wifi 6 is supposed to drastically improve the situation.
devices dont all share the same base frequency, they are spread out across the spectrum. so one slow device doesnt hog all the channels.
target wake time allows better coordination of when devices send, so AP can do some scheduling to prevent congestion within their network.
better power control also means devices should, in most cases, be using less tx power.
wifi 6 greatly greatly greatly enhances spectral efficiency, & i think with it the idea of having PAN clients seems entertainable if not outright appealing. we're already seeing mcu's like esp32-c6 make the switch, & more should follow- the potential power saving is too alluring to pass up, and it avoids so many of the issues wifi had with all mimo channels being blocked while one single channel slow device puttered along.
The esp32-6 is a great example of why these types of clients are never good Wi-Fi citizens - it supports about half the Wi-Fi 6 enhancements a normal client would, does it with a single antenna, and only does it in 2.4 GHz. Better than previous microcontrollers doing 802.11n for sure but it still includes Bluetooth anyways and Bluetooth has more powersaving/congestion avoidance features unless you need lots of data (which this is not good at doing even with Wi-Fi). Even with things like RUs Wi-Fi is still only as good as the worst clients in the airspace let it be.
802.11ah (Wi-Fi HaLow) was intended to serve these types of devices in a separate spectrum with a very different set of features (and 802.11ax later brought some of these features over for normal clients to take advantage of but ignored many specific to IoT only use cases) but it never caught on vs LoRa and Bluetooth.
Do you have any links talking about what wifi-6 capabilities are missing?
I was under the impression capabilities like OFDMA were mandatory on uplink & downlink for "WiFi 6". Ditto for Target Wake Time.
In general, while wifi might not always be the best pick, I also see it working very well in many challenging environments, and I think the expectation that low end devices are necessarily bad for the network isnt going to hold forever. The legacy 802.11b devices have left a lot of emotional scars & pain, and wifi 4 & 5 had many limitations that made low-end devices a network tax, but it seems like we really finally are on the way to better.
I was aware of 802.11ah somewhat, but admit I had forgotten just how many tweaks/optizations went into it. Time division/grouping I'd forgotten about. TWT seems like a good help here. It also seems to lack the key thing that should hopefully be decreasing congrestion, which is ppwer control, which should hopefully greatly reduce congestion in most scenarios, by making devices not always be shoutintg at each other.
Admittedly it seems like a large number of esp32 users do struggle woth throughput, which sucks. A quick scan around the net & Im seeing very few users get above 12Mbps. Many see huge gains from improving bad board design, which seems to compound the issue. WiFi 6 has great potential, but as usual, there's many paths abound to screw up a good thing.
Great. However the real interesting feature was that the controller connected directly to the end server. It'd be great to be able to set that end point
For anyone curious, the Stadia controller is one of the better controllers released in awhile. I would say its on par with the Xbox Series controller.