To be honest, I’m really sad and concerned about people opting to use WSL. It really looks like another EEE scheme. If you want to have freedom in the long run, please, please participate in the FOSS desktop ecosystem: help debug driver issues, help maintain HCLs to make it easy for new users to pick knowingly well-supported hardware, at the very least, report the bugs you see.
WSL opens up a way for MS to push the restricted boot for desktops without people noticing who would otherwise be the first to notice. Monopolistic proprietary software vendors are not friends of FOSS. Never were, and never will be. Don’t trade essential liberty for temporary convenience.
Unfortunately, “the linux desktop” is also wholly captured by monopolistic proprietary software vendors, who are also not friends of FOSS. The systemd debacle, the dbus debacle, all the renderer debacles, the app store debacles, the various gnome debacles – it’s monopolistic proprietary tasteless foisted broken politically-contrived nonsense all the way down. It’s hardly surprising that people use WSL which at least has the benefit of you being able to run productivity software that consistently works.
With big corporations one can never know, but it seems to me that in recent years Microsoft have been a good citizen of the open-source software world. I’m willing to give them the benefit of the doubt.
That being said, I’m also concerned about native Linux desktop losing traction, but I have think this has been happening steadily for a while now. I used to know a ton of people running Linux as their primary OS, and now almost everyone’s on macOS and WSL. Perhaps most people don’t care much about the underlying principles, as long as they get their work done.
Well, and from your very post it’s clear where it eventually leads. macOS isn’t the “polished UNIX experience” people thought it always would be. I believe we should make the system we want or else we risk a situation when there will be nowhere to migrate to.
That said, I do get my work done on a Linux desktop and nothing crashes for me on 8th gen Intel NUC hardware.
I’m one of them. I’m actually the last person to switch out of… about 20 regulars? in what was once a local LUG. I still run Linux and OpenBSD pretty much everywhere I can except for my main working machine, which runs macOS. I’m not happy about it, but it’s also not a matter of convenience.
It’s not that I’ve sacrificed the principles of open source but, realistically, I do not trust the vision that’s currently prevalent in the world of FOSS desktop. It’s a vision that I don’t understand – seeking to provide software for users who understand and care about the technical aspects of fundamental liberties, who can file bug reports and test fixes from a devel branch, or even submit patches, who are willing to follow HCLs before buying new hardware, but are also confused by too many customisation options, and intimidated by small icons and buttons. It produces software that I find less capable with each release, and which makes it harder and harder for me to make things that others find useful, to work on new things, to learn things about computers and things other than computers, and so on.
Showing up with patches – or heaven forbid, bug reports – that do not demonstrate a sufficient understanding of these principles is met with hostility by many communities, and I honestly have no interest in even trying anymore. I’ve no intention of dying on hills held by designers who asked four interns to perform these actions and rate how hard it was, thus unlocking the secret of how to build the UIs that will finally bring upon us the Year of Linux on the Desktop.
And I honestly have no interest in going the underground route, either – running dwm, i3, or resurrecting my FVWM config from 15 years ago and going back to using mc, pine, bc and emacs under xterm. I like file managers with nice-looking icons and proportional fonts, I like graphical email clients and, generally, all these programs that try (and sometimes fail) to figure out what would be good tomorrow, rather than to get to the essence of what was good thirty years ago.
But I also think that it’s not a good idea to develop professional software – terminal emulators, text editors and IDEs, CAD/CAE tools, system administration tools, whatever – by using UX and UI design principles for novice users and freemium apps. All you get is software designed for people who don’t want to run it in the first place, and deliberately made worse for those who do want to run it.
I wish there was a better way, but if the best way to write things that run under Linux and help others do cool things with it is to get a Macbook or a Surface and SSH into a headless machine, or run WSL, I can do that. (And honestly, I totally get the appeal, too. After a few months with iTerm 2 – which is also open source, and I mean GPL, the commie kind of open source – I don’t want to touch Gnome Terminal ever again in my entire life).
In that vein though, can you think of concrete/specific examples that illustrate the conflict of interests/disconnect? The easiest I have from the OSX brand of face-meets-palm would be the dialog:“Terminal would like to access your Contacts”. Unpacking it speaks lengths about what is going on.
My favourite one lately is from Windows, which insists on automatically rebooting in order to install updates in the middle of the night so that you’re greeted by a full-screen ad for Edge, which is now also your default browser a better and more secure Windows experience in the morning. That works out great on paper, but in practice, lots of us who use computers for work just wake up to the Bitlocker screen, and five minutes of please wait, installing updates.
From Linux land… I’ve honestly ragequit a long time ago, I kept poking it for three or four years (by which I mean I mostly ran a bunch of old console apps and a few Qt apps) and eventually gave in and bought a Mac. I mostly have a bunch of bad memories from 2012-2017 or so, after which my interactions with it were largely limited to taking point releases of XFCE, Gnome and KDE for test drives and noping the fsck out back to FVWM. So most of my complaints are probably either out of date, or oddly generic (everything’s huge and moves a lot and that’s not nice to my ageing eyes).
Plus… there’s this whole approach to “improving” things, you know?
When I got this stupid Mac, it was uncanny to see how many of the things I used back in 2006 or 2007, when I last used a mac, are still there and work the same way. Even Finder has the same bugs :). Meanwhile, there are things I liked in KDE 3.5 that literally got rewritten and broken three times since then, like custom folder icons support. Which is now practically useless anyway, since icons from modern themes look pretty much the same below 64x64px – I get looks great in screenshots but guys I have folders with a hundreds of datasheets, there’s no way I can ever find anything in there if I can only see like 12 files at a time.
(Edit: FWIW, I think we pretty much agree on all twelve :P)
My favourite one lately is from Windows, which insists on automatically rebooting in order to install updates in the middle of the night so that you’re greeted by a full-screen ad for Edge, which is now also your default browser a better and more secure Windows experience in the morning. That works out great on paper, but in practice, lots of us who use computers for work just wake up to the Bitlocker screen, and five minutes of please wait, installing updates.
Somewhat ironic how denial of service is rephrased as a double-plus good security measure, no? Coming from a SCADA angle, the very idea of an update of any sort borders on the exotic (and erotic), but the value of that contract seems to have been twisted into a means of evaluating change by forcing it on users and study the aftermath. The browsers are perhaps the most obviously, but regardless of source it is quite darn disgusting.
