It's a bit scary to be honest. Linux is fairly easy to use because it piggy backs on the popularity and relative standardization of PCs thanks to Windows. If Windows were to go fully cloud-based and computers would become simple terminals, the availability of cheap Linux machines could suffer. Not a likely scenario, probably more a microsoft wet dream.

As long as the speed of light in fiberglass is what it is, Gamers will be enough of a Market to prevent that.

On the other hand, cloud gaming has been more of a market than cloud PCs for consumers.

Because you need a basic PC to connect to it which is enough for most usecases anyway. But not for gaming so there it makes more sense.

Cloud gaming will always be hampered by the simple reality of really of poor internet connections being very common.

20m people use Xbox cloud gaming. NVIDIA GeForce Now has 25m users.

I'd say those are some stats worthy of appreciation.

Worthy of appreciation maybe but no one is gonna reasonably play things like CSGO on a cloud server. The difference due to latency is instantly noticeable and very obvious.

I addressed this here: https://news.ycombinator.com/item?id=36057449

It's also bold to assume that everybody plays twitchy FPS. CSGO runs on Linux, the Steam Deck uses Linux in the OOBE, I'm sure that if push comes to shove and MS stops offering native Windows, people who want to play CSGO will just use Linux in the same way that people use emulators and FPGAs and CRT PVMs (hello) to play what they want and have the best experience.

However, what's probably more likely is that latency-dependent games becomes less and less popular if cloud gaming takes over, because more people are probably going to be playing a game that's included in their cheap monthly subscription rather than spending $XY. Making a twitchy niche latency-dependent game makes less sense than reaching as many people as possible where they're playing.

I can't imagine a world where Microsoft kills native Windows though, it just doesn't make sense. How environmentally friendly it is or is not is also a consideration with streaming a desktop.

> It's also bold to assume that everybody plays twitchy FPS.

Never said "everybody" but enough do. https://steamcharts.com/app/730 Current active players 1,458,036 for CSGO alone

But this doesn't negate the fact that cloud gaming works for a lot of genres and a lot of people. That market will likely only grow, and in that scenario, will likely become what most publishers try to cater for as that's where the worthwhile money is.

Look at how many competitive FPS have tried to take on CSGO and have failed since its launch. CSGO is very much the gold-standard of its genre.

Despite modern PC gaming largely being a better experience (minus bad ports, buggy launches), consoles still sell in huge numbers. If cloud gaming grows, I expect it to replace consoles for a lot of people.

Even now, rather than download a game to my living room Xbox, I'll just stream the game via the cloud to it. However, if I'm sat up close with a game at my desk, of course I'll install it on my PC. But at x feet away from my TV in the living room, whatever is lacking visually or in terms of latency, I'm not gonna care about when my input device isn't exactly all that performant. I dare say that about sums it up for a lot of people using the cloud to game.

This is without even bringing up the number of Nintendo Switch games that are streaming-only as the console can't handle it natively. Doesn't seem to bother Switch owners much.

Wall of text, short answer: Both will continue to exist and as long as enough people want to play games like CSGO we will have local pc gaming and thus linux desktop systems

Linux is part of the problem. Linux never has been an OS for personal computer and it's security model is backward and even nefarious for this purpose. Linux, like all Unix and Unix-like is an OS for a big central computer connected to many dumb terminals. What happens here is Windows taking lessons from Linux and Unix. If you want to save the idea of a personal computer you want an OS that empowers users instead of protecting the computer from them.

That is exactly the reason for moving everything to cloud and SaaS: software that is being bought is paid once (when not free), while the one hosted remotely is paid for use, many times, even if it's free (hosting costs). Doing the same thing with operating systems is just the next iteration. We're slowly going back to the mainframes era: a central server that holds everything surrounded by dumb terminals that can't work without connection to the above server. To me this equals to throwing down the toilet 50 years of IT advancement and open networks culture in the name of profit.

Is there an actual technical reason for doing this besides the subscription model success? Better analytics?

In _theory_ using a thin-client and booting into a cloud machine means you can "upgrade" your machine in a couple years w/o having to replace the thin-client.

Could be easier for grandma, could reduce e-waste, will definitely be used to pigeon-hole us into an endless hell of subscription software.

In practice it means your compute time is being micro-time sliced to the maximum so it will probably perform incredibly bad.

It won't happen for a lot of industries. I get a rock solid 0.7ms roundtrip latency using an RME RayDat. I can't play guitar decently with anything more than 4ms of latency.

Hybrid? Maybe. But even so, I think it'll still make more sense to just own the hardware and not rely on WANs with variable latency.

For the average schmoe or office worker? Probably makes sense to have it on the cloud, especially as how well Xbox cloud gaming works for most people.

Something about windows copilot screems either subscription or ads