Submitted by In an interesting video interview, Canonical founder Mark Shuttleworth shares his thoughts on desktop Linux. Some of his most prominent statements include: “I think the bigger challenge has been that we haven’t invented anything in the Linux that was like deeply, powerfully ahead of its time” and, “if in the free software community we only allow ourselves to talk about things that look like something that already exists, then we’re sort of defining ourselves as a series of forks and fragmentations.”
The funny thing is, the lack of wholesale change and innovation is what draws me to Linux on the desktop. I like the paradigm as it is, and change for change’s sake is the reason I left Windows in the first place. The whole control panel vs settings debarcle for example.
Granted, I’m not your typical user though much like everyone else on this site.
“The funny thing is, the lack of wholesale change and innovation is what draws me to Linux on the desktop. I like the paradigm as it is, and change for change’s sake is the reason I left Windows in the first place. The whole control panel vs settings debarcle for example.”
Agree. And it’s one of the major reasons why Systems like Amiga,Atari ST, BE, Vista, Windows 8 ect basically failed. A lot of the changes in these systems weren’t really changes people demanded, they were changes for the sake of change like the mess we see in Gnome 3
I beg to differ about the Atari ST : the TOS (the rom based OS) was quite buggy and evolved really slowly (hardly any update possible without changing chips and/or running ram resident patches). The overall evolution of TOS was quite good imho. The VDI virtual devices allowed easy graphic scaling allowing screen as well as printer driver with little effort. The AES window management was quite good too. What was lacking though was multitasking, what the Amiga Workbench had. Then Atari went bankrupt.
It really just boils down to one simple thing. Package managers mean death. They’re a solution to problems that don’t need to exist. Add a decent sdk like what you would get with Mac, Android, etc. and suddenly you’d have a competitive product. The next problem is the inevitable reactions to this post, the community is 100% incapable of handling constructive criticism or admitting design decisions that don’t win on technical merit win on usability for end users. The only way desktop Linux could become popular at this point is if someone removed “Linux” from all the code and copied what Android, IOS, and MAC have done to become successful. It’s particularly sad such great examples of how to make Linux into a successful product exist, but instead the play store is compared to the monstrosity that is Linux Package managers as the same thing.
Are you arguing for an easier interface to discover/install packages? Because they exist, and have existed for over a decade. Ubuntu with the ppa also provided a very simple solution to installing third party software.
Also, what is this magical ‘Linux’ you want to remove? Are you talking about the unix tool chain? If so, I have to inform you that Mac OS has a unix-like tool chain as does android, the major difference being the very polished GUI and the larger installed base (On mac applications are installed in a dedicated program folder, but since we do the same on linux with /opt/ and /srv/ it’s horses for courses).
I feel the real issue with desktop linux is that it doesn’t provide a compelling use case for a majority of people to jump ship. We have good software, but so does Windows and Mac, and the software base is larger for them even if you can get 95% of stuff done on Linux just fine. Add to that the installation hurdle (which for new users is a hurdle) and people will just stick with what they know.
Where you are right is that Android solved this. They added a good mobile GUI to linux at a time when other options in that space were lacking (everyone had just been winded by the iOS debut upending the market). Adopting in that case was a no brainer for suppliers, and the public didn’t need to jump through any hoops. In that sense, desktop linux occupies the same niche as homebrew android roms.
PS: just because people don’t agree with you doesn’t mean they are “100% incapable of handling constructive criticism”.
You actually sound pretty upset by his criticism lmao
The bit that get’s on my nerves is when someone states an opinion and then goes on to state that disagreeing with said opinion entails a certain moral failure on the part of the other party.
There is plenty ‘wrong’ with desktop linux, I just don’t think the notion of package managers is one of them, especially as Android, iOS and Mac have a similar system with the app stores (as dark2 stated, we’re talking features here and not technical implementation)
“especially as Android, iOS and Mac have a similar system.”
Except when someone develops for those, they pack everything they need in the app. The Linux package manager is more like someone coming into your house, breaking off a table leg, and telling you the table leg was a dependency they need to manage for you. What they do is so entirely different that they should never be compared.
Did you noticed that you are using Strawman arguments to answer dark2?.
I think he is arguing about having ONE PACKAGE manager that will introduce zero effort with a good API to common things(network, vulkan/opengl graphics, access protection to devices, etc), not only a frigging GTK/Qt GUI, or those bloated apps like Snappy/Flatpaks that can’t even handle theme integration and uses a lot of disk space.
And, If you install a “homebrew” Android Rom, and install GAPPS, you will have the full “Google Android” experience while installing apps so, they are not equivalent to Linux distributions. That’s not the same when you switch distros. Different package versions and availability, different release schemes, etc.