From Linux land… I’ve honestly ragequit a long time ago, I kept poking it for three or four years (by which I mean I mostly ran a bunch of old console apps and a few Qt apps) and eventually gave in and bought a Mac.
So amusingly enough I was a die hard FOSS desktop user from the mid 90ies until the arrival of OSX (though raised in Solaris lands). Most of my bills went to pay the cluster of PPC mini macs I used to do my dirty deeds (the biggest of endians). Come 10.6 it was clear that Apple’s trajectory was “fsck you devs, we’re through” and I took it personally. Left all of it to rot, returned to FOSS and was dismayed by what the powers that be had done to the place. The tools I was working on towards, “solving oscilloscope envy by reshaping the debugger” had to be reused to build a desktop that “didn’t change beneath my feet as I was walking around”.
I get looks great in screenshots but guys I have folders with a hundreds of datasheets, there’s no way I can ever find anything in there if I can only see like 12 files at a time.
I would like to run an experiment on you, but the infrastructure is lacking for the time being – here is one of those things where VR performs interesting tricks. Posit that a few hundreds of datasheets are projected onto a flat surface that is textured onto a sphere. You are inside of that sphere with a head mounted display. How long would it take your cognition to find “that one sheet” among the others, versus scrolling through a listview…
the value of that contract seems to have been twisted into a means of evaluating change by forcing it on users and study the aftermath. The browsers are perhaps the most obviously, but regardless of source it is quite darn disgusting.
IMHO this trench war of updates, where users are finding new ways to postpone them and companies (in this case, Microsoft) are finding new ways to make sure updates happen, is entirely self-inflicted at the companies’ end.
Way back when big updates were in the form of service packs, there was generally no question about whether you should update or not. If you were running bleeding-edge hardware you’d maybe postpone it for a week or two, to let the early adopters hit the bad bugs, but that was it. As for the smaller, automatic updates, people loathed them mainly because of the long shutdown times, but it was generally accepted that they at least caused no harm.
Nowadays, who knows. During lockdown I had to go halfway across the city to my parents’ house twice to get my overly-anxious mother (elementary school teacher who’s one or two years away from retirement, so pretty scared when it comes to tech) past the full-screen ads with no obvious close buttons, restore Firefox as a default browser and so on, while she commandeered my other parental unit’s 15 year-old, crawling laptop to hold the damn classes. No wonder everyone dodges updates for as long as they can.
I would like to run an experiment on you, but the infrastructure is lacking for the time being – here is one of those things where VR performs interesting tricks. Posit that a few hundreds of datasheets are projected onto a flat surface that is textured onto a sphere. You are inside of that sphere with a head mounted display. How long would it take your cognition to find “that one sheet” among the others, versus scrolling through a listview…
I’m sure you’ve thought about this for longer than I have, but I suspect there are two things that determine success in this case:
An organisation system that matches the presentation (e.g. alphabetical order for a one-column list view)
Being able to focus on a sufficiently large sample that you can browse the list without moving your eyes back and forth too much
is pretty obvious, I guess. I like to go through listviews because these things are sorted alphabetically, and while many of them have very stupid names like slau056.pdf (which is actually MSP430x4xx Family User’s Guide), I usually sort of know which one I’m looking for, because manufacturers tend to follow different, but pretty stable conventions. As long as they’re laid out in a way that makes it easy to “navigate the list” (in stricter terms, in a way that preserves ordering to some degree, and groups initial navigation options – i.e. sub-directories – separately so they’re easy to reach, wtf GTK…), it’s probably fine.
probably bears some explanation because it’s the reason why large icons suck so much. Imagine you have a directory with 800 items and you’re looking for one that’s in the middle. If you can only see 10-12 at a time, then a) it takes a lot of time to hit the exact window with the file you’re looking for, and the tiniest amount of scroll moves up and down by a whole page of items. So you get to go back and forth between dozens of 10-item pages, and often overshoot, then undershoot the one you’re looking for dozens of times, all while wiggling your eyes all over the window.
I dunno what to think about the inside of a sphere. My knee-jerk reaction is to say I’d get dizzy and that a curved, but field-of-view-sized surface might be a better fit. But gentleman skeptics once complained that trains would be draughty, too, and it turned out they were, but also that it was more than worth it. Just like flying machines became a thing once we finally figured out imitating birds is just not the right way to go about it, we’re probably going to make real progress in organising and browsing information only at the point where we stop imitating libraries, so I think this meets the essential prerequisites for success ;-).
The field of UX has massively regressed in the last 15-20 years, everywhere, not just in the FOSS world. It’s a cargo cult at this point. Even many (most?) of the organisations that allegedly practice “metrics-driven” design routinely get it so wrong it’s not even hilarious – they make fancy graphics from heaps of data, but they have no control groups, no population sampling, and interpretations are not checked against subject feedback (which is often impossible to get anyway, since the data comes from telemetry), so it all boils down to squinting at the data until it fits the dogma.
It’s a common thing for people to discuss what {company X} thinks about {idea Y}. But if you work for a corporation at least a while, you learn there’s no such sentiment. It’s closer to {high level exec A} thinks that {area B} is a great way to expand and {investing in C} is the way to do it. Soon person on position “A” may change, “B” may have a good replacement, and money pumped into “C” may turn out to not have a good return. Microsoft doesn’t like or dislike anything. Managers with enough power temporarily like some strategies more than others.
We had Balmer on one extreme, now we’ve got Satya who seems like the other extreme. In 5 years we may have either: Balmer++ deciding to sue Valve for Proton, or Satya++ deciding to opensource parts of windows kernel, or someone in the middle who approves something nice for FOSS and destroys something nice for FOSS, because they don’t care about that aspect at all, or anyone in between.
Corporations don’t deserve the benefit of the doubt. Some execs do. Just remember they’ll be out in a few years.
I think this phenomenon is laid out very well in the book The Dictator’s Handbook. It claims that countries, companies, and other entities don’t have opinions or preferences; people are the top and at every level do. They are looking after themselves.
Disclaimer: not the author of the book; it’s just one of my favorites, and I recommend it for everyone to read.
With big corporations one can never know, but it seems to me that in recent years Microsoft have been a good citizen of the open-source software world. I’m willing to give them the benefit of the doubt.
What have they done, or stopped doing, to earn this praise?
They’ve released a lot of core tech as open source: VS Code, .NET Core, some Windows programs such as Terminal and Calc, etc. They’ve supported a lot of other open source things less directly: NuGet, Python packages, etc. They’ve more or less stopped campaigning, advertising and litigating against open source stuff the way they did up through the mid/late 2000’s – see here for entry points to some good examples.
