SN2716057That's one thing about all the Linux distros, don't like one switch to another. Which I'm gonna do cause Ubuntu is giving me a headache, let's hope Bazzite will be a breeze.

It has been for me so far, so good luck! :)

Edit: So far, Bazzite has been the only distro that recognised my monitor as Freesync + HDR right from the start without any tinkering. I think this says something.

This is exactly the main problem of why Linux has not already became much more popular - too many distros.
Devs don't agree on something - they create own fork. Such approach is the worst nightmare of hardware makers.
Yeah, I know some distros share the same base (Debian, Arch, etc.) but still, there are too many of them.

Every day some people are born, some die. And apart from that, every day a new Linux distro is made as well. :kookoo:

Since Valve is directly cooperating with Archlinux devs, this is probably the best way to go if you want to achieve maximum SteamOS support.
But I have seen with my own eyes that even Debian-Ubuntu-based Mint is doing pretty well with Steam games. Proton does it's job.
Anyway, people, if you want to go with Linux, choose something popular backed by good community.

I can assure you that, unlike Windows 12 is going to be final final final final version of Windows 10 (11 and 12 were not suppossed to exist),
there won't be more than 2 major versions of SteamOS. Because when it comes to games and software, GabeN can't count to three and above. :D

LittleBroThis is exactly the main problem of why Linux has not already became much more popular - too many distros.
Devs don't agree on something - they create own fork. Such approach is the worst nightmare of hardware makers.
Yeah, I know some distros share the same base (Debian, Arch, etc.) but still, there are too many of them.

Every day some people are born, some die. And apart from that, every day a new Linux distro is made as well. :kookoo:

Since Valve is directly cooperating with Archlinux devs, this is probably the best way to go if you want to achieve maximum SteamOS support.
But I have seen with my own eyes that even Debian-Ubuntu-based Mint is doing pretty well with Steam games. Proton does it's job.
Anyway, people, if you want to go with Linux, choose something popular backed by good community.

I can assure you that, unlike Windows 12 is going to be final final final final version of Windows 10 (11 and 12 were not suppossed to exist),
there won't be more than 2 major versions of SteamOS. Because when it comes to games and software, GabeN can't count to three and above. :D

Thanks to Proton, the distro you choose doesn't matter too much in terms of game compatibility these days. But I see your point, and I agree.

LittleBroThis is exactly the main problem of why Linux has not already became much more popular

It probably plays a part but I think the biggest reason would be that the other brands just pay the hardware companies to use their OS.

I am very excited for this because it means that more gaming devices will be running Linux, which will lead more devs toward developing support for it! :clap:

AusWolfIt has been for me so far, so good luck! :)

Edit: So far, Bazzite has been the only distro that recognised my monitor as Freesync + HDR right from the start without any tinkering. I think this says something.

Yeah, I did a little reading about Bazzite after you started posting about your experiences. Looks really good. Doesn't carry the caveats that small/niche distros usually do, because it isn't really a distro; rather it's just a customized image of Fedora Silverblue/Kinoite. Hypothetically the entire Bazzite team could quit tomorrow and your install would keep trucking along, fully maintained. I think this will become my go-to recommendation for gamers who are Linux-curious.

Bazzite's gaming stuff is evidently packaged in anArch-based distrobox container, which btw you can install on any distro if you want. Pretty cool.

Anyway, I think the Freesync + HDR thing is due to KDE Wayland. On X11 you typically have to jump through some minor hoops to enable Freesync. But KDE Wayland auto-detects everything. Pretty great. Wayland of course exists in other forms, but the ecosystem is quite fractured. Gnome and KDE seem to have the most mature implementations.

Now that I think about it, SteamOS has never been exclusive. You can download an installer and install it on anything any time.

AusWolfNow that I think about it, SteamOS has never been exclusive. You can download an installer and install it on anything any time.

I don't remember if Valve ever published a SteamOS image for people to do so, though?

wNotyarDI don't remember if Valve ever published a SteamOS image for people to do so, though?

