Do we have to get cool people to use Linux? Is it that once enough people say "oh I don't have that problem, my computer is X" (where X is either "a mac" or "a linux" or something else) that it starts to take hold that if people want to not have this kind of thing happen that they should switch their operating system?
Is it even possible for this to ever happen? It seems like there's a race between operating systems that give you ownership and those that don't, and if operating systems that don't give you true ownership of your machine get better/non-intrusive enough to be essentially invisible, there will never be a need/pain-point high enough for people to want operating systems that value this ownership/self-determination.
In my nightmare scenario, everyone uses chromebooks, writes code on remote IDEs, programs with rough edges developed by amateurs never see the light of day, and it all works beautifully for the WALL-E-esque end user.
At some point one laptop was purchased with Windows so one guy could game. He played Fifa for a week then asked me to put Ubuntu since "this is what he was used to".
So, you see, it works that way too!
I've also been introducing Chromebooks with some success. One of them seems to fail spectacularly: I assume the battery is dead and needs a hard reset every once in a while which means going through some weird wizard that's confusing for basic users.
Sadly, Linux hardware support is still hit and miss. My latest Lenovo Legion has a whole lot of issues, the biggest one being that the trackpad just doesn't work. Sound also dies sometimes when waking up from sleep. But other than that it's a beast of a machine with a killer Ryzen CPU!
Unfortunately Ubuntu is really pushing snaps which adopt the Windows practice of forced updates (https://forum.snapcraft.io/t/disabling-automatic-refresh-for...) and in fact go one step further by forcing updates even on metered connections.
 https://www.deepin.org/en/ ?
That, in my opinion, is the deal breaker. My pc is a tool I use for getting things done, not something I want to tinker with every time I have to do something. I'm a literate computer user, programmer and sys admin, and I know how to change settings in Linux but I do not want to spend my time digging through mailinglists for a hack, every time I try to use my webcam/soundcard.
E.g., both my Dell XPS and ThinkPad X1 work flawlessly with Linux.
There was recently a news that all Lenovo laptops will be Linux certified so I guess there's little change to have issues with a Lenovo Legion. Turns out the trackpad just won't work. I never encountered this -- usually it was the WiFi that might cause issues (and of course, waking up from sleep).
Some laptops have shitty windows drivers too. Whatever you end up buying, you should read the reviews for beforehand anyways.
I still hope a good kernel hacker will buy it and fix the trackpad. I would even chip in some money in a fund, but it's not clear to me how that is done.
This actually cuts both ways. I ended up dual-booting for AutoCAD on my laptop. It's apparently at an age where AutoCAD still works well, but trackpad drivers don't work correctly in Windows. Because it's only used for AutoCAD though, I'll only ever be using a mouse, rendering the problem (mostly) moot.
This isn't an issue with Ubuntu, as everything from the trackpad, to the funky keyboard keys, just works.
That's the case on laptops, yes. But on desktops it's really quite good, it's rare to encounter any issue at all. Unfortunately most people want laptops so that's a significant problem.
As it happens my Ryzen desktop did have fewer issue but staring with the monitor off makes me have no GUI and I think I also ran into some sleep/suspend issues after an Ubuntu update.
I'm really not keeping track. If you keep your expectations low Ubuntu is quite lovely.
"I have to press the what to get to the what? When do i press it? Format drive?"
That's how most people have obtained their OS, they never ever touch boot settings. It must either come preinstalled by the manufacturer, or be installable from a GUI launched from their existing OS, which is what they have done if they have installed a newer version of windows than they had originally.
Multibooting is definitely the way to go, it could be seen 20 years ago when drive space began to outpace size needs of Windows' C: volume under typical user conditions.
Every Windows PC eventually was going to have enough drive space for more than one OS, in the long run (now) room for many more.
And the hardware was going to boot so fast, there was going to be no reason not to just quickly and simply reboot to Linux for the internet, or Windows for office work or gaming.
But you have to be able to teach multibooting, and for that you have to practice it yourself reliably.
Well, I'm here to say there are advanced geeks who are still just as bad at multibooting as below-average Windows users are at Linux.
So Linux is still not flying off the shelf even though it's free, and can be added to Windows PC's without Windows users losing anything.
For years now.
The hardware was made for this since the beginning (using multiple floppies, then later multiple HDDs) but it could be considered a manual operation.
Later DOS versions supported the IBM concept of partitioning a single HDD into a maximum of 4 primary partitions, analogous to a 4-drawer filing cabinet being replaced by a digital alternative. A folder(s) containing an OS can safely be stored in any drawer no differently than any other folder. Windows iconized DOS directories as folders as time went on.
The BIOS (later a biosmenu) chooses which HDD to boot to in the original way, then further boot (menu) sectors/files on the HDD choose a partition having a folder containing an OS, then any OS treats each partition equivalently to a separate _drive_ of its own.
This part of the process wasn't even that slow when PC's were 1/10 the speed.
And Windows responded by salvos of bloat with each revision, sometimes in separate inter-revision events where much larger amounts of downloaded material was retained by huge leaps over what the originally issued OS started out doing, as drive space made excellent leaps in affordable space for users, the excellence was nullified. In my estimation a most Ballmerish occurrence.
Also the original NT6 Vista & Windows 7 versions started out with the BOOT folder right there in the C: volume where the default bootmenu it contains could conceivably be edited easier (easier than it was later when hidden in a separate unexpected partition).
Except for the NT6 boot files which became more arcane by orders of magnitude, this is by design.
You need sufficient BCDEDIT skillz at the administrative command line to modify your default NT6 bootmenu, but you only needed a text editor for the human-readable BOOT.INI of NT3-5 or CONFIG.SYS in W9x.
And with NT5 (or NT6 using BIOS/MBR) you could still boot (multiple) Linux(s) from the regular Windows bootmenu, but no Linux for you with NT6 bootmenu in UEFI, this is by design.
No overcoming Windows SecureBoot obstacles in UEFI for Linux either until earlier this year in 2020, this only set Linux back 6 years when it comes to brand-new PC's.
And all you have to do is learn how to multiboot well enough so you only have to teach people how to use a boot menu(s) in the way that everyone should have already been using now for decades.
Now in a FAT32 ESP with UEFI/GPT, Linux installers put their boot files alongside Microsoft's in the generic shared EFI folder, while the Linux OS itself can be completely directed to a single unused partition, /home /var /usr everything on a single type 83 EXTx volume. Without touching Windows on its established NTFS partition. It works so well, it's the closest thing to Linux on a Windows partition.