I’ll happily give Microsoft the benefit of the doubt, but we’ll see if they continue this strategy of being nice for another 5-10 years, or whether we enter the “extinguish” phase, more or less the same way Google has in the last 5 years. If Microsoft thinks they can make more money being nice than being evil, then that’s what they’ll do; that’s the only decision path that matters to them.
They won’t abuse their monopoly too much, because they don’t actually have one. It’s no secret that they open-sourced .Net to make it a viable option for web on Linux servers, and even though they dominate desktop, desktop itself has competition from 1) the browser (see: web apps) and 2) phones/tablets (there’s zero difference between a laptop and a tablet with a keyboard). They’re playing nice because they know they’re an underdog and can’t afford to act otherwise.
Of the three platforms (browser/desktop/touchscreens), the Browser is most controlled by Google, and phones/tablets are most controlled by Google. This conveniently lines up with the “Google is the new Microsoft” meme.
I also think Microsoft has done way better, and I really think they are likely to continue supporting and embracing open source. It seems like the company culture has shifted in a very fundamental way.
That being said, we should still operate under the assumption that they won’t. Trust, but verify. Same reason you should be wary of signing a CLA that assigns copyright, even if you really, really trust the company the CLA comes from.
I used to know a ton of people running Linux as their primary OS, and now almost everyone’s on macOS and WSL. Perhaps most people don’t care much about the underlying principles, as long as they get their work done.
My main principle was always to get certain things done. For a long time (say, from the year 2000 to 2018) Linux was absolutely number one for that, for me. Now it’s Apple, but by a small margin.
I’m fairly certain that it will never be Windows except for games. And that’s becoming a small margin as well.
From the article:
As you can see there’s nothing fancy about it. I wanted to build a decent workstation, not a gaming rig. Still, I opted to get a decent discrete GPU, as there were a few PC games that I was hoping to eventually play (StarCraft II, Tomb Raider, Diablo III, etc).
So this is absolutely and obviously a gaming rig :)
I think that asking people to pay a cost (in time, complexity and frustration) to try and match their desktop workflows from Mac OS or Windows in Linux is a fool’s errand, and in addition, the “real enemy” isn’t Microsoft (Windows) or Apple, but rather Google/Amazon/Microsoft (Azure/Github). I think the real challenges to software freedom are in confronting the big SaaS providers, and I would love to hear what the various freedom advocates think we should do about that.
I don’t want to fall into the opposite cliché “Linux has always worked on all of my machines for the last 10 years”, but this article is a little frustrating… “The wi-fi problem was much easier to address - I just switched to ethernet and never bothered to investigate it.”
I’m not sure who this article is addressed to. Is it for the power user who would investigate and try to solve many of these issues? Or is it for my parents to whom I would just recommend to buy a pre-installed ubuntu desktop, or pre-installed fedora thinkpad.
Benno intentionally misrepresents his opponents in a way I find disingenuous, especially in that video, and it boils my blood.
Tacitly: most people don’t like launchd, but they don’t interact with it either. Find me a person who writes or debugs launchd plists and you’ll find that:
they’re quite rare
they don’t like launchd
launchd and systemd aren’t even comparable, one is part of a monolithic proprietary operating system and the other is an ecosystem that is semi-modular which lends itself strongly to being an entire layer of your computing experience (dubbed: “the system layer” by Lennart Poettering).
There is an element of “do it yourself” with Linux, which leads to people (especially people with entrenched knowledge) being wary of things that cannot be debugged easily- but launchd exists in an ecosystem where “take it to the store” is a legitimate fallback, and even if it weren’t there is a built in backup/restore system that literally restores the machine to its entire previous state.
Disregarding the proprietary or “creep”/complex discussion about systemd/launchd:
Also, I find the systemd comment funny especially from somebody coming from Mac with launchd. (TL;DW: Benno Rice explains how people criticise systemd but love launchd when one is basically the open source clone of the other)
systemd is much more ambitious than launchd. Sure, in some respects, Benno is not wrong to call it a “ripoff” of launchd. But a shift from his use of “ripoff” to “the open source clone of” is misleading both because systemd does much more and because launchd is open source and has been for a while.
I decided to play it safe and I went with Ubuntu 20.04, which was the latest release at the time. Historically I preferred Arch Linux and Fedora, but I wanted to minimize the setup efforts.
Over the years, I have settled on a similar approach.
For personal use I have a Void Linux installed, and I love it. It is minimal and tailored to my needs.
For every company that I work for, I am installing a fresh system. It used to be Ubuntu, but in my opinion Ubuntu degraded, and now I am sticking to Linux Mint. I am replacing mate with i3 and copy all dotfiles. I want a minimal effort system that just works.
I use Ubuntu with Regolith (https://regolith-linux.org/) for both work and personal project use. My work laptop is a Dell XPS 13 that came with Ubuntu pre-installed and my personal box is a System76 Merkat which also came with Ubuntu pre-installed. In both cases it “just works” (so far). I don’t have time anymore to debug OS issues.
I was not aware of regolith, it looks great! I always wanted to use whatever comes by default configured, because I use mostly terminal. The only must-have is i3 and so far this required me to configure a lot after switching session from gnome/mate/xfce to i3.
I’ve had to do close to zero configuration for regolith other than changing the default gap between tiles to fit my personal taste. It has a nice shortcut dialog (super-shift-?) for seldom used keybindings. After using it for a couple of years it is hard having to occasionally use an overlapping window system when I used other machines.
Interestingly, while I also want to forget about OS, my path was exactly the opposite:
I started with Ubuntu, and there things were always broken for unknown reasons.
Switching to Arch was better: now things were broken, but because I had broken them. Finally, I switched to NixOS, and, after spending considerable time learning the thing, I’ve reached the state where everything works, I can’t break it, because the root is read only, and I can always rollback after update by choosing a different boot entry. For the past 4 years or so, I have been essentially running the same system which outlived a handful of machines. This is very different from all my previous experiences, where I had to re-install stuff to unbreak things couple of times a year.
If you work for a lot of different companies and need to rebuild new environments with the same configs, why not use NixOS or Guix? It seems these are better suited to rebuilding from working stable states and sharing declarative configuration setups.
That is a good question.
I never considered it or thought about it. A new setup happens once or twice a year and takes less than 2 hours. I am not sure if investing in any of those would pay back.