I use Linux on desktop daily, and one of the things that got me amused lately is the Proton initiative from Steam. It’s layer at a simplified level that let me play almost all Windows games, without having to tweak Wine or Lutris. This is the kind of experience that the user could be missing with the rest of the Linux desktop environment. I had to do some tweaks in the past, but they are getting better and after each release it’s becoming more no-brainer for game setup and play.
You are dealing with dark2 opition the same way the community in a broader way deals with criticism.
> I think he is arguing about having ONE PACKAGE manager that will introduce zero effort with a good API to common things(network, vulkan/opengl graphics, access protection to devices, etc), not only a frigging GTK/Qt GUI, or those bloated apps like Snappy/Flatpaks that can’t even handle theme integration and uses a lot of disk space.
You have no idea what you’re talking about.
dark2’s thesis is enunciated in one line:
> It really just boils down to one simple thing. Package managers mean death.
I pointed out that package managers don’t have to mean death, and that the idea of a central application repository is now widely adopted. No strawmen needed.
He then goes on to state that:
> The only way desktop Linux could become popular at this point is if someone removed “Linux” from all the code
Linux is an ambiguous term here and could reference anything from the entire software stack to just the GUI. This point requires clarification, hence the comment.
To get the ‘full android experience’ requires installing play services on top of your homebrew, at that point you are out of the realm of FOSS solutions, hence the comparison.
Regarding the idea that one package manager will somehow solve this issue, the ‘One Package manager’ is Snappy or Flatpacks. These applications are heavy on disk space to abstract away all the dependencies of the software (dependencies that are handled via standard package manager otherwise)
Regarding my treatment of dark2, I challenged his assertions, and backed my statements up with examples. If this is how the community handles criticism, then good. OTOH I did let his jab at people who disagree with him being thin skinned and incompetent get to me, as I don’t like ad hominims when we could be debating constructively, so mea culpa on that one.
https://stopthemingmy.app/
You need to read this Risthel For GTK the hard reality theming does not properly exist.
Flatpak has a stack of GTK themes it supports because random themes in fact equal broken applications.
“Snappy/Flatpaks that can’t even handle theme integration”
So with the serous fact that the theme system is basically totally busted should they really attempt for perfect theme integration??
https://www.linuxuprising.com/2018/05/how-to-get-flatpak-apps-to-use-correct.html
There are current 19 approved themes and if you count variations 58 of them with flatpak.
Line in sand has to be drawn somewhere on these theming problems. Yes changing to a non windows provided theme on Windows can also break applications. Lot of ways we do need per application theming. Yes include a global default option but have option of setting individual application themes and means for application to say hey user if you have issue I was only tested with X themes please change to one of these before reporting bugs.
These theming problems are not restricted to flatpak/snap… package management. This is a generic Linux Desktop wide problem. Heck windows has this problem as well. Themes are a nice idea having them work is another problem.
I wasn’t really planning on responding to these inevitable replies, but as for “what is this magical ‘Linux’ you want to remove?” I mean strip out every mention of “Linux” and any copyright from it, and then put a new name on it with all the fixes. Even if I could fix it all tomorrow, I would still need to rename it due to the horrible experience just about everyone who’s ever tried it has had.
“I wasn’t really planning on responding to these inevitable replies, but as for “what is this magical ‘Linux’ you want to remove?” I mean strip out every mention of “Linux” and any copyright from it, and then put a new name on it with all the fixes. Even if I could fix it all tomorrow, I would still need to rename it due to the horrible experience just about everyone who’s ever tried it has had.”
Get a clue, dude. Real Linux users don’t care about losers like you and the lame garbage losers like you want to inflict upon it.
Go create the os of your dreams that nobody wants to use and leave us the hell alone.
It should be pretty obvious I don’t have 100 of my self and the millions of dollars to make that happen. And yes, you’ve proven my earlier point that you can’t take constructive criticism at all. The system could be better, but let’s attack anyone that points out the problems instead of improving it is the path you’ve chosen with your life.
https://blogs.gnome.org/mcatanzaro/2019/05/24/dear-ubuntu-please-stop-packaging-epiphany-if-you-wont-do-it-properly/
I’d also like to double down on idea that Linux needs an official SDK. Something like this would never happen if there were no middlemen repackagers. Package managers are an inefficient a broken system with endless wasted resources on redoing the same work for each distro. An official SDK that includes everything you might ever need would end this and be a godsend for end users, but that would require basically throwing out everything built on top of the Linux kernel so far and starting over from scratch (which is totally necessary as even if I switched to a different open source OS, they’ve just ported all the existing software and package management problems to it as it’s just another Unix one where they could take the easy way out)
My 2c:
Whenever you have a problem with linux you are told that you use the wrong distribution.
It doesn’t matter what the problem is or what distribution you use.