It's here.
www.google.com/url?sa=t&source=web&rct=j&opi=89978449&url=https://store.steampowered.com/steamos/download%3Fver%3Dsteamdeck&ved=2ahUKEwjz46zhjeqKAxUzTUEAHT_IJfwQjBB6BAgmEAQ&usg=AOvVaw0GX5sK4oauDoym4SMeT-0r

vvtunesSteamOS may be the sneaky contraption that will slowly replace Windows as the primary gaming OS. Never loose hope :D

The steam deck (SteamOS) will be helping linux compat no end, in the FF7 modding community, we integrated linux support for the mod manager and also added 16:10 support solely because the steam deck exists.

LittleBroThis is exactly the main problem of why Linux has not already became much more popular - too many distros.
Devs don't agree on something - they create own fork. Such approach is the worst nightmare of hardware makers.
Yeah, I know some distros share the same base (Debian, Arch, etc.) but still, there are too many of them.

Hardware is spoken to on a kernel level. I don't see how distro forks cause an issue there. I also don't see how the existence of forks prevent people from switching. In theory, you can choose a distro that is optimised for your needs (eg. easy to install, good initial support for gaming, ...). But I agree with there being too many pointless distro forks.

What is probably causing more issues is the fact that there is not really a unified GUI toolkit, and even if there was one, it would be different from the Windows one (or Max OS X for that matter). Therefore causing significant extra work to get an application looking good and similar on multiple OSes. But this is not an issue for gaming either, since the game engine is its own GUI toolkit so to speak. In fact gaming should be one of the easier types of software to port to Linux, since it basically only requires the game engine itself to be Linux compatible, and games built upon a Linux compatible engine would almost automatically run on both OSs. Since we even have a unified rendering API amongst the OSs (Vulkan), it's really just laziness from the devs preventing Linux binaries.

TimbalooHardware is spoken to on a kernel level. I don't see how distro forks cause an issue there. I also don't see how the existence of forks prevent people from switching. In theory, you can choose a distro that is optimised for your needs (eg. easy to install, good initial support for gaming, ...). But I agree with there being too many pointless distro forks.

It is a problem for one reason: lack of knowledge.

People don't know what the different distros are, or which one they need, or that they're not really that different for most things. It's easy to get lost in the forest.

People also don't know that it's easy to get an experience similar to other OSes thanks to Wine, Proton and the different desktops which are available for various different distros. Even if you somehow manage to figure out which distro you want, you still get confused choosing your desktop.

People also don't know that you don't need special skills to use Linux. You have the same keyboard, the same mouse, the same click functionality, it works with a GUI, and so on. The idea people have about Linux is a bunch of pizza-munching fat geeks who never leave the house typing frivolously in the terminal all day and night.

Speaking of which, people also don't know where to ask for help. Sure, Linux is a community OS, but what is the community? Even I don't know outside of TPU. And if you happen to know, the answers you usually get for your questions begin with "open the terminal and..." even when there are way easier GUI solutions available as well.

This all paints the picture of a quirky OS made only for true nerds in the eyes of common people. Hardcore Linux aficionados don't get this for some reason. They keep insisting that it's easy, which it really is once you got the above figured out.

AusWolfIt is a problem for one reason: lack of knowledge.

People don't know what the different distros are, or which one they need, or that they're not really that different for most things. It's easy to get lost in the forest.

People also don't know that it's easy to get an experience similar to other OSes thanks to Wine, Proton and the different desktops which are available for various different distros. Even if you somehow manage to figure out which distro you want, you still get confused choosing your desktop.

People also don't know that you don't need special skills to use Linux. You have the same keyboard, the same mouse, the same click functionality, it works with a GUI, and so on. The idea people have about Linux is a bunch of pizza-munching fat geeks who never leave the house typing frivolously in the terminal all day and night.

Speaking of which, people also don't know where to ask for help. Sure, Linux is a community OS, but what is the community? Even I don't know outside of TPU. And if you happen to know, the answers you usually get for your questions begin with "open the terminal and..." even when there are way easier GUI solutions available as well.

This all paints the picture of a quirky OS made only for true nerds in the eyes of common people. Hardcore Linux aficionados don't get this for some reason. They keep insisting that it's easy, which it really is once you got the above figured out.

Unsavvy people still hear "Linux" and imagine green phosphorus command lines. Modern Linux you can go by without ever using the CLI.

wNotyarDUnsavvy people still hear "Linux" and imagine green phosphorus command lines. Modern Linux you can go by without ever using the CLI.

