Updated: August 15, 2016
Several weeks ago, I published an article titled Linux 2017 - Road to Hell in OCS-Mag. Predictably, the piece was met with stiff resistance from diehard fans, bullet-focused Linux users who are unwilling to admit the fact that our favorite operating system has been flaking lately. Like a lot. The linux desktop is not in the bestest of shape.
But that's not why we are here. Disagreeing is fine. It's healthy. What really shocked me, like OMG WOW WTF like is the level of reading comprehension difficulties, yet again. We are talking Distrowatch comments, Reddit comments, organic article comments. Even though I'm possibly giving undue attention to things that need not be repeated, we do need to look at this phenomenon once more. It is the primary reason why the Linux desktop sucks. Because if you are unable to accept feedback, you can't progress.
Reading comprehension issue 1:
All his tests are done from live CD.
I am paraphrasing and omitting some of the grammar and spelling mistakes. but a large number of people think all my tests are done in a live session - and therefore, my results are wrong.
No. YOU are wrong.
First, if Linux has a live session, it should work - otherwise there's no point providing a live session for people to test, now is there? If you can't do anything fun, why would you bother installing?
Second, why do you think there's a difference in how the system works when it's installed versus live CD? Do you even understand how the operating system works? Do you perhaps understand how memory management works? Hardware? Drivers? Of course not. Because if you did, the claim would not be there.
But that's still not the worst part. The WORST part is that people do not bother reading articles before responding. If they did, they would realize that all my tests are done TWICE, in the live session AND in the installed system. That's mega fail.
Repeat after me: All of Dedoimedo's test include live session, install and post-install testing and configuration. All of Dedoimedo's test include live session, install and post-install testing and configuration. All of Dedoimedo's test include live session, install and post-install testing and configuration. All of Dedoimedo's test include live session, install and post-install testing and configuration.
Very simple. Installed. And the Linux desktop still sucks. SUCKS.
Reading comprehension issue 2:
His battery is degrading. I stopped reading after that.
No. You see science is such a beautiful thing. It accounts for all sorts of things, including something called a controlled experiment. One of your measurements serves as a baseline, a control to which you can compare. You see? That's the magic word. Comparison.
So let's explore battery life in more detail. Trusty, 3.5 hours, Xerus 2.5 hours, Sarah 2.5 hours. At a first and hasty glance, you may think that indeed the battery has deteriorated by about 35%.
What about MX-15, which manages some 4.5 hours?
What about Windows 10, which manages some 4.5-5 hours?
What about Fedora 24, which manages 4 hours - and tested AFTER all of the above?
All of this done on the SAME hardware. SAME laptop.
Mathematically, as I've explained in a comment:
A + B = R1
A + C = R2
So we have component A plus component B gives Response 1. And likewise, A + C gives a different Response 2. This means that A is a zero (null) operator in this set of equations.
And A in this case is the Lenovo G50 laptop. Because with distro X, it has one battery life result, and with distro Y, it has a different one. And measured at the same time, so the temporal degradation of the battery is not an element in the equation.
Reading comprehension issue 3:
STFU, and buy a new laptop (from this century).
Not only is this condescending toward hundreds of millions of people who cannot afford to purchase new hardware on a whim, it also ignores the scientific method. Once again, we go back to consistency. Lenovo G50 is what it is. A laptop. It may be good or it may be bad. That's not relevant. What should be relevant is that every single Linux distro provides the SAME CONSISTENT result when running on this hardware.
This is your scapegoat, but you're wrong.
In other words, the network driver may not be the best - but if it behaves differently across four different Ubuntu-based distributions released over a span of less than a year, then the failure is entirely 100% with the software. Furthermore, we also see additional discrepancy in how CentOS and Fedora behave compared to Debian-based systems. Speaking of which, the laptop was manufactured in 2015, and for the first half a year, a large number of distributions struggled to boot on it, because UEFI is so difficult and complex, after all it's been with us only since 2001.
Reading comprehension issue 4:
Samba does not work for him.
Samba works for me. But after some stupidly-named vulnerability in the business space a few months ago, Samba was patched in a shitty manner, and now anonymous access is no longer an option. Furthermore, this has NOTHING to do with hardware. This is security paranoia that cripples the user experience.
The whole breakage pattern is not just limited to sharing files with Windows system, yes for the 90% of the population that still runs it, it also affects media codecs and phones. The things that used to work, work no more. Or differently. Or some random crap.
Reading comprehension issue 5:
Who cares if XYZ does not work?