True. If traveling through their learning curves for your build systems and CI isn’t seen as side benefit, learning curve may not be worth the investment.
I object morally to the way Windows and OSX work and to their parent companies so my only choice is Linux and I’m happy to pay the inconvenience penalty.
I have to say that this 1,600 EUR desktop completely blew out of the water my way more expensive MacBook Pro.
I also bought a 3700X machine last year, which blew my Mac out of the water. Now I have a MacBook Air M1, which goes toe to toe with my 3700X, and beats it for most Rust builds. However, it’s passively cooled and takes a fraction of the power of my Ryzen.
The Ryzen felt like the future for about a year. Now not so much.
I admit that M1 certainly looks very interesting. I might revisit Macs down the road when I’m in the market for a new laptop, as this combination of power efficiency and performance looks fantastic.
Nice write up. I agree that Ubuntu is a reasonable choice for a desktop distro that is easy to setup and get going fast. It facilitates the installation of non-free drivers also, if you’re willing or have to use those. I just can’t help but think it’s the Windows-like version of Linux (bloat wise) and just can’t stand using it, especially for servers. Between Windows and Ubuntu I personally just stick to Windows 10.
I find Fedora a much better workstation distro. It doesn’t push the (IMHO, completely wrong) idea that a different desktop environment requires a derivative distro, and it’s easy to start minimal and add what you want.
If Ripcord’s (alternative Slack+Discord client; shareware) Slack experience is as good as it’s Discord experience then it might be worth of shot for Linux users that have to use Slack but want a nice desktop experience.
Been using Ripcord for a few months now. The only problem is that when people send me a Zoom link using the Slack integration, I can’t open them, so I quickly have to open Slack in the browser version. Thankfully I have to only do this once in a few weeks, so I just suck it up and do the workaround as the benefits of Ripcord still outweigh having to use Slack.
Slacks recent “Huddle” feature is unsupported, so time will tell if Ripcord ends up supporting that.
Multi-tabbed conversations, (and 700 MB less memory usage) in Slack is a game-changer. I hope more such software turns up.
Perhaps someone writes a native GCal client someday? GCal is frustratingly slow on Firefox. Maybe it my fault for continuing to use Firefox on one of the flagship GSuite products hahaha…
As a counter data point, I had Windows 10 with WSL on my old work laptop. Last week I got a new one and decided to install Fedora on it, and everything Just Worked. I only had to resize the windows partition and then it installed on its own, it set up dual boot and found the Windows boot loader without me needing to do anything. All the software I use was easily available either from the official repos or by adding some external repos. Gnome is actually really neat these days. The VPNs I need to use are integrated into NetworkManager instead of being their own shitty separate programs that don’t want to remember my password. I had a few Windows-only programs left that I needed to use, and all I had to do was download the Windows ISO and select the “express install” option on Boxes (the Gnome VM manager that comes preinstalled) which only asked me for username, password and product key and in a few minutes I had a working Windows VM. I paired my bluetooth headphones and they just worked (probably thanks to Pipewire).
The only problem I’ve had so far was with Teams and screen sharing, which I’m assuming is because Fedora ships with Wayland by default and neither Teams nor Chrome run on Wayland directly, but even that might’ve just been a configuration issue on my part.
which I’m assuming is because Fedora ships with Wayland by default
Keep in mind that X is still fully supported and you can install it / switch to it easily if you want things like Teams to work. It’s not listed as a forced change for Fedora 35 either, so it’s not going away any time soon.
I know, but I’ve seen that screen sharing does work at least on Chrome without using its experimental Wayland backend so I think it might have just been a matter of permissions.
I had a few Windows-only programs left that I needed to use, and all I had to do was download the Windows ISO and select the “express install” option on Boxes (the Gnome VM manager that comes preinstalled) which only asked me for username, password and product key and in a few minutes I had a working Windows VM.
I tried that recently, but Windows graphics were unbearably slow. I tried all the different accelerated guest drivers (SPICE guest tools, etc.), but no dice. Various bug trackers and wiki pages said that there is currently no good accelerated solution.
In the end I nuked the Boxes VM and installed VMware Player and it worked like a charm.
I paired my bluetooth headphones and they just worked (probably thanks to Pipewire).
Lucky you. I head constant problems with various Bluetooth headsets.
The only problem I’ve had so far was with Teams and screen sharing,
The same experience with Zoom. Unfortunately, I had another set of issues with X11, where I could not draw in windows (with the Zoom drawing controls) when 2x scaling was enabled (HiDPI screen).
I was pretty happy with my ThinkPad T14 AMD hardware-wise, but there were still so many papercuts in both Linux and Windows, that one day I decided it was enough and sold the ThinkPad and got a MacBook Air M1.
I tried that recently, but Windows graphics were unbearably slow.
Yeah it’s slow, but since I basically just need it to run Citrix and RDP into some Windows VM which is also slow by itself so I don’t mind too much. I’ll keep in mind that VMWare works better though, thanks.
I head constant problems with various Bluetooth headsets.
Were you using pulse? I’ve heard that Pulseaudio had issues with bluetooth headsets, and that Pipewire is supposedly much better at handling them. I’ve also tried Pipewire on my NixOS laptop and it worked right out of the box, maybe it would’ve fixed your issues too.
For me, I’ve never had difficulty with Linux working on hardware (I stuck with good HW; stuff that’s not ewaste OOTB just works IME); it’s just that the desktop experience ranges from being on the cusp of interesting, a half-baked clone, or for Unix retrofetishists. It’s usable (I’ve been doing it for a couple of years on my desktop), but feels depressingly mediocre; always a waste of potential.
OP mentions GPU issues and that he was hoping avoid dealing with a proprietary driver. I wonder if he tried the proprietary driver in the end, though and if the issues went away.
I always buy one-generation-older-than-current graphic cards for this reason. If something doesn’t work out of the box I’ll just return it for something else.
That was my original plan, but I got greedy at the end and went the for a newer GPU from the current generation. I just assumed that with all the praise amdgpu was getting, this wouldn’t matter as it did in the past.
I tried the proprietary drivers as well, but they didn’t help. From what I gathered they are pretty close to the the FOSS drivers (both are developed by AMD themselves), so it’s not very likely that you’d get different results https://wiki.archlinux.org/title/AMDGPU#AMDGPU_PRO
This was a good article, and it pointed out a lot of things that are still present in linux that have been there since the beginning. I have been using linux in various forms since 1996 and like anyone else have switched directions hundreds of times, flopping mainly between an all linux environment, to Microsoft to a homogenous environment where I’m at now with linux, Windows, and macOS.
