Disclosure: I'm 100% Linux since 2005 (except embed devices (game console, Roku)). All the Line-of-Business stuff "just works".
"Try to avoid relying on proprietary software" is strong. "Avoid any option that exists to run software you think you need" feels out of touch, especially when it says "I use Mac for X and Y" - which is barely practical: having a whole extra, expensive computer that's not maintained forever is quite the costly workaround for an arbitrary stance like "don't use Wine" that they don't motivate so much in the end (there's no practical explanation in that comment for avoiding VMs or Wine - they say maintenance, but I don't see what's hard to maintain in running Wine).
The comment argues "The good news. Every bit invested in high quality API/ABI on Linux pays off.". I do agree. I don't know about high quality, and it hurts a bit to say it, but it so happens that Windows might be the only stable API/ABI on Linux, with Wine being a completely libre reimplementation of it. If you need to write a program that you are reasonably sure will run on any Linux in 20 years without intervention, Wine might be your best bet (with AppImage probably your second best bet). What would be the fundamental (philosophical, practical, technical) reason to avoid targeting Wine? What makes winelib so different from other libraries such that you should avoid it? Genuinely curious. What real alternative is there? Qt and Gtk break the API each major version and even the GNU libc doesn't guarantee ABI stability. The only reasonable alternative is "maintained free software" (and that's what I happen to rely on).
FWIW, I have no stake in this: I use only free software, I mostly don't use Wine nor contribute to it, and I wish I were wrong.
Just my take though, I get your point that people spreading this idea and encouraging it have a place and at least its not negative. I just don't think that they really are market movers.
Valve releases Half-Live 3 exclusive as Linux port only. Announces that Windows port will follow at some time undefined time.
That pushes the market. A lot of angry people will become more angry and write posts on the internet. The rest grabs an ISO and decides, that the game matters more than Win11.
That’s not crazy. That is the daily business of Windows, XBox, PlayStation and Nintendo. They even decide deliberately making new games platform exclusive for the newest console, forcing gamers into expensive upgrades for new consoles (and weirdly these people are happy about spending 500 for it).
PS: The shares of Valve will not drop. Valve is - not a stock market company - it is privately owned. That’s the reason why we nice actions from Valve, over long terms. I cannot say what will happen with Microsoft shares.
The setup is usually a pain and needs workarounds. Same for weird SMB or Exchange stuff. It is a hell. The admin changes a setting and you’re in trouble.
It was fixed with the native port of Counter-Strike. Exchange was fixed by EU regulation. The other applications found better native replacements. I’m giving you the advice because the “hacker ideology” doesn’t help users. Users need reliability and defined behavior. The users itself can influence that by using compatible software and APIs. And by requesting it. And we need to pay for it.
Don’t lock back. Don’t stay in a hostile relationship. The same advice applies to government agencies or companies. Prepare. Clean cut.
PS: As mentioned I except Proton (as derivative of WINE). Valve controls Steam. Valve works permanent on fixing new issues. They could told a company to hold the update back to protect Linux.
Valves native Linux ports changes the situation. But when these get native ports Microsoft loses ground in the companies. Not in the stupid ones which deep dives into Azure. But many (young) Admins will cheer, and escape path out of hell.
(For games, there is Proton.)
To cut a very long story short - after Windows 10 restarted on her, and changed default browser and application settings too many times she was going to completely give up using the computer.
I built a new machine (a Dell AIO workstation) for her with Ubuntu, FTM and a few other things.
Worked brilliantly.
Had I listened to your recommendation, I would've never tried Linux.
Sorry, but Linux doesn't run Photoshop. Or Valorant. Or certain VPNs, certain educational software, and doesn't work with a bunch of hardware.
Dual booting is still a hell of a lot better than trying to configure Wine in most cases, but if doing everything natively on Linux was an option, it would've have taken SteamOS so many years to become even remotely usable. And even then people install Windows on their Steam Decks to run certain specific programs or games.
For the same reason native Linux isn't an option, native macOS wouldn't have been an option back when I first tried Linux. And even today, programs like Paint.NET are dearly missed on Linux and macOS (yes, I know about Pinta), and stock macOS is infuriating to use without all manner of tools and background programs reminding me of my XP. I use Windows for my Windows tools, Linux most of the time, and macOS for my macOS work stuff. I'm not getting rid of either non-Linux OS because that would make doing certain things simply impossible.