Windows can make the partition and Linux can do the rest.
Then you have two completely different formatted partitions each containing their native OS, and the original common FAT32 ESP containing the key EFI folder now having both OS's boot files in their own subfolders.
And the proper Linux install just leaves your mainboard with a new UEFI top-priority firmware bootentry pointing to the new Linux bootfiles in their EFI subfolder, pushing the established firmware bootentry for Windows down the list in the mainboard UEFI.
This is after the Linux setup process autofinds and autoadds the Windows install already present on its NTFS volume to the new grub menu so you can choose original Windows or new Linux while the UEFI is set to select the new Linux subfolder within /EFI.
You can always set the mainboard UEFI back to boot to the unchanged established Microsoft UEFI bootentry instead of the new Linux UEFI bootentry, and the PC will act like Linux is not there at all, the new grub bootmenu will not appear and Windows will not normally be able to see or access the Linux partition. You could also use a built-in mainboard firmware bootmenu to manually choose between Windows or Linux during bootup, not much differently than if they were on separate HDDs, without needing any grub access to Windows, but not every UEFI mainboard even has a built-in bootmenu any more like BIOS did (BIOS could only choose from different HDDs, not the individual OS's on the HDDs like UEFI sometimes can do in addition), and they are all different but can be useful when there.
 If a real programmer is interested, what is still missing is a new standard open-source UEFI bootmenu firmware executable.
User-configurable to allow flexibly selecting which mainboard UEFI entries appear on its own menu when called, and their boot priority.
And as simple as possible, to be unchanged afterward, we already have overly complex grub & Windows OS bootloaders downstream on the HDD, as moving targets.
We just need a new hypervisory bootmenu upstream, to run under UEFI before booting an OS or its particular bootfiles, so there will be a standard way to select from firmware bootentries themselves. Without having to properly access the (otherwise confusing but potentially simple) UEFI boot choices manually while booting, using increasingly divergent mainboard Setup routines as more years go by.
In the form of a small .efi executable to be placed on the ESP volume, which can be called from the startup.nsh batch textfile (also store startup.nsh on the ESP volume).
While still being able to eventually obtain Microsoft SecureBoot signatures to any extent needed for such a useful bootmenu.efi & startup.nsh to come into simple routine use.
Which if present, a startup.nsh file is already the thing addressed by UEFI before the mainboard ratchets down to booting the top-priority bootentry in its own UEFI list instead, and startup.nsh is a userfile so we would best start using it.
But nobody's using startup.nsh yet, it's still not significantly underway for over 6 years now.
It's just another obstacle where the resulting UEFI mainboard boot-performance landscape is ridiculously unstandard (and getting worse) by comparison to the previous BIOS approach which was never a formally published standard at all, and BIOS got better over the years.
That's damage from a big UEFI foisting salvo by design, fully forseen.
And Microsoft SecureBoot still looms large as a related obstacle from 2014 since they have not yet signed an open-source UEFI Shell for SecureBoot operation either. Shouldn't be essential for developing a .efi executable bootmenu program anyway, but still a bad lingering deficiency in simple UEFI utilities, and confirms Ballmer's ghost still haunts regardless of WSL or what anyone says otherwise.
I can do that decently with virtualization.
Having to reboot is a really, really big drawback that's justified only if it's otherwise impossible or impractical to achieve what you need - but it is possible and practical, so multiboot is not really necessary.
You could theoretically also shrink a disk partition for a Linux install from a Windows system, but Windows doesn't recognise Linux filesystems (like ext4) natively. As a result, you can't go any further with installing Linux, and I don't know if you can mess around the bootloader to point it to GRUB.
That's talking about dual boot of course, because I'm pretty sure you can't completely replace a Windows install from a running Windows system.
Technically not installing it from windows. But the install action would be initiated from windows.
Would probably end up super janky in the end tho.
You'll just need to have a bootloader that's capable of reading NTFS partitions, that can launch the installation after reboot.
The USB installer still offers to resize existing partitions and install Ubuntu alongside Windows.
What do you mean? You can set up a bootable Linux install from Windows, or just use WSL instead of booting Linux?
How about not wanting to have this kind of thing happen and NOT having to switch their operating system?
Because this sounds like the old political argument "if you don't like this country, move".
Well, changing it should always be an option too.
If the direction where the country is going is not where you want to go and you have no power to change that then moving out is the smartest choice.
Then use free software. Why would you expect to be able to change whatever you want on software you don't own?
Microsoft isn't a democracy. They're not beholden to the majority of people who use their stuff, just to the majority of the money that's paid to them, and their shareholders.
Even companies with shareholders still need customers, or at worse, users/eyeballs.
LibreOffice/OpenOffice is not sufficiently compatible (receiving an e-mailed complex Word or Excel document, changing a single letter, and sending it back is not guaranteed to not change anything else), so that's it, the 'network effect' locks me in.
I could run Windows in a VM on Linux, or Linux on Windows; but that's a bit inconvenient, so OS X gives a good compromise for that.
I do like linux for work, and for use as a basic computer for doing basic things like browsing. It has come a very long way over the years, IMHO surpassed Windows 7 and left Mac OS far behind. But Windows 10 is right up there with linux for ease of use, and most importantly has decades of compatibility.
Both have their quirks, so that's a tradeoff, and neither set of quirks really outweighs the other for me. But that compatibility with the last quarter-century of my life is a clear point in favor of Windows.
For Windows though, I do recommend getting the Pro version and configuring it how you want it. That eliminates almost all the problems of W10 that people complain about, like the random reboots and the tiles on the start menu.
Thankfully WPS Office works with it fine, and allowed me to remove Microsoft from my workflow completely.
(Windows 10 and it's updates, telemetry, no-notification update reboots, and unexplained slowness for me to finally migrate fully to desktop Linux).
The problem with windows for personal use is that it gets in the way too often with pop ups, notifications and crap ads inside start menu.
Modern gnome is quite good already, though our team is using older hardware that's well supported on Linux.
New team members get familiar with the gnome desktop quite quickly and with a dedicated app store installing apps has never been more easier on Linux.
We do still use windows but only inside a vm for office and outlook mainly because we have to
With many people using browser based tools there is simply not much of a difference in people's workflows across desktop oses.
I think the change has to start at the enterprise level, once companies start using linux then it's simply a matter of time before people switch their personal computers to linux.