I may be in the growing minority, but while I’ve always been a secondary user of macOS, I switched over to it full time earlier this year and I quite like it… a lot. I also am deeply immersed in Apple’s ecosystem, so that helped sway my decision.
One key point though the article pointed out however was that maybe macOS will not be suited for development moving forward. With Apple switching over to their own silicon it’s hard not to see the writing on the wall as to where they are heading, and for this situation I think linux will continue to thrive.
I do very little development on linux as most of my development is centered around Python and scripting, so macOS and VSCode works for my needs. I wonder what percentage of folks like myself fall into this category.
To be honest, I’m really sad and concerned about people opting to use WSL. It really looks like another EEE scheme. If you want to have freedom in the long run, please, please participate in the FOSS desktop ecosystem: help debug driver issues, help maintain HCLs to make it easy for new users to pick knowingly well-supported hardware, at the very least, report the bugs you see.
WSL opens up a way for MS to push the restricted boot for desktops without people noticing who would otherwise be the first to notice. Monopolistic proprietary software vendors are not friends of FOSS. Never were, and never will be. Don’t trade essential liberty for temporary convenience.
Unfortunately, “the linux desktop” is also wholly captured by monopolistic proprietary software vendors, who are also not friends of FOSS. The systemd debacle, the dbus debacle, all the renderer debacles, the app store debacles, the various gnome debacles – it’s monopolistic proprietary tasteless foisted broken politically-contrived nonsense all the way down. It’s hardly surprising that people use WSL which at least has the benefit of you being able to run productivity software that consistently works.
With big corporations one can never know, but it seems to me that in recent years Microsoft have been a good citizen of the open-source software world. I’m willing to give them the benefit of the doubt.
That being said, I’m also concerned about native Linux desktop losing traction, but I have think this has been happening steadily for a while now. I used to know a ton of people running Linux as their primary OS, and now almost everyone’s on macOS and WSL. Perhaps most people don’t care much about the underlying principles, as long as they get their work done.
Well, and from your very post it’s clear where it eventually leads. macOS isn’t the “polished UNIX experience” people thought it always would be. I believe we should make the system we want or else we risk a situation when there will be nowhere to migrate to.
That said, I do get my work done on a Linux desktop and nothing crashes for me on 8th gen Intel NUC hardware.
I’m one of them. I’m actually the last person to switch out of… about 20 regulars? in what was once a local LUG. I still run Linux and OpenBSD pretty much everywhere I can except for my main working machine, which runs macOS. I’m not happy about it, but it’s also not a matter of convenience.
It’s not that I’ve sacrificed the principles of open source but, realistically, I do not trust the vision that’s currently prevalent in the world of FOSS desktop. It’s a vision that I don’t understand – seeking to provide software for users who understand and care about the technical aspects of fundamental liberties, who can file bug reports and test fixes from a devel branch, or even submit patches, who are willing to follow HCLs before buying new hardware, but are also confused by too many customisation options, and intimidated by small icons and buttons. It produces software that I find less capable with each release, and which makes it harder and harder for me to make things that others find useful, to work on new things, to learn things about computers and things other than computers, and so on.
Showing up with patches – or heaven forbid, bug reports – that do not demonstrate a sufficient understanding of these principles is met with hostility by many communities, and I honestly have no interest in even trying anymore. I’ve no intention of dying on hills held by designers who asked four interns to perform these actions and rate how hard it was, thus unlocking the secret of how to build the UIs that will finally bring upon us the Year of Linux on the Desktop.
And I honestly have no interest in going the underground route, either – running dwm, i3, or resurrecting my FVWM config from 15 years ago and going back to using mc, pine, bc and emacs under xterm. I like file managers with nice-looking icons and proportional fonts, I like graphical email clients and, generally, all these programs that try (and sometimes fail) to figure out what would be good tomorrow, rather than to get to the essence of what was good thirty years ago.
But I also think that it’s not a good idea to develop professional software – terminal emulators, text editors and IDEs, CAD/CAE tools, system administration tools, whatever – by using UX and UI design principles for novice users and freemium apps. All you get is software designed for people who don’t want to run it in the first place, and deliberately made worse for those who do want to run it.
I wish there was a better way, but if the best way to write things that run under Linux and help others do cool things with it is to get a Macbook or a Surface and SSH into a headless machine, or run WSL, I can do that. (And honestly, I totally get the appeal, too. After a few months with iTerm 2 – which is also open source, and I mean GPL, the commie kind of open source – I don’t want to touch Gnome Terminal ever again in my entire life).
That we agree on principle 1. “professionals first” comes as no surprise ;-)
In that vein though, can you think of concrete/specific examples that illustrate the conflict of interests/disconnect? The easiest I have from the OSX brand of face-meets-palm would be the dialog:“Terminal would like to access your Contacts”. Unpacking it speaks lengths about what is going on.
My favourite one lately is from Windows, which insists on automatically rebooting in order to install updates in the middle of the night so that you’re greeted by
a full-screen ad for Edge, which is now also your default browsera better and more secure Windows experience in the morning. That works out great on paper, but in practice, lots of us who use computers for work just wake up to the Bitlocker screen, and five minutes of please wait, installing updates.From Linux land… I’ve honestly ragequit a long time ago, I kept poking it for three or four years (by which I mean I mostly ran a bunch of old console apps and a few Qt apps) and eventually gave in and bought a Mac. I mostly have a bunch of bad memories from 2012-2017 or so, after which my interactions with it were largely limited to taking point releases of XFCE, Gnome and KDE for test drives and noping the fsck out back to FVWM. So most of my complaints are probably either out of date, or oddly generic (everything’s huge and moves a lot and that’s not nice to my ageing eyes).
Plus… there’s this whole approach to “improving” things, you know?
When I got this stupid Mac, it was uncanny to see how many of the things I used back in 2006 or 2007, when I last used a mac, are still there and work the same way. Even Finder has the same bugs :). Meanwhile, there are things I liked in KDE 3.5 that literally got rewritten and broken three times since then, like custom folder icons support. Which is now practically useless anyway, since icons from modern themes look pretty much the same below 64x64px – I get looks great in screenshots but guys I have folders with a hundreds of datasheets, there’s no way I can ever find anything in there if I can only see like 12 files at a time.