I do. I am not a 16-year-old child. I am an adult, and I respect the money that I earn. I do not have the time nor patience for ideology bullshit games. Products, whatever their origin or price tag are, must work. That is their purpose. If they do not work fully, seamlessly, then they are shitty products and they need to be purged from existence.
Reading comprehension issue 6:
Never. I care about Linux. Which is why this kind of amateurish randomness pisses me off on a monumental level. It smacks of a complete and total lack of vision, drive, passion, or any regard for users. If you're producing something that will be consumed in public domain, please try to pretend you give one quarter of a shit about your audience. If you think having a software product means having a GitHub repo that is actively updated, you are mistaken. Code is the least important piece in this whole equation. Code is irrelevant. Code must be transparent.
Reading comprehension issue 7:
Buy a laptop that is supported.
No. My life does not revolve around choosing software. It resolves around my goals. Software is a means to an end. I will buy a laptop that fits my needs. Not some random model that works well with this or that distro, which is, let's face is, going to disappear in a year anyway and be forked and spooned three times before that.
I will buy whatever I feel like, and I expect software - in Year 2016 - to work majestically on that hardware. And if it does not, then it's a failure on whoever designed the software. It is as simple and brutal as that.
Software as a Failure (SaaF)
The word project when used to describe software is such a horrible crime against humanity. It completely misses the point of what and how things should be done - in the real world, where adults pay with real money and expect results.
You can laugh all you want at the would-be morons who use Windows 10 or iOS or whatever, but the simple reality is, those are infinitely better products than any one Linux distribution out there. You can actually rely on Microsoft and Apple to provide you with a stable, consistent and supported platform for work. You can hate them. You can despise them. You can ridicule them. You can attack their policies and privacy pogroms and whatnot. But you cannot deny the fact they let you get your shit done.
The Linux desktop. Does. Not. Let. You. Do. That.
I make money from Linux. I've been doing that for the better part of my career. On the server side, things work, there's an actual legally-binding owner to everything, and you get software support for a decade or so. SLES, RHEL, take your pick. This is the serious adult game, and this is how it's done. On the desktop side, it's a sad sad little story.
So yes, some of my reviews are bad, some are good, sometimes I'm moody, sometimes elated. I did declare a bunch of distros to be in a pretty good shape. But any one point is insufficient. You need multiple data points. You need to observe trends. Science!
Look at the last three years, and carefully observe the pattern of quality for the top 10 distributions. First, some of the top 10 are no longer there. Second, there's a drastic variation from one release to another. And that's the sum of all evil.
Note: Image credits, Derek J by Phil Watt, freeimages.com.
Windows XP through Windows 10 - I can do all my tasks exactly the same. Run the same software. Everything. EVERYTHING. There's not a single thing that I could do on my desktops in 2002 than I can't do in 2016. Hell, for that matter, when I tested the T42 system, which had then celebrated its tenth anniversary, I could still download Windows drivers for its old, aging card. Not in Linux.
So, just the last three years. Linux. Ubuntu, gone from good to very good to crap. Mint, gone from very good to awesome to crap. Fedora, gone from mediocre to good. OpenSUSE, gone from awful to awesome to meh. CentOS, the one distro holding steady like a rock.
You can play your music then you can't. You can sync your phones then you can't. Your battery life suddenly starts sucking for no good reason. And a million other little bullshit points. Systemd, Wayland. WHO CARES! I don't give a shit about terminology. I want my systems to work and provide me with stability, entertainment and possibly work productivity. That's all.
Do you know the one difference between incompetent people who use Windows versus those who use Linux? The latter feel entitled and opinionated enough to twitter about their superiority complex. You just don't hear the average folks running Windows complain, because they have better things to do, like watch Youtube. Since you can't do that in Linux, because things break all the time, you masochistically waste your energy fixing your system so it can do what any decent Windows 98 box could 9,000 years ago, and then defend your choice with the classic zeal of a Stockholm Syndrome casualty.
This is an unfair comment toward all the Linux people who do care, do work hard, and do try to make better products, but I am not really sure what's the best way, if any, to convey my message to people with reading ability challenges. This is not a call to improve Linux and make it better and blah blah. I've expressed myself enough times on that. You know what the magic formula is and how to make Linux desktop succeed. It is to listen to me, because I'm always right. So the one piece still left is to sit down and read. There. And please, don't link this article in your blogs or whatever. The first response will be: clickbait. Or: he has disabled comments on his site. Don't. No need. Just read. Carefully. That's all.