And that's why I love it. But unsavvy people class me together with the savvy ones and think that "Terminal" is my favourite game. It couldn't be further from the truth. The only command I know is neofetch, and I'd like to keep it that way. :laugh:

If you do find yourself having to use the terminal--and I understand why that's undesirable for a lot of people--I highly recommend tldr. It gives you quick reference for the relevant command with a few simple examples and explanations. This reference is much more helpful for casual users than the man pages, which are fantastically detailed, but their density makes them a chore to wade through.

For example, the man page for "cp" looks like this. By contrast, "tldr cp" outputs the following:

General users, even what you might call "casual power users" (which is basically just a euphemism for "gamers") don't have to use the terminal on Linux these days though. Maybe once in a blue moon. There's usually a GUI solution. The trick is in knowing about the GUI solution, lol.

I use the terminal a fair bit myself, but these days rarely on the machine I'm sitting at--usually I do it when managing a remote machine over ssh. Even for that purpose, you could probably make do with a GUI interface, like Cockpit, most of the time.

Probably 90% of my posts on this forum concerning Linux are made with the aim of informing people about this-or-that GUI solution. The intent isn't to position myself as an expert, or to evangelize about Linux generally, though I do think it's a fantastic (and free!) product. Rather my intent is to simplify some of the pain points I went through. Auswolf is exactly right about Linux advice on the internet. It's mostly terminal copy-pasta, which is fine and good and (usually) accurate/effective, but it can also be intimidating to new users. Also, if you're like me, seeing long terminal commands proffered as solutions will prompt you to figure out what the terminal commands are intended to do before you commit to them. This can add a great deal of overhead to what might already be a frustrating troubleshooting exercise.

On the other hand, there are tasks for which the terminal is undoubtedly the most efficient option.

EDIT: I should add, to bring this rambling back on topic, that GUI tools are easier to find these days, the vast majority of them included as defaults. The amount of progress Linux has made in the last few years, with respect to user-friendliness, cannot be overstated. And that progress owes in large part to the influence of Valve and the Steam Deck.

Back when SteamOS was first announced, with the Steam Machines, I was very hopeful for more of a "HTPC OS" to finally end the pointless console wars.

I'm glad the Steam Deck has kept the SteamOS alive, yet I think for many these mobile gaming devices are more of a gimmick than anything else. Like how many would buy a $600-1000 gaming device that breaks after a few years? Probably a lot of teenagers, but who in their 30s and 40s can stand gaming on these more than a few minutes at the time? :P

AusWolfI've already switched over to Linux on my primary gaming rig, and I can wholeheartedly recommend everyone to do the same. It's a much smoother experience than it was 10 years ago, and it's much nicer for general use than Windows. Feels snappier, too.

There is no question that Linux is way more responsive general use, and Windows 11 isn't responsive no matter how powerful your hardware may be, even on a Raptor Lake the whole OS is jerky, especially keyboard input in all applications like Visual Studio, Notepad etc. Having used every major version since 3.0, Windows is clearly worse than ever. At home you can certainly mitigate some of these issues by disabling spyware etc., but you probably can't do that stuff on your work computer.

But as for gaming, we should keep our eyes on the target - to get gaming companies to use Vulkan and compile games for Linux natively. Wine/Proton is certainly impressive work, but from what I've seen it's still a hit and miss when it comes to individual games. While I haven't tried any top games launched from the last four years in Wine, but from what I've seen having a "top rating" in Steam with Proton doesn't guarantee it will work at all.

TimbalooHardware is spoken to on a kernel level. I don't see how distro forks cause an issue there. I also don't see how the existence of forks prevent people from switching. In theory, you can choose a distro that is optimised for your needs (eg. easy to install, good initial support for gaming, ...). But I agree with there being too many pointless distro forks.

From the user perspective, pretty much any major (regular) distro and derived distros should be fine.
From the developer's perspective shipping software to be deployed on all kinds of distros requires some effort that a Windows dev wouldn't be used to; like do you ship just a tar-ball, a shell-script that's distro independent, or do you want like deb, rpm packages, or perhaps snap(with its own issues)?