(Edit: FWIW, I think we pretty much agree on all twelve :P)
Somewhat ironic how denial of service is rephrased as a double-plus good security measure, no? Coming from a SCADA angle, the very idea of an update of any sort borders on the exotic (and erotic), but the value of that contract seems to have been twisted into a means of evaluating change by forcing it on users and study the aftermath. The browsers are perhaps the most obviously, but regardless of source it is quite darn disgusting.
So amusingly enough I was a die hard FOSS desktop user from the mid 90ies until the arrival of OSX (though raised in Solaris lands). Most of my bills went to pay the cluster of PPC mini macs I used to do my dirty deeds (the biggest of endians). Come 10.6 it was clear that Apple’s trajectory was “fsck you devs, we’re through” and I took it personally. Left all of it to rot, returned to FOSS and was dismayed by what the powers that be had done to the place. The tools I was working on towards, “solving oscilloscope envy by reshaping the debugger” had to be reused to build a desktop that “didn’t change beneath my feet as I was walking around”.
I would like to run an experiment on you, but the infrastructure is lacking for the time being – here is one of those things where VR performs interesting tricks. Posit that a few hundreds of datasheets are projected onto a flat surface that is textured onto a sphere. You are inside of that sphere with a head mounted display. How long would it take your cognition to find “that one sheet” among the others, versus scrolling through a listview…
IMHO this trench war of updates, where users are finding new ways to postpone them and companies (in this case, Microsoft) are finding new ways to make sure updates happen, is entirely self-inflicted at the companies’ end.
Way back when big updates were in the form of service packs, there was generally no question about whether you should update or not. If you were running bleeding-edge hardware you’d maybe postpone it for a week or two, to let the early adopters hit the bad bugs, but that was it. As for the smaller, automatic updates, people loathed them mainly because of the long shutdown times, but it was generally accepted that they at least caused no harm.
Nowadays, who knows. During lockdown I had to go halfway across the city to my parents’ house twice to get my overly-anxious mother (elementary school teacher who’s one or two years away from retirement, so pretty scared when it comes to tech) past the full-screen ads with no obvious close buttons, restore Firefox as a default browser and so on, while she commandeered my other parental unit’s 15 year-old, crawling laptop to hold the damn classes. No wonder everyone dodges updates for as long as they can.
I’m sure you’ve thought about this for longer than I have, but I suspect there are two things that determine success in this case:
An organisation system that matches the presentation (e.g. alphabetical order for a one-column list view)
Being able to focus on a sufficiently large sample that you can browse the list without moving your eyes back and forth too much
is pretty obvious, I guess. I like to go through listviews because these things are sorted alphabetically, and while many of them have very stupid names like slau056.pdf (which is actually MSP430x4xx Family User’s Guide), I usually sort of know which one I’m looking for, because manufacturers tend to follow different, but pretty stable conventions. As long as they’re laid out in a way that makes it easy to “navigate the list” (in stricter terms, in a way that preserves ordering to some degree, and groups initial navigation options – i.e. sub-directories – separately so they’re easy to reach, wtf GTK…), it’s probably fine.
probably bears some explanation because it’s the reason why large icons suck so much. Imagine you have a directory with 800 items and you’re looking for one that’s in the middle. If you can only see 10-12 at a time, then a) it takes a lot of time to hit the exact window with the file you’re looking for, and the tiniest amount of scroll moves up and down by a whole page of items. So you get to go back and forth between dozens of 10-item pages, and often overshoot, then undershoot the one you’re looking for dozens of times, all while wiggling your eyes all over the window.
I dunno what to think about the inside of a sphere. My knee-jerk reaction is to say I’d get dizzy and that a curved, but field-of-view-sized surface might be a better fit. But gentleman skeptics once complained that trains would be draughty, too, and it turned out they were, but also that it was more than worth it. Just like flying machines became a thing once we finally figured out imitating birds is just not the right way to go about it, we’re probably going to make real progress in organising and browsing information only at the point where we stop imitating libraries, so I think this meets the essential prerequisites for success ;-).
For what it’s worth, the Reduce Motion accessibility setting helps with some of that.
I think the worst thing that happened to FOSS is UX designers (and that includes things like Flatpak).
It doesn’t matter whether the reason for “software won’t do X anymore” is some evil mega-corp or some arrogant UX designer, the result is the same.
(And before anyone slides in with a “let me mansplain UX to you”: I’m good, thanks.)
The field of UX has massively regressed in the last 15-20 years, everywhere, not just in the FOSS world. It’s a cargo cult at this point. Even many (most?) of the organisations that allegedly practice “metrics-driven” design routinely get it so wrong it’s not even hilarious – they make fancy graphics from heaps of data, but they have no control groups, no population sampling, and interpretations are not checked against subject feedback (which is often impossible to get anyway, since the data comes from telemetry), so it all boils down to squinting at the data until it fits the dogma.
It’s a common thing for people to discuss what {company X} thinks about {idea Y}. But if you work for a corporation at least a while, you learn there’s no such sentiment. It’s closer to {high level exec A} thinks that {area B} is a great way to expand and {investing in C} is the way to do it. Soon person on position “A” may change, “B” may have a good replacement, and money pumped into “C” may turn out to not have a good return. Microsoft doesn’t like or dislike anything. Managers with enough power temporarily like some strategies more than others.
We had Balmer on one extreme, now we’ve got Satya who seems like the other extreme. In 5 years we may have either: Balmer++ deciding to sue Valve for Proton, or Satya++ deciding to opensource parts of windows kernel, or someone in the middle who approves something nice for FOSS and destroys something nice for FOSS, because they don’t care about that aspect at all, or anyone in between.
Corporations don’t deserve the benefit of the doubt. Some execs do. Just remember they’ll be out in a few years.
This is very true.
I think this phenomenon is laid out very well in the book The Dictator’s Handbook. It claims that countries, companies, and other entities don’t have opinions or preferences; people are the top and at every level do. They are looking after themselves.
Disclaimer: not the author of the book; it’s just one of my favorites, and I recommend it for everyone to read.
What have they done, or stopped doing, to earn this praise?
They’ve released a lot of core tech as open source: VS Code, .NET Core, some Windows programs such as Terminal and Calc, etc. They’ve supported a lot of other open source things less directly: NuGet, Python packages, etc. They’ve more or less stopped campaigning, advertising and litigating against open source stuff the way they did up through the mid/late 2000’s – see here for entry points to some good examples.