It's no less a problem for enterprise use. On my Windows 10 work machine I constantly get pop-ups begging for credentials and attention for stuff that I simply don't care about or that doesn't need the credentials. Some of these are particularly aggressive and will always steal keyboard focus from any window, while daring to put up another pop up notification near the systray (which are so extremely important according to microsoft, that these notifications have even higher priority than the pop-up menu of the systray -- what'cha thinking Microsoft, do I wanna interact with the thing I just clicked on to open, or with the thing you pooped up over it under my cursor?!) when I dismiss them. I made a spreadsheet once and got about two dozen work-interruptions from Microsoft popups.
Not to mention the absurd aggressiveness of Microsoft Teams notifications. I avoid the app when I'm not actively participating in a call (it randomly uses too much CPU and is also obviously tracking everything you do on the PC), so I usually have Teams quarantined in a Firefox tab. Other software manages to have hit-me-once notifications, but not Teams. Somewhere someone writes a message? Let's put multiple notifications on the screen and start bleeping away. What, you didn't look at the message? Let's send a reminder email!
Modern Microsoft software would make Kafka cringe, for we are morons allowing this.
I installed it on my work phone which I don't use to make it stop.
At another point during an upgrade from one LTS edition to another, the DWM was switched out but something happened during the transition and the logon screen booted up to the old, unusable DWM/interface. Had to boot at a different runlevel and point the boot config at the correct DWM.
These are the kind of issues which keep Linux on the desktop limited to enthusiasts only. I have worked professionally with Linux for many years, but at home I just want shit that runs itself. I can recover from these kinds of problems but that's my leisure time shot to shit for the evening and that's a very different equation to fixing issues on the clock (or in leisure time where I'm specifically electing to tackle something tech-related e.g. in the homelab).
Then I discovered my ageing Windows 7 licenses on the machines would still upgrade to Win10 just fine and that was the end of Linux on the desktop for me for a while.
$ ls -1 /boot/
It's a hard problem to tackle since distros don't have the huge telemetry data that Windows has to diagnose and fix these issues.
This is primarily why I suggest to people to buy hardware with a special focus on Linux support.
Companies are now offering linux laptops, Dell with developer edition XPS and lenovo think pads. The Linux experience on these will be vastly superior.
On the other hand, my Ubuntu mountain lion desktop so far:
1. Died because somehow kernel upgrade failed and I had to manually roll back
2. Had a weird interaction with vscode causing a system freeze where I had to figure out what to manually modify in config.
3. Occasionally dies when browsing some js heavy pages on chrome
I definitely prefer Linux as a dev environment and want my backend apps deployed on Linux servers, but the desktop experience is just not smooth. I can’t imagine my elderly dad dealing with the Linux issues I’ve had so far.
I turned those off within a day after I installed it and haven't had those problems in years. I do recommend getting the Pro version as it makes that stuff more easily configurable. Some people will take issue with that, but for me as a power user it's worth it to have that additional control over something I use all day every day.
ETA: I use linux and MacOS at work, and linux is quite good for that. I find MacOS infuriating though - worse than Windows even for basic window and file management functionality, and worse for configurability.
Also I can't remember the last time Ubuntu crashed on me.
Sure, I've messed up configuration and myself broken pieces of it, but I could do that on Windows too.
But memes won't die, arch is unstable, Linux is hard and for nerds only.
(I should add that I'm not even a careful Arch user - one probably should check news before upgrading; I've never bothered.)
Windows makes me restart just to update .Net for Christ's sake.
Somehow you've got decision about Linux from Ubuntu. There are not as much changes on rolling release though I had once similar problem on Arch Linux in a decade.
I've heard it is impossible on Nix (easy rollback).
There are 443 Canonical employees  and 13400 Red Hat employees . I've heard "Red Hat" a lot in presentations of core stack - X, Cairo, Pango. Linux kernel contribution by company — Red Hat 8%, no Canonical . Ubuntu made upstart, mir, unity, now snap. Red Hat made systemd and flatpak.
Ubuntu experiments with users more than Arch Linux does. It is among the first on pulseaudio, systemd, wayland. Some reverted, some lived through. It is not stable. So why is Ubuntu user friendly?
It used to work out of the box and look great at that.
Ubuntu actually made Gnome 2 look and work so well that for a few years that was my favorite.
Red Hat always was a hassle to install for end users and their default Gnome was dated and at times poorly configured.
It is 2020 — live USB, graphical installer, video review on https://manjaro.org/download/
I have not heard of this ever happening with macOS or Linux.
That said, Proton has gone a long way towards making Windows completely redundant for me.
Most people are not that sort of person.
They are shown a particular app like Excel, and they are told the big X in the corner closes the app. They don't wonder what the difference between the big X and the little X (I think this style of document-in-app might be old now, but anyway). You give them another app with an X in the corner, and they come and ask you what the X means. You show them Outlook, and they think Outlook is the only app that can send email. You show them a web based email, and they wonder how they can do all the things Outlook can do, but in a browser. You show them how to google for their problems and they think coming to YOU is the best way to do the search. You drop them into a new OS, and now you are the interface.
I definitely wouldn't peddle Macos or ChromeoS as a reasonable alternative where ownership of your computer is the root solution. They all come with their own versions of middle fingers to users, just that the users are not as vocal about it.
Also, I think the fundamental business model of Macs running OSX and Chromebooks running ChromeOS are very different, and I'd trust apple before I trusted Google.
Every time this comes up I am here to offer a perspective perhaps few have.
I started using Linux before it was 1.0. It was so long ago, quite a few ppl here weren't even alive (and then again, I am sure there a few here who started in '91 or '92). For a while, only at university department level servers. Then dual booting it at home, with Windows 2000 (maybe even Win98 at the tail end of that, I don't remember) and having Linux as my daily driver first on desktop then on laptop 2004-end of 2017.
After this long, I must say, anyone using Linux on a laptop has Stockholm syndrome. You can choose between Ubuntu where infrequent OS updates shatter your system so throughly you can't work for days or you can use Arch Linux which rolls and breaks smaller parts of your system from time to time, mostly scanner, printer, Bluetooth. That's how it goes. https://xkcd.com/619/ this is 1000% true. Oh yes, I use Linux on servers to this day, it's great.
Sure, if your usage is plain home wifi and browsing the web, basically something just a tiny bit beyond a Chromebook, it might work but heaven forbid you'd try to connect to some funny enterprise VPN or even worse, enterprise wifi or as I mentioned, have a Bluetooth speaker or an MFD.
