> What I want to highlight is that for some people (including this author) running Linux on some machines make them less useful than whatever they begun with.
Ok so this is a pro-choice defense.
I wouldn't try to counter-argue that perhaps many people would be more capable on whatever they use today. There's the fact that Linux hasn't been a particularly well supported possibility for people to mature & develop (themselves) on of course factors in here, that most people are path dependent on what is available, but whatever.
To me a lot of the assessment boils down to whether we care about what helps us the most today, or whether we have time & interest & capacity in getting better in the long term. This author, like many others, would undoubtedly do better with what they already know, today.
But over time, the advantages of living in Linux, in learning a posix-ish shell well, in coming to understand & thrive with the OS you probably deploy on: those long term investments, in learning a learnable, open source, malleable system, make you a better wiser more experienced dev to who better understands how your application fits together with the rest of the computer, and it probably will teach you a bunch of great tricks along the way. A lot of time it doesn't really matter, and you can happily develop away, leave the funny bits to someone else, continue cranking happily away in VSCode either way, and this investment wouldn't be returning. But there definitely is something to be said for a holistic approach, for gaining a better perch on computing & picking an Open Source way.
But if we only evaluate how productive we'd be right now, as this article certainly does, and ignore future prospects: everyone's going to pick what everyone already saw in the world, and no one is going to invest in learning Linux. You have to want to be able to go further. That's why it's exciting seeing big companies like HP starting to ship Linux laptops.
The author has their own defense of not liking Linux.
> Heck, I don’t know how people can think that their subjective personal experience running Linux can simply be extrapolated to be applicable to 100% of the developers in the world. There is no pinnacle of computing, there is no perfect machine for developers. Each person will have a different idea of what they want and what is good for them, let them be.
"There is no pinnacle of computing" is something I agree with strongly, & is, in my view, by far one of the best & strongest advantages of Linux. There's such a wide gamut of Linux experiences, from GNOME to PopOS to Sway to xmonad to Sugar/XOPC, there's a lot of "something for everyone", many different styles of Linux, and great adaptability within each system. We can layer in the more outside contenders too perhaps... GoboLinux, ChromeOS, WebOS... Linux isn't the pinnacle, but it certainly looks like the right platform for building peaks, has the most versatility & potential.
Perhaps Linux's worst fault is that Linux challenges us to explore in ourselves what we want, and what best serves us, as opposed to being one well known paved road we all travel down.
> I like open-source, I felt that I needed to run a FOSS system. I was mostly running Linux due to self-imposed expectations and peer pressure. I was not happy with it at all. That was not because there is something wrong with Linux, it was because what I wanted was something different.
I'd like to see a little more elaboration. They've described a host of external factors, said Linux failed to be what they wanted. But they haven't at all labelled what wasn't working as they wanted, how they tried to adapt Linux or themselves. It feels like a very pat forgone conclusion; it doesn't allow for progress or change: there's no possibility Linux could ever improve it's way through this, nothing the author could have done to improve themselves. This iconifies what feels to me a lack of useful genuine candor & argumentation throughout this article, makes it nothing more than a huge example of what the author is most protesting: social pressure, just not in a pro-Linux direction.
Thanks for taking the time to compose such deep feedback. :-) I appreciate it a lot. I didn't go deep into why Linux didn't work for me because the article was already too long and I felt no one was going to read it. I'll summarise here and if you feel it deserves a longer post, I'll try to add it to the blog.
First it is important to contextualise that I am talking about desktop workflows, even desktop development workflows, not server stuff. I think that Linux makes a terrific server and platform to develop industrial and hobby stuff.
It just didn't click with me for desktop usage for the following reasons:
I come from MacOS. I used Macs all the way back to MacOS 8. I enjoy Mac ways of doing things, looking for MacOS in another system is recipe for frustration. That is true for any kind of looking into a system wanting it to be something else. Someone who loves Linux and keeps trying to find it in a Mac will be equally frustrated.
I feel that desktop applications on Linux are not as polished as the ones on a Mac. That is not true for every single application of course, there are many gems on Linux, but due to the way that native Mac apps are built and able to leverage an Apple provided collection of features and "components", they tend to look and behave in ways that feel more in tune with the system.
On Linux it feels like a patchwork of technologies and toolkits that not necessarily talk with each other well. I like the integration of MacOS and what we used to call Cocoa, how apps would adhere to a HIG and you could trust features to work across them.
I never liked the battery life on any Linux machine I had (mostly macs running Linux including PowerPC ones, some PCs). To be honest, BSDs always worked better on a Mac for me than Linux.
I wholeheartedly agree that Linux is a wonderful base to build something. GoboLinux is a great example of that, it solved one of my pet peeves with Linux which was the messy filesystem. WebOS was the best operating system for mobile devices I have ever used, I had all the devices, they were wonderful.