The answer is always “Wrong distribution!”.
This leads to the conclusion that there simply is no “right” distribution.
So why even bother with linux?
Honestly, sorry you’ve had that experience.
I feel that you could probably get any distribution to do anything with enough elbow grease, that said most distributions are nich fillers, and as such there will probably never be one distribution to rule them all.
Out of curiosity, what issues were you dealing with?
“Out of curiosity, what issues were you dealing with?”
Not working mice, keyboards, soundcards, GPUs…
smashIt,
Surprised you didn’t mention network cards, modems, printers, scanners, and raid controllers, haha.
I’ve had all the same problems at some point or other, however sometimes people complaining about linux are guilty of taking any random (windows branded) hardware they’ve got lying around and expecting linux to support it out of the box. If you want linux compatibility, then you ought to buy from a manufacturer that explicitly offers linux compatibility. I fully appreciate that most people do not have linux certified hardware and instead experience linux on arbitrary hardware, heck I’m guilty of the same thing, but the truth is that’s naturally going to be hit and miss.
I *wish* there was a linux distro that installed on everything perfectly out of the box, but that’s an impossibly high standard for any operating system. Even windows falls short and requires fidgeting with drivers sometimes. Ever had to use windows prior to proper GPU & network drivers? I have, it’s not good. My newest keyboard and mouse have spare buttons that I’ve tried in vein to get working on windows using the manufacturer’s drivers, yet they work just fine in linux, shrug. Any operating system can be frustrating at times, we need to have perspective though, if you want a linux computer but you don’t want to mess around with unsupported hardware, then maybe consider buying a pre-built system that supports it explicitly (just as one does with windows).
I know that there is a ton of obscure hardware out there.
But mice and keyboards are not obscure.
They are as basic as you can get.
For the GPU.problem:
I don’t blame them for not supporting every GPU out of the box.
But not even an intel GPU in a thinkpad?
And then the official guide to get it running tells you the complete opposite of what you should have done.
That’s the developer just “not giving a fuck”.
smashIt,
I get what you are saying, but it may not have been the mouse or keyboard at all, sometimes the system bios is actually at fault. One of the more frustrating aspects of operating system development is that sometimes your code is technically correct but won’t work without adding some system specific quirks/workarounds. Sometimes windows users are shielded from these with windows drivers that work around the bugs, but the bugs are still there technically. Personally I’ve had to work around ACPI bugs in linux.
Obviously I can’t say what the problem was in your case, and I don’t mean to downplay your frustrations, I’ve experienced plenty it too! However our instincts of blaming the OS might be turn out to be incorrect when all the facts are laid out.
It reminds me of this where Thom had similar problems with intel GPUs.
https://www.osnews.com/story/29406/microsoft-isnt-forcing-lenovo-to-block-free-operating-systems/
I still stand by the recommendation to use linux on a purpose built linux computer as any incompatibilities should already be sorted out. Unfortunately linux computers are niche products and priced accordingly, but if you want something that just works, that’s undoubtedly the best option.
“I feel that you could probably get any distribution to do anything with enough elbow grease”
And therein lies the problem. Instead of listening to the problems, finding root causes, and then pushing back on the community for actual long term fixes; you’re going to try telling people to just keep doing things the hard way and offer advice on how to work around every problem that doesn’t exist if I just use some other OS. Hence the desktop Linux experience is more like death by a thousand cuts.
We shouldn’t be so naive to think that the so-called free market works in a way that those with better, more innovative products gain a greater market share. Unfortunately, this is a second-order effect: could be relevant but usually it doesn’t play a role. The prime forces that determine market winners and influence the market dynamics are those which we usually turn heads away from, like access to capital, political power and lobbying, personal relationships and acquaintances, historical development of a market niche etc. We ignore those because we want to believe in the American dream: those hard-working and capable will be eventually awarded. Georg Carlin said for it “you have to be asleep to believe it”. The Linux ecosystem due to its collective nature is unable to employ for its own advantage many of these forces, although the ecosystem technically, on a paper, offers a lot compared to its competitors, especially if you take into account its freedom and price. Mark Shuttleworth’s wealth only was certainly not enough to beat bigger players in this game.
adundovi,
This! We often fight over package managers, desktop GUIs, distros or whatever the debate of the day is, but in terms of the ordinary consumers who will ultimately determine the long term popularity of a product, these things will have very little influence as long as it’s “good enough”. Superior marketing, brand recognition, and captive audiences will win every time even with an inferior product. Most of our energy in the FOSS community focuses on things that make us happy but do not have much potential to significantly increase our standings in the market. Many of us resent corporate influence as it compromises on our values: design integrity, privacy, vendor locking, against planned obsolescence, etc. Yet ironically standing up for our values often decreases our own spheres of influence since the money and power will gravitate elsewhere.
Mark said something interesting in the interview about making linux stronger by putting someone legally in charge to stop fragmentation and enforce direction, but then it would cease to be linux (paraphrasing here).
Pretty disingenuous, as Unity wasn’t why people really got up in arms over Canonical. It was Mir. By muddying the waters and reinventing the wheel from scratch with Canonical favouring licencing, Mir threw an unneccessary wrench in the progress of Wayland.
Canonical had pledged support for Wayland and at that time there was a clear track forward. X would be superceded by Wayland. Clean, neat and certain.
Suddenly Canonical does a 180 and introduces Mir. Homegrown behind closed doors, it was to give Canonical a leg up over the competition. At least that is what they thought. Never mind the FUD they spread about Wayland. Chaos ensued. Nvidia anounced only to support X. Intel initially wouldn’t support Mir. Developers weren’t certain if they would have to write and compile their software for three display systems. It created a nasty divide.
In the end it was all for nought. The whole convergence story from Canonical didn’t get full community support because of very ruffled feathers. Mir was chronically late and dragged Unity down with it. In the end Canonical pulled the plug as they lost momentum.
Now Unity and Mir are refactored into a Wayland compositor and Wayland Desktop Environment. Showing the whole debacle was pretty much unnecessary. Who knows what would have happened if Unity was fitted on a Mir Wayland server from the start. I might have been typing this on an Ubuntu Edge phone on a converged workspace.
This is really one of the reasons why Desktop Linux has had less adoption. Not that there are different package managers, or fragmentation in that regard, but that for whatever reasons, Ubuntu keeps trying to steer away from being with the rest of the distributions. They should stick to their 6 month release schedule, with latest gnome based on Debian Sid. Instead of putting so much effort into some fork of Gnome, actually try to improve Gnome. Try to improve Wayland, create a united front for potential developers.
From the video Mark seems to only refer to their efforts with Ubuntu Phone as ‘failing’ so I find the headline misleading. Yes he does point out the community’s reaction to Unity but that’s just one flavor of desktop environment, and I don’t think Unity’s success (perceived or actual) defines the success of Linux as a desktop OS.
“From the video Mark seems to only refer to their efforts with Ubuntu Phone as ‘failing’”
Of course Ubuntu Phone was going to fail. If Microsoft couldn’t con people into buying/supporting a phone and other products they didn’t want, just how in hell was Ubuntu going to do the same? By having asses like dark2 run around telling people “I know better than you what you want?
As Willam Shatner said all those years ago on SLN-Get A Life.
Did I say Ubuntu Phone wasn’t going to fail?
you cant walk into to the compu-hut and walk out with a Linux based computer. most people do not install there own operating systems.
‘Desktop Linux’ is doing just fine. The problem is and always has been a lack of quality applications. There are also too many of either poorly designed or crappy applications in the repositories.
There is no single or `main` reason why the Linux desktop hasn’t become a real competitor or hasn’t been able to poach more than the tiny sliver of market share its always had. Blame Microsoft, lack of quality apps, poor marketing, fragmentation, instability, being “unpolished”, lacking or missing drivers, high-end graphics, audio, and whatever else. Point your finger and make all the excuses you like, nothing is going to change for the foreseeable future because the Linux community lacks the ability to acknowledge these (and more) are all areas that need improvement, and it lacks the will & resources it would take to actually tackle them. After decades, a large chunk of the Linux community still hasn’t figured out that its greatest strengths are also its greatest weaknesses, and for that reason it will remain stuck in its sandbox – at least in the desktop space.
“2019-05-25 5:42 pm
dark2
It should be pretty obvious I don’t have 100 of my self and the millions of dollars to make that happen. And yes, you’ve proven my earlier point that you can’t take constructive criticism at all. The system could be better, but let’s attack anyone that points out the problems instead of improving it is the path you’ve chosen with your life.”
You and those like you are just basically Comp SCI Leeches who think they can con people into doing the crap they espouse and then later run around claiming “credit” for working on a project they had no real involvement with.. I’ve seen tons of asses like you from my days in school.
It seems you don’t understand the scope of writing and maintaining an entire SDK for an operating system. (And are back to personal attacks as you can’t win the argument.) I have an assignment for you. Watch a bunch of episodes of Gordon Ramsey’s “Kitchen Nightmares.” Pay particular attention to the owners or managers that want Ramsey to tell the public their food is good, and then Ramsey points out tons of obvious problems that they refuse to listen to. Those owners are essentially how the desktop Linux community behave and react to criticism. Just keep watching until you finally notice the similar things you and your fellow open source friends do.
And no, I have no plan to take credit. I don’t even program for a living anymore. Frankly desktop Linux has proven to me that commercial software development is the only way for large projects to not be garbage.