And there's no help. None. Nobody knows any more what's going on. We had a pretty good handle on how the various parts of the system fit together because it was damn simple but post-systemd noone understands it any more or if they do, they don't share. I posted this more than once, my favorite example here is the DisplayLink adapters (which are awful but they are often a necessary awful) and people said, read the Arch wiki page, well, back then, even like three years after I did, the meat of that page was what I could scrape together from mailing list and blog posts of the relevant driver authors but that's it. I certainly didn't understand it and people blindly copypasting out of the wiki didn't either and if you hit a problem, bam, who is there to help? Or to the previous point and this no help together: the IT lead at my workplace asked some years ago how to make the F5 VPN w/ two factor work on Linux and the deafening silence is the answer. I managed to make it work by running an old enough Firefox as root (!) which could run the F5 NPAPI extension. It's just plain nuts. Yes, F5 should have provided a native Linux client which supported 2FA on Linux but they didn't. It doesn't matter whose fault it is but at the end of the day, I am out of work if I can't log into the VPN.
I am running Windows 10 since 2018 January. Not a problem. I ran O&o shutup, I use Classic Shell, so I never even see the Win 10 Start menu, I use WSL and -- everything just works. I have no problems with the system installing security updates at 5am, I set my system once so it only installs those, big updates are postponed by six month or forever and the whole thing just. works. Yes, Windows Subsystem For Linux, because the Linux command line tool are just great. It's everything else that is less great.
https://www.reddit.com/r/linuxquestions/comments/iej5y6/netf... look at this and despair. Once again: I don't care, I have translated Doctorow's speech to Hungarian to spread it, to help fight DRM, no matter how little, even that was 16 years ago, I fought yeah, but the reality is, it's here and I don't want to bother with this. Look here, Linux people say, it-works-or-at-least-ought-to-with-this-workaround this is what I am saying: people will bend over backwards to defend their OS... I am not sure whether I am able to get my point across. sigh.
I've tried Windows 10 and WSL for several months. It feels like consumption OS for me, I am not that productive. I do not like ecosystem, I do not feel secure installing 3rd party binary software.
I want distribution with plenty of open source solutions that just works. I have not had issues with Arch Linux in years. That said I do not use nvidia, radeon, printer, bluetooth, netflix and VPN. I had unsupported hardware experience and check twice beforehand.
Maybe Stockholm syndrome is not about OS but how it suits user? I certainly had it on Windows, no more.
(a) People who do very little with their computers. A tiling window manager, an Emacs, and a few terminals will never break, sure.
(b) People who selectively block all the breakage out of their memory ("yeah, I had to update some low level settings, that's not breakage, that's just normal as the driver interface changed in this version, etc").
But not people who use their computers for all kinds of things besides hardcore coding, especially multimedia (hobbyist or professional)...
Don't be so optimistic ;). As of recently, Emacs started freezing regularly for 10+ seconds when editing Rust code while rust-analyzer is used as an LSP plugin. First, I thought it was rust-analyzer, but I did some tracing and all time was spent in FreeType functions during the hangs (probably something going rogue, triggering a lot of font rendering). Lo and behold, use terminal Emacs (no X11) and the problem is gone.
But since life was busy, I didn't have the time to debug it further and installed CLion, which works great.
(c) People who use their computer for all kinds of hardcode coding except multimedia.
And actually (a) and (c) could be merged.
I had one issue with Arch Linux - partial init, boot to console - check issues . There was an issue with Flash player, solved first by HTML <video> than by bundled Chromium Flash. Pulseaudio - yes, solved by removal, ALSA still works. Can't run windows software - solved for me by web.
I have a ton of programming languages installed. When I experiment with xmonad and it breaks it is not Linux, it is me. If I am going to experiment with CUDA etc I know it would be uncharted territories.
In another episodes I saw Linux completely and irrecoverably (at my level of skills) break xwindows after dist-upgrade (on a simple business line HP laptop with a single Intel GPU), stuck in dependency hell trying to run 3rd party software because libc6 had incorrect version and apt couldn't resolve that (again, way above my skill level).
I'll use Linux on servers and clients where it is great, but my work host system will remain Windows for the foreseeable future, thank you very much.
PS: and to address OP article - both my home Win10 and work Win10 never force reboot for updates for some reason, at least in the space of up to 2-3 weeks of continuous uptime. Yes, the sometimes nag me for updates (just like Linux) and I simply install them in some downtime periods (just like Linux).
2. dist-upgrade breaks a lot, 3rd party software does not play nice with distributions
My rule book — stick to software that plays nice with distributions, use rolling release. I can rebuild my entire system from source with ease. Hence I can patch my system and I do it. It is a great power and it is unimaginable on Windows.
Now I have some business lenovo with intel gpu and no issues running ubuntu 20 and daily updates.
Anything from bluetooth speakers (Marshall's nice) to external monitors to whatever just works...
Also running Arch on a home server with an amd gpu for games (steam/paradox games).... funny enough that also works great.
When I did have issues it was due to nvidia (crappy drivers).
For hobbyist stuff... depends on hobby i guess... for me most projects involve a pi zero or similar... good luck doing that with windows...
Regardless, having experience with Linux and macOS -- macOS often has annoying breakage too. I can imagine that it is the same on Windows.
All software is broken. But at least with open source software you have the possibility to fix it if you have the technical chops.
Software wouldn't be valued as much if more people were willing to do the work lol
On the other hand, it is also easier to pick up those skills on an open source OS than a very closed system such as iOS.
I can't take seriously people who affirm that they are both "Linux user since 91" and at the same time pretend that Debian / Ubuntu don't work on any of their machines, ever. Come on.
Seriously, I don't get it. What doesn't work for you that would work better on Windows? I have absolutely no experience of windows (the last version I used somewhat regularly was 98SE).
Issues with things like dual monitors on Linux laptops and other peripherals.
If you're opening that new tab in Chrome or Firefox, you're watching Netflix in 720p, even in Windows.
I just watch it in 720p in Linux and don't worry about it. Life's too short.
Once again, using absolutes to try and make a point seem stronger. No, it's not "anyone". Sure, yeah, sometimes setting it up is hella annoying when your WiFi doesn't work out of the box (one laptop 8 years ago and one this year), printers won't connect (OpenSuse is terrible with network printers unlike Ubuntu), NVIDIA decides to stay dicks, and some media keys on your keyboard don't work, but those are all driver issues. Once those are out of the way, it also "just works" TM.