I’ll happily give Microsoft the benefit of the doubt, but we’ll see if they continue this strategy of being nice for another 5-10 years, or whether we enter the “extinguish” phase, more or less the same way Google has in the last 5 years. If Microsoft thinks they can make more money being nice than being evil, then that’s what they’ll do; that’s the only decision path that matters to them.
They won’t abuse their monopoly too much, because they don’t actually have one. It’s no secret that they open-sourced .Net to make it a viable option for web on Linux servers, and even though they dominate desktop, desktop itself has competition from 1) the browser (see: web apps) and 2) phones/tablets (there’s zero difference between a laptop and a tablet with a keyboard). They’re playing nice because they know they’re an underdog and can’t afford to act otherwise.
Of the three platforms (browser/desktop/touchscreens), the Browser is most controlled by Google, and phones/tablets are most controlled by Google. This conveniently lines up with the “Google is the new Microsoft” meme.
I also think Microsoft has done way better, and I really think they are likely to continue supporting and embracing open source. It seems like the company culture has shifted in a very fundamental way.
That being said, we should still operate under the assumption that they won’t. Trust, but verify. Same reason you should be wary of signing a CLA that assigns copyright, even if you really, really trust the company the CLA comes from.
My main principle was always to get certain things done. For a long time (say, from the year 2000 to 2018) Linux was absolutely number one for that, for me. Now it’s Apple, but by a small margin.
I’m fairly certain that it will never be Windows except for games. And that’s becoming a small margin as well.
From the article:
So this is absolutely and obviously a gaming rig :)
I think that asking people to pay a cost (in time, complexity and frustration) to try and match their desktop workflows from Mac OS or Windows in Linux is a fool’s errand, and in addition, the “real enemy” isn’t Microsoft (Windows) or Apple, but rather Google/Amazon/Microsoft (Azure/Github). I think the real challenges to software freedom are in confronting the big SaaS providers, and I would love to hear what the various freedom advocates think we should do about that.
What do the big SaaS providers have to do with the state of linux on the desktop?
Not much! But they have a lot to do with the state of software freedom, or rather, the lack thereof.
In summary: “My hardware doesn’t work on Linux”, “systemd is complex”, and “therefore I switch to WLS on Win 10”
Also, I find the
systemdcomment funny especially from somebody coming from Mac withlaunchd. (TL;DW: Benno Rice explains how people criticise systemd but love launchd when one is basically the open source clone of the other)I don’t want to fall into the opposite cliché “Linux has always worked on all of my machines for the last 10 years”, but this article is a little frustrating… “The wi-fi problem was much easier to address - I just switched to ethernet and never bothered to investigate it.”
I’m not sure who this article is addressed to. Is it for the power user who would investigate and try to solve many of these issues? Or is it for my parents to whom I would just recommend to buy a pre-installed ubuntu desktop, or pre-installed fedora thinkpad.
Benno intentionally misrepresents his opponents in a way I find disingenuous, especially in that video, and it boils my blood.
Tacitly: most people don’t like launchd, but they don’t interact with it either. Find me a person who writes or debugs launchd plists and you’ll find that:
launchd and systemd aren’t even comparable, one is part of a monolithic proprietary operating system and the other is an ecosystem that is semi-modular which lends itself strongly to being an entire layer of your computing experience (dubbed: “the system layer” by Lennart Poettering).
There is an element of “do it yourself” with Linux, which leads to people (especially people with entrenched knowledge) being wary of things that cannot be debugged easily- but launchd exists in an ecosystem where “take it to the store” is a legitimate fallback, and even if it weren’t there is a built in backup/restore system that literally restores the machine to its entire previous state.
Disregarding the proprietary or “creep”/complex discussion about systemd/launchd:
launchd is an init.
systemd is an ecosystem.
systemdis much more ambitious thanlaunchd. Sure, in some respects, Benno is not wrong to call it a “ripoff” of launchd. But a shift from his use of “ripoff” to “the open source clone of” is misleading both because systemd does much more and because launchd is open source and has been for a while.I find the systemd comment funny from someone saying they eschewed using Gentoo. Tinkering: use it or lose it.
Over the years, I have settled on a similar approach. For personal use I have a Void Linux installed, and I love it. It is minimal and tailored to my needs.
For every company that I work for, I am installing a fresh system. It used to be Ubuntu, but in my opinion Ubuntu degraded, and now I am sticking to Linux Mint. I am replacing mate with i3 and copy all dotfiles. I want a minimal effort system that just works.
I use Ubuntu with Regolith (https://regolith-linux.org/) for both work and personal project use. My work laptop is a Dell XPS 13 that came with Ubuntu pre-installed and my personal box is a System76 Merkat which also came with Ubuntu pre-installed. In both cases it “just works” (so far). I don’t have time anymore to debug OS issues.
I was not aware of regolith, it looks great! I always wanted to use whatever comes by default configured, because I use mostly terminal. The only must-have is i3 and so far this required me to configure a lot after switching session from gnome/mate/xfce to i3.
I’ve had to do close to zero configuration for regolith other than changing the default gap between tiles to fit my personal taste. It has a nice shortcut dialog (super-shift-?) for seldom used keybindings. After using it for a couple of years it is hard having to occasionally use an overlapping window system when I used other machines.
Interestingly, while I also want to forget about OS, my path was exactly the opposite:
I started with Ubuntu, and there things were always broken for unknown reasons. Switching to Arch was better: now things were broken, but because I had broken them. Finally, I switched to NixOS, and, after spending considerable time learning the thing, I’ve reached the state where everything works, I can’t break it, because the root is read only, and I can always rollback after update by choosing a different boot entry. For the past 4 years or so, I have been essentially running the same system which outlived a handful of machines. This is very different from all my previous experiences, where I had to re-install stuff to unbreak things couple of times a year.
For work I still use Void - just works.
If you work for a lot of different companies and need to rebuild new environments with the same configs, why not use NixOS or Guix? It seems these are better suited to rebuilding from working stable states and sharing declarative configuration setups.
That is a good question. I never considered it or thought about it. A new setup happens once or twice a year and takes less than 2 hours. I am not sure if investing in any of those would pay back.
True. If traveling through their learning curves for your build systems and CI isn’t seen as side benefit, learning curve may not be worth the investment.
I object morally to the way Windows and OSX work and to their parent companies so my only choice is Linux and I’m happy to pay the inconvenience penalty.
I also bought a 3700X machine last year, which blew my Mac out of the water. Now I have a MacBook Air M1, which goes toe to toe with my 3700X, and beats it for most Rust builds. However, it’s passively cooled and takes a fraction of the power of my Ryzen.