TimbalooWhat is probably causing more issues is the fact that there is not really a unified GUI toolkit, and even if there was one, it would be different from the Windows one (or Max OS X for that matter). Therefore causing significant extra work to get an application looking good and similar on multiple OSes.

I would argue the options for Windows aren't that great either since they moved on from win32, and MS has never been good with API versions, so having a complex GUI application for Windows means a lot of workarounds or rewrites for different Windows versions. While I don't know nearly all major client applications out there, I rarely see any heavy application use any of MS' GUIs(especially not .NET), it's usually one of two: 1) custom widget API built on top of OpenGL etc., or a well established widget toolkit like Qt (or sometimes GTK). But when it comes to enterprise stuff, it's usually much worse.

TimbalooBut this is not an issue for gaming either, since the game engine is its own GUI toolkit so to speak. In fact gaming should be one of the easier types of software to port to Linux, since it basically only requires the game engine itself to be Linux compatible, and games built upon a Linux compatible engine would almost automatically run on both OSs. Since we even have a unified rendering API amongst the OSs (Vulkan), it's really just laziness from the devs preventing Linux binaries.

For sure, the ingame GUI is obviously rendered in the graphics API of choice. But games still usually needs a way to show error messages and sometimes some type of launcher. Doing this with pure X11 isn't that hard, ~200 lines will give you a styled message box, the probably only issue there is that fonts doesn't look too nice.

Those that use libraries like SDL will get such features built-in, along with lots of others including audio, input handling and special controllers such as joysticks. This is comparable to the feature set DirectX offers for Windows, and is appealing to many indie developers. But SDL is more like an engine framework, as it controls some the main control flow of the engine for you and offloads it to a library, so you wouldn't get the precise timing and threading control you want for a high-performance game engine. I consider it more of a learning tool than anything else. Taking an already built game engine and porting it would be cumbersome. Using something more barebone like glfw is probably the smarter choice, or better yet build your own, which is what I've done for 20+ years with OpenGL. The others will still remain though; like IO from the OS and audio. The latter of which is an area where I think we need an upgrade/replacement of the very outdated OpenAL. Linux does have the advantage of providing a very bare metal access to the physical devices in addition to events from X11 or Wayland. One last thing outside my expertise is networking; I haven't touched network programming since the early 2000s, so I don't the state for game development there.

My point being, if your game is already in Vulkan or OpenGL, porting the graphics is very simple as you say, the other parts probably needs some effort. :)

efikkanBack when SteamOS was first announced, with the Steam Machines, I was very hopeful for more of a "HTPC OS" to finally end the pointless console wars.

I'm glad the Steam Deck has kept the SteamOS alive, yet I think for many these mobile gaming devices are more of a gimmick than anything else. Like how many would buy a $600-1000 gaming device that breaks after a few years? Probably a lot of teenagers, but who in their 30s and 40s can stand gaming on these more than a few minutes at the time? :p

Call me a child then, because I have a Steam Deck, and I find it to be awesome for indie titles that play well with a controller. I'm not really into AAA gaming these days anyway.

With that said, I mostly game on my PC, and I wouldn't give $600-1000 for a secondary device.

efikkanThere is no question that Linux is way more responsive general use, and Windows 11 isn't responsive no matter how powerful your hardware may be, even on a Raptor Lake the whole OS is jerky, especially keyboard input in all applications like Visual Studio, Notepad etc. Having used every major version since 3.0, Windows is clearly worse than ever. At home you can certainly mitigate some of these issues by disabling spyware etc., but you probably can't do that stuff on your work computer.

But as for gaming, we should keep our eyes on the target - to get gaming companies to use Vulkan and compile games for Linux natively. Wine/Proton is certainly impressive work, but from what I've seen it's still a hit and miss when it comes to individual games. While I haven't tried any top games launched from the last four years in Wine, but from what I've seen having a "top rating" in Steam with Proton doesn't guarantee it will work at all.

Proton has been more of a hit than a miss lately. I don't think I had a problem with any game recently. I honestly don't miss Windows one bit. :)

efikkanBut as for gaming, we should keep our eyes on the target - to get gaming companies to use Vulkan and compile games for Linux natively. Wine/Proton is certainly impressive work, but from what I've seen it's still a hit and miss when it comes to individual games. While I haven't tried any top games launched from the last four years in Wine, but from what I've seen having a "top rating" in Steam with Proton doesn't guarantee it will work at all.