The amount of time I spent having to hunt down driver files for Windows XP and Windows 7, defrag the HD, clean up the windows registry, find a crack or freeware alternative for every little thing out there, or try to find out why something crashed and have absolutely no logs available, was inordinately high. I don't miss it one bit.
Sure, the community isn't the greatest and support is not always easy to find (or existent), but I dare say when things go wrong on linux at least you have logs and paste something into a search engine. When something crashes on Windows, most of the time you're shit outta luck - at least that how it was back in the day. It would be really surprising to me to see GUI apps on windows being runnable from the terminal and output something you can search for to debug. Hell, on Linux you can often even download the debug symbols, run the whole things in `gdb` or `strace`.
Your experience is yours and it's easy to find similar experiences and opinions, but it'd be great if you strayed away from absolutes like "anybody" or "everybody". Your experience isn't universal.
>> "Do you have support for smooth fullscreen Flash video, yet?"
Seriously, who uses that?
I'm your guy. But no, I will not use Linux.
I will tell you why.
Unfortunately, you won't like it because I'm an outsider, and it'll feel like a personal attack no matter how I phrase it.
( The only time Linux people listen to criticism like this, is the few times it comes from insiders. ESR's rant titled "The Luxury of Ignorance: An Open-Source Horror Story" nearly perfectly highlights the attitude and the issues that keep me away from Linux, so you can just read that if your prefer: http://www.catb.org/esr/writings/cups-horror.html )
Every time I have used Linux, it feels like using something written by a visitor from a parallel Earth, someone who brought with them an operating system from a place that never had a Microsoft or Windows.
Maybe Linus Torvalds really is from another plane, a different dimension, stuck here in our world? It certainly feels like it to me. It's a thousand little things where if there's some option or a choice, it's virtually guaranteed that Linux will do the opposite of what Windows does.
Things like: Does a click get registered on mouse down or mouse up? Is the task bar on the bottom or some other side? Where are the minimise and close buttons? Etc...
Now, I get that Linux is copying POSIX and UNIX and many other "standards" that are older than Microsoft the company, let alone Windows. This explains why even though for the last two decades 90% of the entire planet has gotten used to the back-slash (C:\foo\bar) as the path separator, but Linux uses the forward-slash (/mnt/foo/).
That's probably the one excuse I've seen that's legitimately too difficult to change, where there is a real reason that nothing can be done.
It feels purposefully contrary, doesn't it? Coming from Windows, like the vast majority of the planet, it's the first slap in the face telling you in no uncertain terms that you and the rest of the majority can go get fucked, that it is YOU that must adapt to Linux, not Linux to you.
This is the exact same thing as the United States, with a mere 4.23% of the global population forcing inches and the m/d/y date format down the throats of the other 95.77%.
This isn't just some superficial thing, like which character is the path separator. It's a thousand, thousand things that should have been fixed decades ago.
But no, Linux people are obstinately conservative, clinging to 50+ year old "standards" that 90-95% of the rest of the abandoned long ago or never heard of. Most of those standards boil down to: "Whatever Berkeley happened to do in the 1970s in their computer labs is the One True Way, and can never be changed."
This is why it cracks me up that whenever I talk to open-source developers, the type that carefully avoid all-things-Microsoft like the plague: they actually use Apple Macs to write code.
Because Steve Jobs, like me, had no tolerance for conservative, backwards, contrary bullshit. He wanted a usable operating system that didn't waste time. He didn't want something suitable only to a university lab setting, he wanted an operating system for consumers. He wanted an operating system that just worked and didn't require the user to read a textbook and monkey around with text-based config files just to print.
Every few years I download a Linux ISO from some random distro and give it a good go, just to see if it has grown up yet.
One of my favourite acid tests is this one: Can it join an Active Directory domain?
This is something 99% of corporate environments have.
Linux people will say: "Yes! It can! There's Kerberos support!"
That's like saying that a pile of unshielded copper cable can be used power a computer. Technically true. Not what I want. I want a plug that I can plug in.
In Linux, "joining a domain" means some random combination of:
Providing dozens of parameters manually that Windows does not ask for and does not provide.
Hard-coding IP addresses instead of using DNS names.
Running command-line tools on the Windows Domain Controller to export "kerbtab" files.
Monkeying about with PAM config files to actually make the authentication show up.
Having to do hideous things to the system to actually support ACLs and more than one group membership. You know... more than one. MORE THAN ONE! Crazy, crazy stuff like that.
Linux fanboys will rant that this is normal, that this is expected, that Windows is somehow "hiding things from the user". Just seconds ago I searched for "kerbtab" and this is one of the first things that came up: https://news.ycombinator.com/item?id=22338327
The end result is this: Every time one of two redundant domain controllers on a network has a hiccup, or is simply down for maintenance, Windows and Mac clients will smoothly truck along like nothing happened.
Linux clients will lose their shit. They will take minutes to authenticate, or simply time out and lock out the users.
So who cares, you ask? Why is Active Directory support important?
Because when I looked into it more deeply, I found that the core of the issue was that Linux DNS client is hot garbage. Windows and Macs can fail over to the secondary DNS server is milliseconds and transparently maintain the ability to look up the needed SRV records for locating the LDAP endpoints. ( https://petri.com/active_directory_srv_records )
Linux... can't. That's why it uses IP addresses for the Kerberos configuration parameters.
Because some lab in Berkely or MIT or wherever "did it that way" back in 1985 or some shit.
I'm not kidding. I looked it up. I read through the decades of rants where network engineers where begging the Linux maintainers to see reason and implement DNS server failover properly. Where they begged them to support TTLs and caching properly.
The arguments were absolute bonkers. Crazy town stuff. People just foaming at the mouth as they went on and on about how it's somehow critical for preserving the legacy of POSIX to not fail over quickly.
Apple Macs fail over instantly. Google's Android fails over instantly. Windows fails over instantly.
Linux... has to be contrary. It has to use the primary DNS and sit there patiently waiting for the timeout for every. Single. Request.
This breaks more than just Active Directory. Web browsing in this state is virtually impossible. It's just that it still kinda-sorta works, and people just shrug it off. But with Active Directory, it's a show-stopper. Total disaster. You can't even log on!
So this is the point where I sigh, delete the virtual machine, and go back to Windows, which "just works" and lets me get my job done.
Http URLs use forward slash. MacOS uses forward slash. Virtually everything else uses forward slash. Windows is clearly the oddity here, but since that's what you're used to you incorrectly assume that backslash should be standard.