The Ryzen felt like the future for about a year. Now not so much.
I admit that M1 certainly looks very interesting. I might revisit Macs down the road when I’m in the market for a new laptop, as this combination of power efficiency and performance looks fantastic.
Nice write up. I agree that Ubuntu is a reasonable choice for a desktop distro that is easy to setup and get going fast. It facilitates the installation of non-free drivers also, if you’re willing or have to use those. I just can’t help but think it’s the Windows-like version of Linux (bloat wise) and just can’t stand using it, especially for servers. Between Windows and Ubuntu I personally just stick to Windows 10.
I find Fedora a much better workstation distro. It doesn’t push the (IMHO, completely wrong) idea that a different desktop environment requires a derivative distro, and it’s easy to start minimal and add what you want.
If Ripcord’s (alternative Slack+Discord client; shareware) Slack experience is as good as it’s Discord experience then it might be worth of shot for Linux users that have to use Slack but want a nice desktop experience.
Been using Ripcord for a few months now. The only problem is that when people send me a Zoom link using the Slack integration, I can’t open them, so I quickly have to open Slack in the browser version. Thankfully I have to only do this once in a few weeks, so I just suck it up and do the workaround as the benefits of Ripcord still outweigh having to use Slack.
Slacks recent “Huddle” feature is unsupported, so time will tell if Ripcord ends up supporting that.
Multi-tabbed conversations, (and 700 MB less memory usage) in Slack is a game-changer. I hope more such software turns up.
Perhaps someone writes a native GCal client someday? GCal is frustratingly slow on Firefox. Maybe it my fault for continuing to use Firefox on one of the flagship GSuite products hahaha…
As a counter data point, I had Windows 10 with WSL on my old work laptop. Last week I got a new one and decided to install Fedora on it, and everything Just Worked. I only had to resize the windows partition and then it installed on its own, it set up dual boot and found the Windows boot loader without me needing to do anything. All the software I use was easily available either from the official repos or by adding some external repos. Gnome is actually really neat these days. The VPNs I need to use are integrated into NetworkManager instead of being their own shitty separate programs that don’t want to remember my password. I had a few Windows-only programs left that I needed to use, and all I had to do was download the Windows ISO and select the “express install” option on Boxes (the Gnome VM manager that comes preinstalled) which only asked me for username, password and product key and in a few minutes I had a working Windows VM. I paired my bluetooth headphones and they just worked (probably thanks to Pipewire).
The only problem I’ve had so far was with Teams and screen sharing, which I’m assuming is because Fedora ships with Wayland by default and neither Teams nor Chrome run on Wayland directly, but even that might’ve just been a configuration issue on my part.
Keep in mind that X is still fully supported and you can install it / switch to it easily if you want things like Teams to work. It’s not listed as a forced change for Fedora 35 either, so it’s not going away any time soon.
I know, but I’ve seen that screen sharing does work at least on Chrome without using its experimental Wayland backend so I think it might have just been a matter of permissions.
I tried that recently, but Windows graphics were unbearably slow. I tried all the different accelerated guest drivers (SPICE guest tools, etc.), but no dice. Various bug trackers and wiki pages said that there is currently no good accelerated solution.
In the end I nuked the Boxes VM and installed VMware Player and it worked like a charm.
Lucky you. I head constant problems with various Bluetooth headsets.
The same experience with Zoom. Unfortunately, I had another set of issues with X11, where I could not draw in windows (with the Zoom drawing controls) when 2x scaling was enabled (HiDPI screen).
I was pretty happy with my ThinkPad T14 AMD hardware-wise, but there were still so many papercuts in both Linux and Windows, that one day I decided it was enough and sold the ThinkPad and got a MacBook Air M1.
Yeah it’s slow, but since I basically just need it to run Citrix and RDP into some Windows VM which is also slow by itself so I don’t mind too much. I’ll keep in mind that VMWare works better though, thanks.
Were you using pulse? I’ve heard that Pulseaudio had issues with bluetooth headsets, and that Pipewire is supposedly much better at handling them. I’ve also tried Pipewire on my NixOS laptop and it worked right out of the box, maybe it would’ve fixed your issues too.
Pipewire, on Fedora and NixOS.
Ah well, guess I was luckier than you then.
For me, I’ve never had difficulty with Linux working on hardware (I stuck with good HW; stuff that’s not ewaste OOTB just works IME); it’s just that the desktop experience ranges from being on the cusp of interesting, a half-baked clone, or for Unix retrofetishists. It’s usable (I’ve been doing it for a couple of years on my desktop), but feels depressingly mediocre; always a waste of potential.
OP mentions GPU issues and that he was hoping avoid dealing with a proprietary driver. I wonder if he tried the proprietary driver in the end, though and if the issues went away.
Likely his issues would be alleviated by running a newer (non-LTS) kernel with the appropriate drivers mainlined.
I have never had a good experience with the AMD proprietary drivers.
I always buy one-generation-older-than-current graphic cards for this reason. If something doesn’t work out of the box I’ll just return it for something else.
That was my original plan, but I got greedy at the end and went the for a newer GPU from the current generation. I just assumed that with all the praise
amdgpuwas getting, this wouldn’t matter as it did in the past.I’ve been trying newer kernels over the course of several months, but they didn’t help either. Probably they fixed the problem by now, though.
I tried the proprietary drivers as well, but they didn’t help. From what I gathered they are pretty close to the the FOSS drivers (both are developed by AMD themselves), so it’s not very likely that you’d get different results https://wiki.archlinux.org/title/AMDGPU#AMDGPU_PRO
This was a good article, and it pointed out a lot of things that are still present in linux that have been there since the beginning. I have been using linux in various forms since 1996 and like anyone else have switched directions hundreds of times, flopping mainly between an all linux environment, to Microsoft to a homogenous environment where I’m at now with linux, Windows, and macOS.
I may be in the growing minority, but while I’ve always been a secondary user of macOS, I switched over to it full time earlier this year and I quite like it… a lot. I also am deeply immersed in Apple’s ecosystem, so that helped sway my decision.
One key point though the article pointed out however was that maybe macOS will not be suited for development moving forward. With Apple switching over to their own silicon it’s hard not to see the writing on the wall as to where they are heading, and for this situation I think linux will continue to thrive.
I do very little development on linux as most of my development is centered around Python and scripting, so macOS and VSCode works for my needs. I wonder what percentage of folks like myself fall into this category.