My experience is exactly the opposite: even if a game has a Linux-native version, you're almost always better off using Proton/Wine, which works fantastically well. Probably 90% of the time, you don't have to do any tweaking, just pick a recent Proton/Wine prefix and press play. A substantial fraction of the remaining 10% can usually be resolved after five minutes of reading on ProtonDB. There are, of course, rare instances where you'll need to jump through hoops, but that's true on Windows too. What we're discussing is a difference of degree, not kind.

Steam and Heroic will cover the vast, vast majority of single-player games released in the last 15-20 years. Lutris installation scripts will cover most any of the exceptions (e.g. Rockstar games purchased through Epic). As for older stuff, it depends. In some cases, old stuff runs better on Linux, and of course the really old stuff needs to be emulated regardless (e.g. DOSBox), but there is sort of an "anti-goldilocks zone" of old-but-not-super-old games that might be a pain in the ass to set up. Something like Knights of the Old Republic's widescreen resolution patch, for example. I haven't tried to set that one up using Linux tools; it might be possible, but I couldn't be bothered investigating the matter--I just installed the resolution patch through a Windows VM. Of course, the game is still playable without widescreen. It was certainly good enough for me the first time I played it, ~23 years ago. Time flies.

There are other cases where even relatively complicated modding schemes turn out to be surprisingly simple on Linux. Baldur's Gate I, for example, illustrates my point above about Linux-native solutions: there are a ton of intricate "guides" on the internet about modding this game in Linux, and all of them are a waste of time. It turns out that the easiest and most effective way to mod BG in Linux is to pretend that you're on Windows. Just double-click the mod installers and Wine will handle the rest. UnityModManager is likewise almost effortless to set up if you ignore the readme and just add it to Steam and run it through Proton.

In keeping with the theme from my prior post, the main trick is knowing that there's a simple way to do things. That fact isn't always readily apparent. And it's no one's fault--things have changed drastically in just the last few years, particularly with respect to gaming.

WastelandMy experience is exactly the opposite: even if a game has a Linux-native version, you're almost always better off using Proton/Wine, which works fantastically well. Probably 90% of the time, you don't have to do any tweaking, just pick a recent Proton/Wine prefix and press play. A substantial fraction of the remaining 10% can usually be resolved after five minutes of reading on ProtonDB. There are, of course, rare instances where you'll need to jump through hoops, but that's true on Windows too. What we're discussing is a difference of degree, not kind.

Steam and Heroic will cover the vast, vast majority of single-player games released in the last 15-20 years. Lutris installation scripts will cover most any of the exceptions (e.g. Rockstar games purchased through Epic). In some cases, old stuff runs better on Linux, but there is sort of an "anti-goldilocks zone" of old-but-not-super-old games that might be a pain in the ass to set up. Something like Knights of the Old Republic's widescreen resolution patch, for example. I haven't tried to set that one up using Linux tools; it might be possible, but I couldn't be bothered investigating the matter--I just installed the resolution patch through a Windows VM. Of course, the game is still playable without widescreen. It was certainly good enough for me the first time I played it, ~23 years ago. Time flies.

There are other cases where even relatively complicated modding schemes turn out to be surprisingly simple on Linux. Baldur's Gate I, for example, illustrates my point above about Linux-native solutions: there are a ton of intricate "guides" on the internet about modding this game in Linux, and all of them are a waste of time. It turns out that the easiest and most effective way to mod BG in Linux is to pretend that you're on Windows. Just double-click the mod installers and Wine will handle the rest. UnityModManager is likewise almost effortless to set up if you ignore the readme and just add it to Steam and run it through Proton.

In keeping with the theme from my prior post, the main trick is knowing that there's a simple way to do things. That fact isn't always readily apparent. And it's no one's fault--things have changed drastically in just the last few years, particularly with respect to gaming.