Inches were a standard too, used by the majority.
A standard being older has no relevance when trying to 'sell' something to the majority group that has switched to a newer standard. A group literally born into that standard.
Now, I get it, the back slash vs forward slash thing is an irrelevant distraction, and not even worth discussing seriously, which is why I called it out as such.
But there's thousands upon thousands of things in Linux that are worth changing, that are worth the while being updated to match what the 95% are used to.
This is basically the gist of the arguments for and against systemd.
From the perspective of someone used to Windows or MacOS, systemd is catching up to NT4.
From the perspective of a Linux user, it's forcing them to use kph instead of mph over their dead body.
The fact that the Unix-type world is not like the Windows world is precisely what makes it relevant and worth using.
I completely dislike the Windows way of doing things, it makes me uncomfortable. That's why I use something different.
This is the most triggering thing I've ever read on this website. First, it assumes that average every day users navigate the windows file system by that URL bar or power shell which is definitely false.
But wait, where would every day users actually have more hands on experience with a path structure? Oh that's right, their web browser where it's all forward slashes. Windows is the outlier here, not Linux/Unix systems. They're the ones who should change.
Linux is not a private company to please you. It is you who can fix issues. If something does not work it is a signal — no one bothers enough to implement. You don't. Quite a lot of people like you don't. Entire niche can't code or hire to code.
Linux users is not homogeneous group. I do not care about you and your use case. Maybe you've found some trolls, good for you. Maybe it works for some users and they've shared how it can be done but you can't listen. Who knows.
Linux works perfectly for web development — a ton of languages, popular browsers, open source tools. It works out of the box on hardware it supports. It allows GUI customization and there are so many choices. I do not want window decorations and task bar — fine .
It is fascinating how many projects with deep history merged together — vim, emacs, man, tex, X, bash, perl, etc, etc. It is still unfolding right here and now. Communities formed like islands. It takes years to comprehend how it differs from Windows, to find your own path.
You can pay for better Linux AD support (e.g. Centrify) that IME just works. Just like you paid for the same in Windows.
It's one of the first things for me, too. In fact, it's one of the only results that Google returns for that search term, almost certainly because usage of it is practically non-existent (except for your "mis-usages").
A search for "keytab" -- the proper term -- produces the expected results.
I'm not even going to bother responding to your other arguments -- neither the flawed ones nor the one or two that are actually almost borderline legitimate.
When you continue to insist on blaming Linux for the inability to flawlessly integrate and co-exist with a closed-source, proprietary system designed by a company who -- at the time it was actively designing and creating that system -- considered Linux a "cancer" which need to be eradicated and wiped out, well, I get the feeling there's really no point.
Since Microsoft now "loves Linux" so much perhaps they'll be motivated to someday publish the technical details of all of their proprietary "extensions" to the standards that everyone else uses.
Until then, I'll just leave it at "Linux is obviously not for you" and move on.
You don't think that's a problem?
@home I used pdnsd, MaraDNS and dnsmasq in the past which did that trick for me, where dnsmasq was the best performing solution, though not the most convenient, that would have been MaraDNS, and pdnsd today seems ancient and bitrotten.
Anyways, Linux is like Lego, and you prefer already assembled kits. Which is understandable. Anything else is delusional, like starry eyed preachers on some street corner in downtown, promising whatever and the one true faith.
Active Directory is not important, and it has never been. But if it's important to you then use FreeIPA.
If you are entering IP addresses manually then you don't understand DNS.
Then instead of doing everything by hand, create a script, package, or image that solves the problem.
Mobile OSes are just as bad if not worst
To give an example, I got a graphics card with ray tracing capabilities, and I wanted to try the Minecraft RTX beta to see what this ray tracing thing was all about. In order to install it, I needed to create an account in this special XBox program, and then I had to delete and re-install the Minecraft application several times in order for the windows store to install the correct version. I think I rebooted the entire computer at least 3 times in the process. And on reboot, after signing up for this "insider" program, I got a new prompt about sending my usage data back to MS.
Once I did get it installed, I found that the Minecraft home-screen is now a garish display designed to put micro-transactions in front of you, to the point I had to hunt a bit for how to start a new game. After all of this the RTX portion turned out to be a half-finished mess, and I had to read a tutorial online to figure out how to even access RTX content.
So basically the user experience I had, in trying to play a game for children, felt akin to being forced to walk down a dark alley and shaken down by MS in order to access the content I wanted.
Honestly it's part of the reason I'm not a big fan of VSCode. The "suggested" popup feature just seems like it follows this trend of MS finding ways to put content in front of the user without asking, and I'm just not confident it's not going to be abused in the future to up-sell me on Azure services or something.
I don't leave updates hanging about for weeks and months, which explains how forced reboots are not a problem. Forced reboots are your own stupid fault, I have no pity at all.
What doesn't make sense is that I get none of this crapware on my system. It's clearly possible to disable this garbage, I don't know how I have done it without restorting to any scripts.
Edit: to clarify, as someone who very recently installed Win10 and Office. You do not have to reboot to install Office. Windows had to reboot because some critical update had to be installed, and office so happened to be installed at the same time. Don't get me wrong, I uninstall a few things after a fresh Windows install (edit 2: and that is unacceptable behavior from Microsoft). This article, however, is way off with the primary complaint: the reboot. Play stupid games, win stupid prizes. Allow Windows to update, because it really does know best.
Edit 3: this is the Verge, who put bloody thermal paste onto a CPU when the AIO had it pre-installed. There isn't a high bar of journalism here.
As far as I'm concerned, the problem is that the system is unable to update itself without rebooting.
You are absolutely correct! There was a brief period during XP, I think, where this was the case. Irrespective of that, reboots are a fact of life if you have Windows installed, and allowing and choosing the best time for Windows to reboot is a part of having Windows installed.
Don't use Windows if you can't deal with the reboots. Use Mac or Linux.
They force it on you too. I'm not sure if they ever go ahead and automatically restart / update. But the OS constantly nags you to update.
I yearn back to the good ol days, when you would issue a command to a computer and it would execute that command and then wait for the next command. The user called the shots. Now computers do all sorts of things that their manufacturer wants, or the OS vendor wants, or the application developer wants. My desire as a user is secondary or unimportant. Automatic software installation, sudden reboots, notifications, Try YouTube Premium!, analytics collection, ads on my browser, ads in the start menu, ads inside applications, Try YouTube Premium!, Setup Cortana!, Setup Siri!, Setup Windows Defender!, Setup your Microsoft Account, Turn on virus protection, Take a Tour, Here's What's New, Keyboard Batteries Low, Disk Not Ejected Properly, Try YouTube Premium!