But on a desktop it always felt short in my subjective experience. I don't have the same love as the majority of the developers for a terminal. I'd prefer to use it only as a last choice, Linux is built in a way that it rewards command line usage and composition of command line based tools. That is the correct way to use it, just look at how much loved minimal tiling setups are these days and how much of terminal activity happens in them. I prefer rich gui apps.
I also am very fond of some old ways of Apple third-party development, when apps would all ship with AppleScript support and I could do my workflow pipelines by connecting them instead of writing a shell script.
It boils down to desktop apps being the star of the show, and the OS vendor shipping a toolkit for developing the that bakes in a lot of features which makes it easier for said apps to integrate well with the OS and each other. Linux has a lot of choice in that and choice is good, but it often leads to apps that feel like they were built for different systems running side by side with totally different UIs, menu placement, and in the horrible past different clipboards (that last one works fine now).
Anyway, it circles back to what I said at the beginning, it didn't worked for me because I kept trying to have the kind of experience I have in MacOS in a different system. As I said on the original article, I'm happy that people enjoy Linux. I'm not trying to get them to change. I run it for many years as well, I just prefer something else. It is OK to prefer something else.
As for "pro-Linux", I'd rather let people chose whatever they want to run. If it is Linux, I'll cheer them all the way, if it is something else, I'll still do the same.
I find a bit disingenuous the section where you say:
> everyone's going to pick what everyone already saw in the world, and no one is going to invest in learning Linux. You have to want to be able to go further.
It felt a bit like subtexting, it might not be the case, but in my personal case I gave it a try. I'm quite decent Linux user. I still maintain my Linux servers, I still have Linux machines. It is not that I didn't invest in learning it. I learned it and decided on something else.
If I was to run something other than macOS and wanted to go the FOSS route, I wouldn't pick Linux. I'd probably try to stick with Haiku which for me is a much better desktop experience than Linux. I actually have a Thinkpad x230 with Haiku these days instead of Linux, it is my secondary machine.
Ok so this is a pro-choice defense.
I wouldn't try to counter-argue that perhaps many people would be more capable on whatever they use today. There's the fact that Linux hasn't been a particularly well supported possibility for people to mature & develop (themselves) on of course factors in here, that most people are path dependent on what is available, but whatever.
To me a lot of the assessment boils down to whether we care about what helps us the most today, or whether we have time & interest & capacity in getting better in the long term. This author, like many others, would undoubtedly do better with what they already know, today.
But over time, the advantages of living in Linux, in learning a posix-ish shell well, in coming to understand & thrive with the OS you probably deploy on: those long term investments, in learning a learnable, open source, malleable system, make you a better wiser more experienced dev to who better understands how your application fits together with the rest of the computer, and it probably will teach you a bunch of great tricks along the way. A lot of time it doesn't really matter, and you can happily develop away, leave the funny bits to someone else, continue cranking happily away in VSCode either way, and this investment wouldn't be returning. But there definitely is something to be said for a holistic approach, for gaining a better perch on computing & picking an Open Source way.
But if we only evaluate how productive we'd be right now, as this article certainly does, and ignore future prospects: everyone's going to pick what everyone already saw in the world, and no one is going to invest in learning Linux. You have to want to be able to go further. That's why it's exciting seeing big companies like HP starting to ship Linux laptops.
The author has their own defense of not liking Linux.
> Heck, I don’t know how people can think that their subjective personal experience running Linux can simply be extrapolated to be applicable to 100% of the developers in the world. There is no pinnacle of computing, there is no perfect machine for developers. Each person will have a different idea of what they want and what is good for them, let them be.
"There is no pinnacle of computing" is something I agree with strongly, & is, in my view, by far one of the best & strongest advantages of Linux. There's such a wide gamut of Linux experiences, from GNOME to PopOS to Sway to xmonad to Sugar/XOPC, there's a lot of "something for everyone", many different styles of Linux, and great adaptability within each system. We can layer in the more outside contenders too perhaps... GoboLinux, ChromeOS, WebOS... Linux isn't the pinnacle, but it certainly looks like the right platform for building peaks, has the most versatility & potential.
Perhaps Linux's worst fault is that Linux challenges us to explore in ourselves what we want, and what best serves us, as opposed to being one well known paved road we all travel down.
> I like open-source, I felt that I needed to run a FOSS system. I was mostly running Linux due to self-imposed expectations and peer pressure. I was not happy with it at all. That was not because there is something wrong with Linux, it was because what I wanted was something different.
I'd like to see a little more elaboration. They've described a host of external factors, said Linux failed to be what they wanted. But they haven't at all labelled what wasn't working as they wanted, how they tried to adapt Linux or themselves. It feels like a very pat forgone conclusion; it doesn't allow for progress or change: there's no possibility Linux could ever improve it's way through this, nothing the author could have done to improve themselves. This iconifies what feels to me a lack of useful genuine candor & argumentation throughout this article, makes it nothing more than a huge example of what the author is most protesting: social pressure, just not in a pro-Linux direction.