Last I checked, Office 365 didn't work, Basically anything modern Adobe didn't work, even the latest version of Visual Studio (not VSCode) didn't work. Things may have changed, I just learned to live without that stuff.
Wouldn't have believed it if I didn't first see and then use it myself.
Think it's because JUCE is relatively well-supported on Wine and natively on Linux, there are hardly any dependencies outside of system libraries and a DSP library.
Sadly, after moving my music production setup from Windows to Linux, I'm locked out of some of my expensive sample libraries because while the plugins run fine, the licensing programs do not. Very frustrating.
Solution: don't use those apps and maybe people will learn. Eventually, apps and technologies like this die in our digital landscape. Rest in piece Flash, you will be missed. 3D max and Photoshop, you're next.
Real solution (for now): just don't give these assholes money. If you need to run the software, fine, but at least have the decency to steal it.
I happily pay Autodesk their stupid $600 a year because I get that much value out of the application and then some.
This idea that they are purposefully hostile because they don’t want you to steal their commercial product, or they don’t support an operating system with 2% marketshare is ridiculous. I totally understand why they don’t support Linux. It’s my choice to use an incompatible system.
Yes, you will, but big picture: eventually the software will die.
Software isn't just a product, it's an extension of education and expression. Using paid-for commercial software is risky.
What happens is young, aspiring professionals get taught that software, and then it's ripped from their hands because of the price. A lot of these young professionals give up or, more likely, switch to OS alternatives. And they make up the industry of the future. Eventually, the commercial product is pushed out because it can't keep up.
But even if this doesn't happen, the product still dies. Because commercial software is ultimately held hostage. The BEST CASE scenario is you're eventually bled dry and squeezed. The common scenario is that the software disappears or isn't popular anymore, and you've lost hundreds of hours of skills.
This even happens in dev spaces.
If you chose COM+ for an application, you fucked up. You're stuck with it and it's painfully behind, well, everything.
Or maybe you were seduced by how cool Oracle is. And it was... 20 years ago. Now, it's not a competitive database. But you're stuck paying for it. Had you just used MySQL or Postgres from the beginning, this wouldn't have happened.
And then flash. Oh flash. Great technology, but it doesn't matter. The world moves on. If you chose flash for your website, you fucked up.
And now 3D max. I give it 10 years. Blender is already ahead of it and all the professionals of tomorrow have jumped ship.
If you think companies like Oracle are going to die, go ahead and short their stock. It's just about doubled in the last year. Good luck!
I don't follow. All proprietary software IS destined to die.
Even extremely popular proprietary software. Windows 10? That's destined to die. iOS 14? Destined to die.
It's inevitable. If you're lucky, the overlords making the software will give you a new alternative. Hopefully the new alternative has all the functionality and isn't 100x more expensive.
But OS doesn't work this way, because it can be forked.
The way it works is that, in the beginning, proprietary software is really good compared to OS. As time goes on, the proprietary software rots.
It either stays the same, obsolences, or actively gets worse. Usually the last one. OS software, by comparison, gets better and better. Eventually they intersect, and if you chose the proprietary tree, you lost.
I thought when you were making the argument that all proprietary software is destined to die that you were actually making a real argument to say that it will die because it will become inferior to open source software that customers will prefer.
But no, actually, you’re making a completely pointless argument about how a specific version of software will die because it has been surpassed by a future version of it, and you’re just cynically assuming that all proprietary software eventually gets worse, that you’ll actually want to go back to an old version and fork it to maintain it.
But this argument is just plain false for a lot of people who use that software. No, my Autodesk isn’t going to “die” because it’s updated monthly and I can’t use the old version, just like a McDonald’s doesn’t “die” because the hamburger I ate from them yesterday doesn’t exist anymore.
Well that’s just Adobe badumtsssss
Agree that modern apps are hit and miss.
Done
I like using two identical miniPC's, one for each monitor.
Well, actually each monitor has two inputs and each PC two display outputs, and I had a couple extra cables so they are cross-connected too but that's besides the point.
Seems like RDP is almost intended to work like this from the beginning. Deficiencies are a lot easier to tweak side-by-side too.