Just stop it!
I wish computers would stop just doing things on their own, and that I could go back to my computer only doing things that I ask it to do.
Perhaps that's why you aren't asked?
Younger peers think that Microsoft being evil was a fad or a meme, but I remember bribing lawsuits where Microsoft would earn literal billions of dollars shoving their software to the state offices, schools etc., and vendor locking it to the point where anything FOSS would just bork out.
Too bad we don't care for the modern corporate history as much as we do for the non-modern one.
If it's a userspace-based model with kernel modules (like nvidia) - yes, you can insert/remove kernel modules without rebooting and exchange userspace libs freely.
If it's an in-kernel driver part like the i915 module that comes bundled with your distro's kernel, you can kexec the new kernel, which is simple with systemctl kexec.
It will likely require to exit and restart the graphical session though.
I don’t either, but also the forced reboots only happen once a month for me, right after patch Tuesdays. I’ve found that setting updates to manual and proactively updating on that day gets rid of almost all forced reboots. Of course now you have an extra thing to remember, but it does generally fix the problem.
Unless you have matching SHA-256 hashes, the default heuristic should be to assume that you might have a "customized"/testing variation of the software. We really need to get in the habit of including hashes when talking about software behavior.
The annoying thing is that the advance warning is gone, probably some default setting says that it will reboot sometime, but I very rarely get notified that updates are pending and that today/tomorrow I'll get a forced reboot. It just does it when I lock/close/sleep the laptop/desktop.
It stops windows store auto-updates, and I’m not sure about this, but I think it might stop you installing apps from the windows store without granting the permissions back temporarily, but either way I’ll take that trade-off
I don't see why people pay for that stuff.
- switch to linux (ubuntu might be the exception)
- Do what they want. Install the app. Turn on automatic updates. Do not worry when they change something and adapt to the new normal.
Think of the jobs that require people to use a cloud based software for 90% of their work, these jobs just need a good browser like Firefox and a office suit like LibreOffice.
Take 2D/3D design work. 10 Years ago it was almost impossile to get any decent 2D/3D work done on foss now there are powerful tools like GIMP, Inkscape, Scribus, and Blender, Godot.
I agree the non FOSS alternatives are a lot more smoother than the FOSS alternatives, but this is a compromise worth making when we consider how user hostile some software companies have become in the recent years and if more people turn to FOSS their pain points can be converted in to tools that helps us all improve FOSS.
We are at a time when a few software companies have so much power they are even able to play with our ability to participate in a democracy and I think its our responsibility as software folk (most of us here) to help guide people in the FOSS direction than trying to nudge everyone away from it because of a few pain points.
Most of these have been around a lot longer than 10 years. Heck, The Gimp was released in 1996, Blender exists since 1994 (I remember playing with it at the end of the 90ies or beginning 00s when it was still freeware), Scribus since 2003, and Inkscape since 2003.
Sure, each of them have improved the last 10 years, but the deltas are not that big that you can get many things done now that you couldn't get done 10 years ago.
I primarily use Linux, but with the exception of Blender, these tools are not as easy to use or as extensive as their counterparts on macOS (and probably Windows, but I don't really know Windows). In fact, until very recently Gimp and Inkscape did not even have proper HiDPI support (since they were still on Gtk+2).
This is not a criticism on the developers of these tools. They are amazing feats of engineering. But at some point they would probably need funding so that someone can put in the time to polish all the sharp edges.
So even given decent funding, i'm not sure GIMP could replace photoshop. Network effects are a serious moat.
Okay, but what do you win, really? Are you going to use Google Apps or Microsoft 365? The exact same thing is going to happen (or is happening) there already.
I think that replacing somewhat flawed, non-OSS local systems with OSS 'thin clients' that connect to fully closed, centrally managed cloud services is a really bad choice if you want to regain control.
- sneaky practices like what the article expains
- one-off price of e.g. hardware
- recurring price of e.g. services
- stability of services
- control over your devices
- control over your data
- seamless integrations
Apple: high integration, high stability, high privacy, high price of hw and services, low freedom, low control over devices, high control over data, low amount of sneaky practices
MS: low integration, medium privacy, high amount of sneaky practices, low price of hw, medium price of services, high freedom (odd thing to say), high stability, medium/high control over devices, medium control over your data
Google: low privacy, medium sneaky practices, low price of hw, low price of services, high/medium freedom, low stability, high/medium control over devices, low control over data, medium integration
You can't get all the good things at once, unfortunately.
If you don't care that random X11 applications (e.g. your browser) can snoop keystrokes and mouse events or make screengrabs of any other X11 application.
But it doesn't matter anyway, because any application can grab any data from your home directory.
Unless you use Flatpak, bubblewrap, or some other sandboxing technology.
Disclaimer: I use Linux 95% of the time. But we should be honest about the shortcomings. Desktop Linux is not secure and you only have privacy if you trust all of your applications.
No operating system is secure when it comes to installing untrusted applications. The strength of desktop Linux is the vast amounts of available free and open-source software, ergo software that you can actually trust. To my knowledge, no other OS beats that - usually you can't even trust your OS itself these days.
Plus we already have a universally supported sandboxing technology, it's our web browser. I can use Microsoft's software inside a Firefox container when I'm forced to, without giving them access to my home folder, keystrokes/mouse events, etc. Even better, half of the requests those things would make are blocked by uBO.
There's also Wayland as a replacement for X11, but granted not all applications support it.
Your comment makes it sound like this is a problem that only X11 has.
Sandboxed macOS apps (e.g. anything from the Mac App Store) cannot log keystrokes or make screen grabs without the user giving explicit permission to do so.
How can one trust closed source software?
On Mac: There was an application that could do that, but Apple sued them and so was not available anymore.
On Windows: Was not able to get those mp3's out (don't remember why exactly)
On Linux: Was able to mount it and get those files out with some Open Source iTunes clone.
At that point, I totally ditched Apple.
It is not for everyone, but for us who love one of them it is fantastic :-)
So it seems sometimes you can get all the good things at once.
- even lower price (upcycled hardware)
- even higher privacy (how can one trust closed source?)