The problem is that a lot of people don't know that there's a simple way to do things. You ask any Linux aficionado how to do X, and the answer will begin with "just open the terminal and type..." which is completely unnecessary for the vast majority of things these days. This is why people are afraid to switch. They think Linux is more complicated than it actually is.

efikkanFrom the user perspective, pretty much any major (regular) distro and derived distros should be fine.
From the developer's perspective shipping software to be deployed on all kinds of distros requires some effort that a Windows dev wouldn't be used to; like do you ship just a tar-ball, a shell-script that's distro independent, or do you want like deb, rpm packages, or perhaps snap(with its own issues)?

I would assume snap/flatpak/appimage or whatever container format is the preferred way of deploying a multi platform software on Linux as of today. You can bring your own dependencies/libs with the package, pretty much most problems solved. It's not like this is already done with windows installers anyway, a lot of software ships with a number of prerequisites (.NET, DX runtime, ..., some even ship their own trusted JVM version...). While I despise something like flatpak for open source programs, where I'd rather have the advantages of the "native" package management system of the distro, flatpak (or the likes) is a godsend for deploying closed source software, imho.

AusWolfCall me a child then, because I have a Steam Deck, and I find it to be awesome for indie titles that play well with a controller. I'm not really into AAA gaming these days anyway.
With that said, I mostly game on my PC, and I wouldn't give $600-1000 for a secondary device.

No, I'm referring to most people in their 30s and 40s who's body can't handle anything close to what they did in their youth. If you can play on such a device for hours without any issues with your neck, arms etc. then consider yourself blessed. :)

If I were to play a game like that, I would rather play it on a TV, or of course on a desktop PC.

WastelandMy experience is exactly the opposite: even if a game has a Linux-native version, you're almost always better off using Proton/Wine, which works fantastically well. Probably 90% of the time, you don't have to do any tweaking, just pick a recent Proton/Wine prefix and press play.<snip>

Then let me be more precise; I'm talking about rendering engines using Vulkan directly, vs. being written for DirectX 12 and then slapping some sloppy abstraction layer on top of it to translate it to Vulkan or OpenGL, which is what most "ports" to Linux have been over the years. Especially if this work is outsourced to a third party the end result will be garbage. Principally this isn't very different from what Wine does, nor what the DirectX 9 emulation for Intel GPUs do too, they all emulate another API by translating each API call to another. This will never work 100%, which is why we risk strange glitches and stutter, as graphics APIs are very complex state machines (do not be fooled into thinking a such API can easily be translated the way a rest API can).

TimbalooI would assume snap/flatpak/appimage or whatever container format is the preferred way of deploying a multi platform software on Linux as of today. You can bring your own dependencies/libs with the package, pretty much most problems solved. It's not like this is already done with windows installers anyway, a lot of software ships with a number of prerequisites (.NET, DX runtime, ..., some even ship their own trusted JVM version...). While I despise something like flatpak for open source programs, where I'd rather have the advantages of the "native" package management system of the distro, flatpak (or the likes) is a godsend for deploying closed source software, imho.

Yes, this is pretty much standard practice with any commercial software, whether it's for Windows or Linux. Most such applications for Windows will include even the visual studio C++ standard library. And to some of us this is very obvious, it's because that's the specific version they have spent extensive resources validating on.
(Many newer web developers wouldn't get this though; like a NodeJS application checked out with the same dependencies at different times will get completely different code, resulting in bugs that are hard to find and reproduce.)
So even newer versions of software will often not include the latest dependencies, not unless they have a good reason to, as the benefit of tried and tested is much more valuable for a quality product. It's what I would do my self too; have as few dependencies as possible, and bundle any which makes sense.

Whether it's best to bundle a specific version of a dependency or rely on the installed library in an OS depends greatly on how "standard" this dependency is, how stable it is, etc. Like there are many libraries that are crucial to the OS, so those better be validated. In a perfect world where all libraries would be stable and bug-free, this wouldn't be much of an issue. But if you sell a product you take the reliability for the customer a little more seriously, making the trade-off worth it. it has more to do with that than whether it's open source or not.

efikkanNo, I'm referring to most people in their 30s and 40s who's body can't handle anything close to what they did in their youth. If you can play on such a device for hours without any issues with your neck, arms etc. then consider yourself blessed. :)

If I were to play a game like that, I would rather play it on a TV, or of course on a desktop PC.