Decades ago I just had to accept that a key purpose of introducing multi-partitioning to HDD's was so that multibooting from a single HDD would be extremely straightforward. And once set up, very closely mimics the hardware behavior of having a dedicated HDD or SSD for each of Windows and Linux, on the same PC.
Previously, with two different HDD's connected, each completely unaware of the other one upon power-up, when you reboot you can always use the motherboard's built-in BIOS boot menu to choose when you want to boot to a drive other than the one designated as the default choice.
That way there is nothing related to Windows on the Linux HDD at all, and nothing having to do with Linux on the Windows HDD. You can physically remove either drive before powering up and everything works completely dedicated to a single OS as expected, because each HDD is complete including its own boot files, exactly the same as it is in a non-multibooting arrangement.
As long as each HDD is capable of booting on its own, you choose the one you want, and that's the one that boots.
Well it actually took a while in the '90's before most motherboards had a built-in BIOS bootmenu to choose between different HDD's, but this feature became universal so users wouldn't have to physically reconnect their intended boot drive to the Primary Master cable. Which was the only bootable connection before the BIOS bootmenu made Secondary-connected HDD's as bootable as Primaries, your choice. You don't really have to get the most out of the electronics, but some things like this are really nice to have.
Now this was the time when it got real fancy, and both Windows and Linux bootloaders were crafted to accommodate "chainloading" from a Primary HDD to a non-Primary, so physical reconnection would not be necessary to accomplish the same behavior. This was ideal for all the remaining motherboards at the time which were not issued with a BIOS bootmenu. This is where you start to get a mixture of Windows and Linux on the same HDD, at least in the boot files. It doesn't have to be confusing, but it can be.
Once one set of boot files can boot either OS from any HDD, then each HDD no longer needs its own boot files, however that also means that those HDDs not having boot files would not boot if they are the only HDD connected.
I say the BIOS bootmenu is the fundamental that is best not abstracted too far.
Fortunately, multibooting to various SSD's using one single (Linux) bootloader [0] can be configured to have the same hardware workflow as choosing separate HDD's through the motherboard bootmenu.
And to be the most consistent I like to use the same workflow to choose from multiple partitions whether they are on the same HDD or not.
Now you can figure it's all moot, with separate miniPC's for Windows and Linux. Which really could be considered more of a luxury than multibooting a single-drive PC at will, and even more versatile than having two SSD's in the same PC.
But wait a minute, each one of these drives on each PC is a massive multibooter . . .
[0] The Windows bootloader works as always on MBR-layout SSD's on PC's supporting traditional BIOS mode, but still too defective under UEFI, where Microsoft drops the ball completely since Windows 8 in the key area of multibooting Linux. Which for decades was as easy as intended by the hardware design. But negative progress is accepted as progress by those who are supposed to be experts, as we have been convinced.
My main use is work, so I plug in the dock’s C cable and then my second monitor to the laptop HDMI. When I want to switch to my desktop I just unplug the USB-C from the laptop and pop it in the top of my tower.
Back when I had more free reign on my work laptop, I used Synergy to use them both at the same time but that was mostly so I could do “x personal thing” on my desktop while responding to Slack all using the same keyboard and mouse - without it I have to just stretch a little to the laptop keyboard.
I tried everything. I tried some dude’s GitHub project to get it to work with Wine. It’s just not working for me.
Something like this seems perfect for that use case.
Sorry boss I can't do work today, I decided to go full Linux and our CRM doesn't support it!
Every one of these needs more intense tweaking before it will run as well as the same offices 20 years ago.
Too bad most users are locked out and IT may not know how to do it or may not be motivated anyway.
It may even be at the point where less tweaking may now be needed for Linux to become a higher-performance office machine than Windows/Office was 20 years ago. With less undocumented effort than it would take to get the same performance from the latest Windows. But who's going to do it?
All other things are not being equal though, 20 years ago PC's were lower-performing hardware in a number of ways, so that probably should be brought under consideration.
But it just seems so unfair then.
Telling people not to even think about using their favorite piece of software is a good way to make sure they don't consider switching. A lot of popular Windows apps run perfectly fine in WINE. I've been using foobar2000 in it for a decade at this point, and have yet to find a native alternative that gives me the same feature set. So why shouldn't I keep running it?
This is one of the big, but less obvious, benefits to Wine/Proton. Games with native Linux builds run into all kinds of distro-specific issues that you don't really get on Windows. It's an issue for new games and an even worse issue for older games that aren't being updated anymore. Just look at Steam on macOS to see how big of an issue this is - so many games are not compatible on the latest Macs because they were built for x86 (32-bit).
My experience has mostly been that Linux native versions just aren't as good as the Windows-on-Proton version. (Shout out to Larian for their recent BG3 release, a much better native version.)
Totally agree that Proton only works so well because of the constant effort that Valve put into it.
Shouting at game devs to make better native Linux versions isn't going to work. What will work is that the market demographics are slowly moving over to Linux, mostly thanks to Valve, Proton and the Steam Deck.
> All modules that call a Unix library contain WoW64 thunks to enable calling the 64-bit Unix library from 32-bit PE code. This means that it is possible to run 32-bit Windows applications on a purely 64-bit Unix installation. This is called the new WoW64 mode, as opposed to the old WoW64 mode where 32-bit applications run inside a 32-bit Unix process.
They will never ever receive native Linux ports.
Understand what each OS is good at. Back in my younger days, experimenting with Linux was my defacto CS education.
I use desktop Linux when I don't want distractions I need my computer to do what I want it it.
Window's is much much better for music production. I'm not switching DAWs.
Primarily I'm a .net developer, I NEED Visual Studio to really be productive.
OSX is when I have an important interview or something. Although I did interview using Fedora recently. Fantastically stable distro!
You don't take a Lambo off road, you generally don't take a Jeep to the race track.
This is terrible advice. Many people want to use Windows apps while using Linux, and Wine works just fine for that. And for those few that don't work in Wine, dual boot works great.
Rather: don't try to be compatible with inherently unstable APIs:
https://sporks.space/2022/02/27/win32-is-the-stable-linux-us...
https://blog.hiler.eu/win32-the-only-stable-abi/
Just to be clear: I consider it to be a good idea to write native apps for GNU/Linux, but first stabilize the APIs so that they stay basically stable for at least 20 years.
But also, compiling everything with dynamic libraries is kind of an interesting side effect of having all the source code in a single context of a distribution; maybe you should always statically compile if you're not part of this system.
Under Windows, this is also done. But there is a difference: changes in a system-wide DLL better are really backwards-compatible.
If there are breaking changes, a new "package" gets introduced with new names for the DLLs. This is why on many Windows systems, lots of versions of, for example, the "Microsoft Visual C++ Redistributable" are installed. Nevertheless, the old versions are still available and get maintained for a very long time at least with respect to security fixes.
Also, the API design under Windows tends to be much more "future-proof". For example, a lot of data structures contain some size information as a first field so that the API can detect which "version" of a data structure has been passed so that future changes can be implemented in a backwards-compatible manner, e.g. WNDCLASSEXA:
> https://learn.microsoft.com/en-us/windows/win32/api/winuser/...
The first field is cbSize.
To me, it seems in the GNU/Linux ecosystem, the API developers care a lot less about such topics.
My recommendation for happiness with Linux is: Always use native apps. Don’t use WINE. Don’t try to be compatible to inherent hostile things. Don’t use VMs. And especially don’t use Dual-Boot. It sucks.
Basically migrate and go full Linux. Don’t look back :)
Proton (which is WINE derivative) works somehow, because Valve invests every single day tremendous efforts into it. But that’s the problem, tremendous efforts.
The good news. Every bit invested in high quality API/ABI on Linux pays off. Valve contributions to MESA and amdgpu are invaluable. Valve should honor native AAA-Titles and Indie-Titles for Linux - with exclusive Steam Awards. There is awesome stuff like Unrailed. Make the game developers think:
PS: I missed Counter-Strike so much on Linux for years. And the Valve came, ported everything natively, and it is wonderful :)PPS: I use a Mac for two incompatible applications (Garmin Express and Zwift). Less maintenance than Windows. Less possibilities than Linux. Horrible file-browser. Window management is a pain. But it covers the gap without ruining my day. I have to admit, the Mac cannot run Counter-Strike 2. That’s a task for Linux :)