- even higher freedom (replace any part of the stack)
- no sneaking
- high stability
- high control oven device
in right hands
For the last week, every morning I wake up, all my editors, terminals, servers, etc... are closed because it rebooted.
There is an entire stackoverflow thread on hacks to avoid it because there is no official way to prevent it.
I thought my Macbook Pro was bad usb-c issues were bad, this thing is worse!
That should get you past most of the forced update reboots, because it’s usually “patch Tuesday” updates that force reboots.
Not saying this is a good thing or excuse for the behavior. It’s just how you can get past the particular issue.
I imagine there are good reasons in force to update, like security patches. But I'd be lying if I didn't find it frustrating that my work laptop needs to update at what always feels like an inopportune time.
In general, when I am prompted about a software update I default to selecting "not now" (I've been refusing Google Photos' nagging for what feels like forever now).
What does it say about our industry that users have developed an automatic aversion to updates? Shouldn't updates be something exciting? Interesting? Somewhere along the line we burned our relationship with users and lost their trust.
Nobody wants to spend hours trying to get their computer back to the way it was yesterday, redo all the work they were doing, or learn new workflows when they have a deadline tomorrow or an important presentation to do this afternoon. When updates have become malware, people will block them.
When it does restart, it's often accompanied by flashing up some Mozilla property in the UI like pocket or lockwise that I really don't want.
Since Windows 10 came out, the OS has become a search engine for stuff, ranking MS products highest of course.
People and enterprises insist to use this OS calamity, despite the fact there are free and open source alternatives. I guess we will still see news about MS windows non sense in 2030
The fact is that Microsoft Office is still lightyears ahead of Open/Libre/Whatever-Office. And enterprises need that office suite.
And the integration between Office, ActiveDirectory and Office365 (their cloud service) is just awesome, spot on. It's impressively good as a groupware solution.
Outlook on Windows is fairly okay, and covers the needs of 99% of the office workers (Outlook on Mac OS is superb, by the way).
The thing is: until the FOSS alternative becomes a no-brainer, enterprises will keep paying for Office, and thus for Windows.
I would like so much to replace Windows with Linux on my laptop. But every time I try (which I do every 5 years or so, despite knowing better) it ends in tears and frustration. I don't even have exotic hardware. Just some mainstream Thinkpads and Dell laptops. Usually, I give up after literally (!) weeks of trying to get sound / hardware accelerated graphics / hibernation / power management / etc. working properly. Ironically, I get the best Linux experience with WSL or when I run it in a VM on top of Windows.
I can only imagine how upset people would be when they have more serious jobs interrupted.
What is going on in this world that some people think this is ok?
I will never update the device again and disconnect it from the internet. So the people who thought it was a good idea gained absolutely nothing.
It is also the reason I keep using Linux.
What I’m trying to say is, if you use Windows you accept that you’re using a walled garden. If you don’t want to, use Linux, or use both. But what’s the point of driving activism towards change? There’s already options if you care, and it’s not like any of the two models are completely without flaws.
1. Freedom to play any game.
2. Work uses a Windows specific VPN software.
Absolutely no other reason. It's the main source of frustration in my computer-related lifestyle.
I firmly believe that unless you have incredible insight, or incredibly expensive certifications, Windows is absolutely worthless to any advanced power user who wants control over their computer.
edit: I have some fairly strict settings on my laptop, so maybe that stopped it. I also have Office365 so that could be related too.
1. Some guy says something about Linux.
2. He gets several replies that are multiple variations of "well, Linux doesn't work for me!"
In this case, it's Windows users mentioning how the issues in TFA don't happen to them.
Maybe this will help some understand that replying with "for me, it's broken!" (Whether it's about Linux or Windows or anything else) is very likely not helpful at all, or completely pointless at worst, because it might be perfectly possible that it isn't broken for the people who made the original comments so there's not much discussion to be had.
I've had to roll back twice now and the only technique I've found which works to block updates on Windows 10 Home is to have the C: drive so full that MS can't download the update.
Even worse is when nation-states use Windows in critical and sensitive systems. There's absolutely no guarantee Microsoft isn't siphoning out information. People think China is bad, but as Snowden showed us, the U.S has been proven to be worse in terms of cyber espionage.
The only way to completely prevent them is to completely incapacitate Windows Update and forego all updates.
It's not a good idea to discuss techniques to disable Windows Update in open fourms as Microsoft has a habit of sabotaging anti-update techniques that become too well known.
My dad switched back to windows because he missed his favourite apps, which i could not find a good alternative for (like Windows Movie Maker).
Windows just works, it's bundled with every PC ever, and it's hard to want to leave that for something so much more difficult.
Nowadays, I know I could get a Linux box up and running without issue, and that friction is an order of magnitude lower than it was a decade ago, and I'm itching towards trying once more given that these thousand cuts in Windows are killing me, but it's gonna take time to gather up the drive to undertake the jump.
I use Linux for the last 7 years. Zero issues so far. I have dual bot for gaming and windows have more hardware issues than Linux. On Windows my Wi-Fi card would sometimes fail to start.
- Why I have to install drivers?
- Why terminal is so ugly?
- Where is my package manager?
- How many Settings are there?
- I do not like desktop manager, what alternatives do I have?
- I want to patch Explorer. Where can I find sources?
- I have some troubles with programming language, why?
Since Windows XP I find its outlook ugly. And it is getting worse.
I just don't find Linux (or Mac) a compelling alternative. I can do everything I possibly need on Windows; Visual Studio, gaming, VS code, docker, node, browsing, whatever. I absolutely think Linux is so much better for servers since it can all be done from the command line..NET Core was a game changer in that regard, but I'm still perfectly fine with Windows desktop.
I think it comes down to what you prefer, are you a tinkerer or just a user? I just want my desktop to run, I don't care what drivers I'm running or anything like that. Windows comes with the computer, I'm used to it and I've never had the problems described by people. I can plug in any peripheral and it just goes, it's frictionless for me. As far as restarting I'm fine with it, I don't run my computer 24 hours a day, I shut it down after work until I need it for gaming or other reasons, I'm not running software for the space shuttle so I don't need 10 nines uptime
Which is exactly why I use Linux as my main desktop OS, specifically openSUSE. All of the drivers are just there stuff just works.
When I started out with Linux in 2001, it was with a tinkerer mindset, but now I really just want something that gets out of my way. Windows just keeps getting in my way and Windows 10 is the worst yet.
The only reason I dual boot is for gaming, Proton is still not quite there yet.
for me, Steam for video games and Visual Studio