Ah, I see what you mean. I wouldn't use my Deck to game for half a day without a break, that would be torture. The battery doesn't last that long anyway, and having it plugged in kind of defeats the purpose of a handheld, imo.

efikkanBack when SteamOS was first announced, with the Steam Machines, I was very hopeful for more of a "HTPC OS" to finally end the pointless console wars.

I'm glad the Steam Deck has kept the SteamOS alive, yet I think for many these mobile gaming devices are more of a gimmick than anything else. Like how many would buy a $600-1000 gaming device that breaks after a few years? Probably a lot of teenagers, but who in their 30s and 40s can stand gaming on these more than a few minutes at the time? :p

Spoiler: *Snip*

There is no question that Linux is way more responsive general use, and Windows 11 isn't responsive no matter how powerful your hardware may be, even on a Raptor Lake the whole OS is jerky, especially keyboard input in all applications like Visual Studio, Notepad etc. Having used every major version since 3.0, Windows is clearly worse than ever. At home you can certainly mitigate some of these issues by disabling spyware etc., but you probably can't do that stuff on your work computer.

But as for gaming, we should keep our eyes on the target - to get gaming companies to use Vulkan and compile games for Linux natively. Wine/Proton is certainly impressive work, but from what I've seen it's still a hit and miss when it comes to individual games. While I haven't tried any top games launched from the last four years in Wine, but from what I've seen having a "top rating" in Steam with Proton doesn't guarantee it will work at all.


From the user perspective, pretty much any major (regular) distro and derived distros should be fine.
From the developer's perspective shipping software to be deployed on all kinds of distros requires some effort that a Windows dev wouldn't be used to; like do you ship just a tar-ball, a shell-script that's distro independent, or do you want like deb, rpm packages, or perhaps snap(with its own issues)?


I would argue the options for Windows aren't that great either since they moved on from win32, and MS has never been good with API versions, so having a complex GUI application for Windows means a lot of workarounds or rewrites for different Windows versions. While I don't know nearly all major client applications out there, I rarely see any heavy application use any of MS' GUIs(especially not .NET), it's usually one of two: 1) custom widget API built on top of OpenGL etc., or a well established widget toolkit like Qt (or sometimes GTK). But when it comes to enterprise stuff, it's usually much worse.


For sure, the ingame GUI is obviously rendered in the graphics API of choice. But games still usually needs a way to show error messages and sometimes some type of launcher. Doing this with pure X11 isn't that hard, ~200 lines will give you a styled message box, the probably only issue there is that fonts doesn't look too nice.

Those that use libraries like SDL will get such features built-in, along with lots of others including audio, input handling and special controllers such as joysticks. This is comparable to the feature set DirectX offers for Windows, and is appealing to many indie developers. But SDL is more like an engine framework, as it controls some the main control flow of the engine for you and offloads it to a library, so you wouldn't get the precise timing and threading control you want for a high-performance game engine. I consider it more of a learning tool than anything else. Taking an already built game engine and porting it would be cumbersome. Using something more barebone like glfw is probably the smarter choice, or better yet build your own, which is what I've done for 20+ years with OpenGL. The others will still remain though; like IO from the OS and audio. The latter of which is an area where I think we need an upgrade/replacement of the very outdated OpenAL. Linux does have the advantage of providing a very bare metal access to the physical devices in addition to events from X11 or Wayland. One last thing outside my expertise is networking; I haven't touched network programming since the early 2000s, so I don't the state for game development there.

My point being, if your game is already in Vulkan or OpenGL, porting the graphics is very simple as you say, the other parts probably needs some effort. :)

Weee! I'm teenager again! :p
No $600-1000 or in my case € necessary, bought a Steam deck, that had a run in with gravity (broken OLED screen and pressed in right thumb stick). Got a replacement screen and hall effect sticks, everything is fine now for under 250€. :D
If you don't want to tinker yourself, there are refurbished units on steam or other marketplaces, if you prefer similar handhelds.
Emulation, Indie games, old games, light 3D or just 2D games, even newer games on lower settings run perfectly fine. Desktop mode adds a lot of versatility (hello modding! :love:).

It's nice to see Steam OS getting traction on other platforms as well, hope Valve and the Linux community keep up the got work, that has been done so far. :cool: