Re: Why do new people give up on Linux?
Posted: Mon Jan 18, 2021 12:45 am
I can give you my perspective as someone who has recently switched to Linux.
I've always been interested in computers and programming a degree. I wouldn't say I am a "power user" at all, but I have been definitely slightly above average on computing skill.
There are a few main issues in my opinion with standard users making a clean switch to Linus as a primary personal OS. They can be summarized as follows:
1) The idea of Linux and a "Distro". Most people don't understand what Linux is. They think it is a singular thing. They are confused by the concept of Linux and different distros and what it means for them, their device and their experience. As a result, they don't even know how to start unless they get pointed in a clear direction kindly by a Linux user suggesting a proper distro for their use case.
2) Stuff working out of the box. Linux mint 20 has been a great experience for me. Everything is working very well out of the box. But in the past that was not my experience with other distros (even old versions of Mint years ago). The WiFi isn't working, keyboard lighting isn't working, etc. Even some relatively benign "bugs" cause the user confusion, and makes them question whether things are working properly at all. Most of the issues can be fixed, but people don't know how to go about doing it (and they think they shouldn't HAVE to do it if this is a quality software)
3) User experience / aesthetics. I will again say my experience with Mint has been fantastic. But in the past this was not the case. Things were confusing, poor tutorials (if available), codecs not being available, and things looking a bit old and clunky. The linux community has traditionally valued function over form, and that totally makes sense in many instances. However, we need to understand human psychology. People don't want the best appliance if it looks like an old clunky device. They are allured by the latest and greatest shapes, curves, colors etc. I'm not saying this is GOOD for the user, but this is reality.
4) Work arounds. I had a linux machine many years ago. I gave it to my wife since it was super small and light for her to take to her school for showing presentations etc. It had a problem with the projector, multiple monitors etc. It had a work around but it was a pain in the butt. Even though I could clearly use it, she didn't want to deal with using terminal and working around what should be a plug in, click and go. I can understand this frustration and myself feel someone embarrassed when explaining the value of Linux and at the same time explaining why I can't simply plug in a monitor and go.
5) Games / software. This is an honorable mention because I think 1-4 are 90% of the problem. If they were solved I think most people would be happy browsing, checking mail, and using office 365.
I know this is getting better with Steam on Linux etc. But the stigma is there that games don't / can't run on Linux. Also many proprietary software (CAD, financial, MS Project, etc.) used in industry is not on Linux making the switch impossible for professionals. That being said, I think over time this will slowly go away as apps become more and more web based (we are already seeing it with Office 365, MS Project PWA, etc.). I'm even somewhat hopeful long term streaming services like Stadia type services will be available so gaming on a linux machine with low end hardware is as good or better than a Windows PC.
Re: Why do new people give up on Linux?
Posted: Fri Jan 29, 2021 11:47 pm
I can't resist, many good comments, but didn't read the entire thread.
I can say a few things, I've been stuck helping windows and osx users fix their issues, which to me is fairly comical because I haven't even seen windows 10 beyond a quick look or two over the years, and the only time I ever see osx is when I'm stuck with the unenviable task of helping someone work through the super shoddy excuse for programming I find the apple corporation puts out to its totally uncritical and abuse accepting fan bois.
So I can tell you why I give up on osx, it's badly written software, extremely unstable, and very hard to get to do even a fraction of what I do without even thinking about it on GNU/Linux. Plus it makes me want to hurl the device across the room every time I interact with apple's stuff. So that may offer a slightly different take on the assumption that the software put out by microsoft or apple is actually that good or robust or stable in the first place.
My second problem with this type of claim is that I've done, or did anyway, windows based networking for a living for a while, at least part of my living, and the notion that windows systems are particular good or fit for purpose implies you've never encountered a hive corruption, a total failure on motherboard replacement, or a host of other issues fairly specific to the entire windows corporate model of computing operating systems.
As I like to tell people when I get stuck fixing their annoying windows or apple problems, thank you for reminding me that no matter how frustrated I may get by some gnu/iinux issue, the alternatives are much worse, so it's worth keeping that in perspective sometimes.
Linus Torvalds once noted quite correctly that all operating systems suck, they just suck in different ways, and gnu/linux sucks a little less than the alternatives. But that again depends on which type of sucking you find worst, that will dictate, or should dictate, your os choice. If you want stuff to be fixable, then windows is a bad choice. If you want robust reliable software written to a very high level for users who are super demanding, then osx is a terrible choice. If you want to have a long term fairly stable api for installing a massive range of free and non free software, then linux is probably a bad choice. And so on.
Speaking for myself, when I first switched more or less full time to a gnu/linux os, I remember a few things quite distinctly, one being that kde 3.5 was everything I had ever wanted windows 2000 to be but it wasn't. Second was opening a 1 MB big text file in a code editor and thinking something had gone wrong because rather than watching the opening little icon bob there for some 60 seconds while windows tried to figure out how to handle such a big job, it was open instantly. So fast that I was sure it had not even opened since I'd never seen anything like that on windows, ever. And that was before the even slower, and cludgier, windows xp, with it's virtually non existent ability to do serious multitasking became mainstream. In fact, for many years afterwards, when I'd start up a friend's system, and later with vista too, I thought it was broken because it was taking so absurdly long to boot
If you've ever done any type of support, you should be aware that windows users give up on windows constantly, it's called if they are tech savvy, reinstalling,, and it not, getting a new computer, because windows has grown so corrupted and rotten. OSX does much better in this regard, being a mutated unix at core, though they go way too far in their walled garden catering to non critical users approach for me to ever contemplate using it. The giving up however of average users isn't recognized at all by many since they are still running windows, they just gave up on their last install, over and over. I don't give up in general on linux installs, but I have given up on many windows installs over the years because experience showed me that once a certain situation or set of situations occurred, it was very unlikely the install could be saved.
With this said, I think there's a HUGE mistake being made fairly continuously by the linux community, which is ignoring the free as in freedom, and open as in open source, nature of the free servers/desktops, whether bsds, linux, whatever, and trying to focus on the free as in beer, and the fantasy that it's a drop in free of cost windows replacement, which it isn't. Which is a good, possibly even a great feature.
I'm trying to find a windows os where I can pick between i3, xfce, or something else, as my desktop, running 4 to 8 virtual desktops natively, no problems, no issues, it all 'just works', where I can take an install, move it to another motherboard, reboot, and tweak a network setting, and it's all ready to go. In other words, this is what I want an operating system to be able to do, and, oddly, it's what the engineers who spend their time, paid or unpaid, to write this software, want their desktops to do. So because they spent the time on their priorities, the desktops do what they want them to do, in a robust, stable, almost comically so at times, way.
It's a strange thing to try to explain to people that yes, really, there is no 'they' out there making a free windows replacement, there is a bunch of people, some working for companies, some for organizations, some for themselves, who are making the stuff work they way they want, and since they spend the time or money to do that, they get first vote on how it works, runs, functions, etc. In other words, gnu linux tends to favor tech oriented people with decent problem solving skills, who can read the f#cking manual, or at least, who won't panic at having to do that now and then.
I could not run my free desktops without at least once a year researching a fix for an upgrade issue, it's usually not a big deal to fix it, and given I worked with windows and am forced to touch macs now and then, I have no illusions about reality, I've seen the failures of windows, and I can guarantee you, when 99% plus of people encounter them, they give up, and either pay someone to try to fix it, or buy a new computer. If you don't know what an event ID is, or that microsoft does not give even come close to reasonably useful documentation on event id error numbers or causes, you also probably don't know which third party sites, like event-id.net, you have to use to try to resolve these thorny internal windows failures and bugs. And even then, you often can't fix a bug that microsoft says was fixed years ago but which isn't fixed. In gnu/linux desktop/servers, the odds are very good you can in fact fix it, and find the solution.
I think by now the split has gotten fairly clear, some 80+% of users basically find windows good enough, or osx so annoying they'd rather have at least some control over their system. Another 10-15% of users want that full corporate walled garden, and will pretend that all the failures and issues they encounter aren't there, and that apple is wonderful. Or are satisfied with the osx pseudo unix, somehow. Then a very small slice of more tech savvy users who want to have powerful robust operating systems that can be maintained over years, call it 2% or so, run straight gnu/linux desktops, or even more rare now, a BSD. Not including the android or other google OS type systems, which are I think designed more like smart phones in general, to be thrown away after a while, while google handles your system.
So this probably very roughly defines the overall end user computing market, it hasn't changed much for a decade, so it seems safe to say we're roughly where we will be.
The guys who write the stuff that runs your desktop prioritize certain things that average computer consumers care less about, and the linux engineers care less about the stuff that normal consumers care more about, so those things don't work, or don't work as well, as they do on the fully corporate non free desktops that pay thousands of programmers to make those things work, at least superficially. That goes back to what matters to you as a user, if what matters most is the feature sets developed by the microsoft or apple corporations, then why are you even trying to use linux in the first place? You have to be clear on what matters to you, then pick the os that fits that most closely, and accept that the others won't be as good a match, and go on with your lives.
I mean, if linus cared deeply about kernel api stability, which the core guys in the linux kernel claim, nonsensically, to care about, then you could quite literally take any driver written in the last 10 years and pop it on the kernel and it would just work. But they don't care about that, they care about stuff working today, and in particular, they care about the stuff that the companies that pay their salaries care about, heavy duty server stuff, so you can rest assured, linux is totally ontop of the latest 100 core ARM servers running pcie nvme drives, but they don't really care if your nvidia driver fails on the newest kernel upgrade, nor do they care if your old printer works.
This is why linux, as the kernel, is so comically and insanely robust and powerful, and why linux runs most of the world's servers now, and I think maybe all of the world's super computers? That's because to engineers, these are interesting problems. gnome sucking and breaking again with gtk 4 is not an interesting problem, nor is it likely to attract the best or brightest developers. So the consumer desktop side of linux works about as well as the core developers want it to work. If you know what to look for, which is good documentation, good code, then you can also find a good desktop or window manager, like i3, because it's written by people who care about that problem and have the skill to execute that caring and maintain it over years.
When I started inxi, I was solving a set of problems I could not readily solve on windows when I was a sys admin, basically a lot of the stuff I put into inxi to this day is for sys admins, but also for end users, since I've been both, I make it cater to its main user, me, and then to set of users who I know and trust. This is how the linux kernel also works as a project in many ways, though their interests are much wider of course since they get paid for the work and cater to the interests of those who pay them, which tends to be server oriented stuff. Server oriented stuff reflects in far greater robustness and stability of the kernel running your desktop, but i don't believe that's the main motivation, the main motivation is to run servers really well. osx gave up completely on that market, and boy does it show, junk code, junk stability, junk robustness. Nobody to keep apple honest anymore now that they are 100% consumer software.
So people give up on linux because it was misrepresented to them, by well meaning people, and they shouldn't have been using it in the first place. Usually because of an upgrade failure, that was a problem I tried to resolve with a core group of people for years, and we almost had the problem licked, but it was a full time job, and then we all moved on, but it was a problem that could with cooperation be solved, never has been though. It's better now than before, but still not that great. One upgrade failure is all it takes most users to give up. Or some windows hardware not working, and them not understanding that if they got the right hardware for linux, their issues would be over, permanently. Or they need to use fancy printers or scanners, and get the wrong ones. Or they want windows but for free.
So my long term view now is, people aren't giving up on linux, they weren't meant to use it in the first place, they discovered that fact, and moved on to what they were or are meant to use, which is totally fine.
It's not a religion where it dies if you don't gain more adherents, taking them from the other competing factions, it's just a bunch of people writing the os they want, the tools they want, the desktops or window managers they want, and if those happen to meet your own wants or needs, then you are indeed in luck!!! But if not, it's not like you paid for it, and have some claim to be able to demand x or y from the developers, what did that person themselves do to make it more the way they wanted? Odds are good they did nothing, and prefer to complain about linux users, because they do not understand that this is FREE/Libre software, and they have no right to complain at all, that's a somewhat disgusting sense of entitlement that I do not fully understand, I mean, yes, sometimes stuff sucks, really bad, and lord knows, I've complained, but to me, if you aren't doing anything other than taking, then your rights to complain are very small.
This is, again, what I view as the biggest mistake free software based distributions make, not making it clear to their users that you cannot compare corporate run non free systems like windows or osx with free software, because it's apples to oranges, you can compare non free with non free (osx vs windows), or free with free (like comparing reactos with gnu/linux with freebsd), and work to make free more the way you want it to be, but for instance, take me, if someone asks me to do something really dumb with inxi that I don't care about, why should I do it? they aren't paying me, I think it's a dumb idea, so why on earth should they believe they have a right to demand my time for free like that? Total confusion, sigh. Focus on free as in freedom is my advice, focus on making the stuff you like as good as you can make it, focus on the problems you can solve, if it's support, give the best support if it's code, give the best code.
If the coders aren't out there willing to solve a problem then that problem is not important to them, and that's how it is. Coding in the real world is worth maybe $100 an hour and up, contemplate that next time you reflect on why something on a consumer gnu/linux desktop isn't working perfectly, whose job is it to fix it? why should they, particularly if it's an annoying and not fun problem in the first place? Since engineers have a much greater role in determining the BSDs, Linuxes, etc, futures and features, those operating systems and tools are going to reflect what matters to them more than what matters to an average windows or osx user, and there's very little point in trying to whitewash that fact. I don't have queues of people lined up offering me inxi patches, so if I don't do something in it, it's unlikely to get done, and that's how it is.