I remember having this argument 20 years ago with a bunch of people talking about the great graphics of some games. My response was always “yeah but the gameplay isn’t good.”

I’ll take a pretty game, sure. But I’ll take 2600-level graphics and good gameplay over a lot of the AAA garbage we’re being fed these days.

Many more companies than Valve are making financial investments into Linux gaming, including companies that own various Linux distributions (Red Hat, Canonical, etc.), CodeWeavers (who amongst other things have been contracted by Valve on a lot of Proton work) and to a lesser extent Humble Bundle.

You’re making my point for me though. Each of the other things you’ve suggested is more work than requires more expertise. Popping up an emulator on an existing box and dumping a ROM in there is something an intern can do.

All of these other things can be done, but they’re not as quick and simple, and that’s why we’re seeing this in the first case - Nintendo went with a quick and simple solution, and someone found a bug (it still plays Windows noises).

I take it you’ve never ported an application to a different platform running on a different hardware architecture before.

This looks a whole lot like it’s probably some random emulator they grabbed and full screened?

Making an FPGA for all of this is far more work than pulling an open source emulator and sticking it on a machine…

An FPGA seems like a lot of effort, but an SNES emulator running on a Raspberry Pi seems like it may have been a better option IMO.

Yeah I’m not sure this is a good precedent unless it’s attached to “only because of your monopoly position.” Like… Does that mean every app store needs to give a way for every other app store to use their selection? That’s not how it works in any other industry (I can’t buy Aldi brand cereal at Meijer).

Yeah, Epic are being pretty hypocritical, but hopefully this will result in some positive changes. I’m just shocked at how much judges are taking Apple’s side in those lawsuits.

Yeah, I had this sort of experience back with LastPass, but Bitwarden works beautifully for me.

There’s a potentially valid criticism if that occurs because iOS’s mechanism is robust to poorly implemented password managers and Android’s isn’t, but that’s also not the criticism being provided here.

Super weird. My laptop had a new enough GPU that Windows didn’t even have proper drivers yet and it worked out of the box on Linux.

It really depends what one’s doing, also. For many things, including many games, 30fps is fine for me. But I need at least 60fps for mousing. Beyond that though I don’t notice the mouse getting smoother above 60fps, but some games I do have a better experience at 120fps. And I’m absolutely sold on 500+ fps for simulating paper.

If someone’s saying that about 30fps they should just set their refresh rate to 30 and move their mouse.

Seriously. I’m running the same version of the same distro on machines manufactured over a decade apart. And even if my distro dropped support for my older machine in its next version, I have 10 years to find a replacement.

Pretty amazing that years of effort from massive competitors like Epic and Microsoft haven’t managed to crack this. I wonder what they’re doing wrong?

(Ok I lied. I know exactly what they’re doing wrong and there’s zero chance of them changing.)

Obviously piracy would be terrible and nobody would do it, but imagine how great GBC games would be on the Steam Deck if such a thing were possible! I’d probably play those pretty regularly!

So “warm plugging” is a thing - it means a piece of hardware is detachable while the machine is asleep.

In cases like what you listed I believe they normally mark it as playable with a note that the button names might not align.

Yeah, I think their rating of unsupported is correct. It’s pretty easy to play with the default configuration though (touchpad as mouse).

Which one of the three games I listed are you talking about?

I don’t want to argue the semantics of whether or not it is a handheld PC

And yet that’s exactly what you’re doing, and missing the forest for the trees as you do so…

If it’s “just a handheld PC,” how do I setup cups as a service that starts on boot?

The answer: as a minimal change I have to enter developer mode and understand that the next update will wipe that away. (If I want to use it for this purpose regularly, it’d be better to install another OS, at which point you might as well call the Switch a handheld PC too.) This is a limitation of the Deck, but it’s a very intentional one. It’s a gaming appliance first and foremost, which makes the comparison to other gaming appliances apropos.

Are “Steam Deck has X many games that Valve have certified on it, up from Y 17 minutes ago” articles particularly useful? Only really to people who are using them for ad revenue. But that doesn’t make the comparison to the Switch a bad one. In fact, comparing the Steam Deck to the Switch is a better comparison than comparing it to a gaming desktop.

Form factor very clearly isn’t the only consideration though, and when combined with the other factors it becomes very useful. The comparison here is of two devices for playing video games, and form factor makes a difference there. Back in the early 2000s, people weren’t really comparing the GBA to the PS2, the Xbox, the GameCube and the Dreamcast. These simply were different markets. Likewise, the Steam Deck and the Switch are more comparable than the Deck and the PS5 specifically because of form factor, even though as computing devices the Deck and the PS5 are more similar (both are running custom AMD Zen 2 CPUs with custom RDNA 2 GPUs, for example). The Steam Deck is, in practice, only slightly more “a PC” than the PS4 is.

Doesn’t matter for how they’re measuring. None of those games are marked as either Verified or Playable by Valve, so they’re not included in these statistics, despite working fine on the Deck.

Yeah, a few examples from my own Deck include Rymdkapsel (works perfectly, not verified or marked as Chromebook Ready), Surviving Mars (playable, but not rated) and Turmoil (explicitly unsupported, but works fine).

Are you trying to equivocate motherboard form factor (a specific terminology used for the sizes and locations of screw-holes in motherboards) with entirely different types of device that happen to have similar shapes and sizes?

The motherboards in those computers could have the same motherboard form factor, but that doesn’t make comparing a rack-mounted router with specific design constraints to your gaming PC a reasonable thing to do. Your gaming PC is, most likely, far better at its job than a router in a data centre would be, and the DC router is most likely far better at its job than your gaming PC would be, because they’re totally different categories of device. Likewise, a Wifi router and an Xbox are totally different categories of device. Even discussing just form factor alone, the Wikipedia summary you posted includes:

other physical specifications of components

An Xbox lacks many of the things considered as basic functionality for a router, such as a second Ethernet port. Likewise, a wifi router tends to lack many of the things considered crucial for modern game systems, such as a GPU and video output. In both cases it’s perfectly possible (at least in theory, though I’m sure at least one person has actually done it) to reconfigure one for the alternative purpose, but that is utilising the device well outside of its design parameters.

What I’m saying is that the PC comparison simply misses the point, whereas the Switch comparison is a comparison of how the devices are intended to be used (and for the most part actually used). The Steam Deck is not a good device for running a development environment or most of the things one thinks of as “PC tasks.” It’s not designed for that.

One can use it in that way, but one of the biggest differences between the Steam Deck and its “more pc-ish” competitors like the Ally is that Valve has done a lot of work to make desktop mode unnecessary for the vast majority of users. The user experience puts it much more directly in competition with the Switch than with other handheld PCs, and that’s a strength of the Deck, not a weakness.

One can install Linux on a Switch too, at which point it’s basically an ARM-based handheld PC. But a reviewer who reviewed the Switch for its power as a handheld Linux machine would be missing the point too.

While the Steam Deck is technically a handheld PC, as a Linux enthusiast who’s tried to use the desktop mode for laptoppy things… No it isn’t.

It works in a pinch (well… Not for my job, but I also don’t expect Valve to put lxd into SteamOS), but the comparison to the Switch (which I also own) is much better than comparing it to even a gaming laptop. In fact, if I were the type of person to emulate stuff (which, don’t worry Nintendo, I totally am not), I would say my Steam Deck makes a better switch than my switch. If I’d emulated a Switch to play Mario Kart (which obviously I haven’t) I’d say it was a better experience on my Steam Deck than on my switch.

Here’s a real-world use case that also won’t require insane GPU power.

Here’s a real-world use case where this difference is noticeable to the average person. We don’t need to render video games at 1000 Hz, but many things that can be rendered with comparatively low GPU power could be made a better experience with it. The real question is whether/when the technology becomes cheap enough to be practical to use in consumer goods.

Here’s a big part of why they want 1000Hz. You don’t need to fully re-render each frame for most cases where 1ms latency is desirable - make a 100 Hz (or even 50 Hz) background and then render a transparent layer over it.

I think this is mostly just practical. Tainted kernels are more problematic for a variety of reasons these days (thinking about things like TPMs), so moving that stuff out of the kernel lets NVIDIA keep their proprietary drivers without sacrificing certain use cases.

I’m more interested in seeing NVIDIA hardware usable with mesa, which would be a much longer term project.

I thought one of the main advantages of sodium-ion batteries was price? Great for the applications you listed

I’ve had three sets of earbuds where the right one has died of no apparent cause shortly after the warranty expired.

I’m starting to think it’s intentional.

Valve is making money on Proton (if indirectly), so they would very quickly get sued for it. It’s not that the copyright owners would want to stop them from distributing the library - just that they’d want a cut of Valve’s massive revenues. A community project, on the other hand, isn’t likely to be able to provide any revenue. So since it’s just redistributing runtimes that are already available for free and the only likely result is the project getting taken down, it’s not really worth doing.