I think the main reason, by far, that more people don't use Linux on their personal computer is that you can't easily buy a computer with Linux pre-installed.
In my experience, something like 95% of all Linux problems are at install time--hardware support, drivers, necessary pacakges...etc. Once somebody has gone through the difficult bits at the very beginning, Linux systems tend to be very stable--after the first week or so, I've had less issues with my Linux computers than I have with Mac OS or Windows.
I set Ubuntu up for my roommate, who is rather inexperienced in computers, and he has had no issues. In fact, some things worked even better than on Windows/Mac--he was able to use my printer without installing any drivers; my roommate last year had to spend several hours and get somebody else to help him to print from his Mac.
Apart from the difficulty of installing, most people don't even realize that they can install a different operating system--they believe that whatever they have is a magical part of their computer. Or something. You'd think Mac people would be better--they are, after all, using the less popular platform--but they are, if anything, even worse in this regard.
Thus, I think the immediate and most important prerequisite to widespread Linux use is not Linux's quality (which is already superb) but its presence in retail outlets.
However, there are two big obstacles to this: Mirosoft and Apple. Microsoft has had a history of anti-competitive behavior in regards to Linux, and goes out of its way to strong arm OEMs into only supplying Windows; both companies ruthlessly market that Linux is not an option--the false Mac vs PC dichotomy is a perfect example.
However, even if Linux isn't particularly popular, it already delivers the most compelling desktop and laptop environment to those who bother installing it; in that sense, it is already the year of the Linux desktop--I can be more productive on Linux than my friends on Windows and Mac OS; if less people use it, well, it's just less competition ;)
Main reasons why I am not using Linux on desktop are as follows: the complexity and broken-ness of audio and video drivers and frameworks, poor state of unified user interface. Poor state of backwards compatibility on the user experience, not on software necessarily.
If everyone would have just stopped at KDE 3.5 and improved upon it I would have been happy.
As an ecosystem, there are too many options and ways of doing things. Yet, there is no single obviously good one.
Ubuntu ameliorates most of this.
But... There's still the huge problem of support. Flash works much better on Windows. Games are made for Windows, not for Linux. Interfacing with Microsoft Office environment works better on Windows.
Where Linux completely eviscerates Windows is on server and web development based on command line. New language and platform innovation.
>However, even if Linux isn't particularly popular, it already delivers the most compelling desktop and laptop environment to those who bother installing it; in that sense, it is already the year of the Linux desktop
This is simply incorrect. Only reason I would use Linux is for development and servers, Windows is still superior in all other aspects. There's hardly no mental overhead moving from Windows interface and software platform of 2000 to that of 2011. There is a huge mental overhead from moving between Xmonad, Window Maker, Gnome, KDE, various versions of Ubuntu, Gentoo, Debian, Arch, everything.
But... You do realize that if we count devices, Linux on desktop has already gone past Windows. It's called Android.
In my experience the audio and video issues vary greatly with hardware; for example, my old laptop had endless audio issues that I never fully fixed while my new laptop just worked. This is yet another issue that a computer sold with Linux would not have.
As for the unification of the interface, I do not see why it would be good for all versions of Linux to be the same. Linux is, after all, just the kernel. Within Unity or Gnome or KDE (4.7 is really awesome) the UI is consistent; you're only going to run into weird issues if you customize heavily.
Besides, you can also run XMonad on a Mac (and I know a couple of people who do); this does not mean Mac OS is inconsistent, does it?
The whole consistency thing, of course, is also something you worry about only when choosing your distro--if you didn't care and could just buy a computer with Ubunutu (or whatever) it would not be an issue at all.
Games are an issue on Linux; however, I think the first niche it would occupy would be cheap $300-600 laptops that people wouldn't use for games anyhow. On my old computer, flash just worked; on my new one, I've decided to go all out on free software and use Gnash instead which works for Youtube (but not for ads :)).
If you absolutely have to use Microsoft Office, you would obviously want to stick to Windows. However, most people can get by with LibreOffice/Google docs. Serious people, of course, use LaTeX and GnuPlot exclusively ;).
As for backwards compatibility, I agree that Linux UIs do change a lot. However, you are also underestimating how much Windows has changed (perhaps you are just more familiar with it?). My mother is not particularly computer literate, but could use XP fairly well. When she first saw Windows 7 she couldn't even launch a browser.
I really don't understand why you're constantly moving between all the different window managers--just because they're there doesn't mean you have to use them. These days I exist almost solely in KDE and am completely happy; admittedly, I'm probably a bad example because I was happy when I was moving between KDE,FluxBox, Gnome and some weird Sun DE for school.
Just because Linux gives you lots of options does not mean you have to use them, or even consider them--you're always free to exclusively use Ubunutu or Gnome or KDE or whatever. As I mentioned earlier, if you don't care and you had gotten a computer with Linux on it, you wouldn't have to make any choices--just keep whatever came on the computer.
Agree with the above - lack of pre-installs is a major, major barrier, no matter how good the software is. Clearly there have been a number of missed opportunities over the years, with HP, Dell, Acer, probably others. Yes, they each have done suboptimal things with their preinstalls, but my sense it isn't just big companies messing up. Something is lacking somewhere in the FLOSS ecosystem such that there hasn't been a single mass market manufacturer stick with it. I don't think it is the software (anymore at least), but not sure of anything beyond that.
(Obviously there's System76 and Zareason and even smaller companies; I wish them wild success and will eventually purchase from them when I stop getting hand-me-downs.)
...
The "top 10" list from Linux.com is pretty reasonable, but it's proof that Linux.com is not the home of real journalism -- the list isn't presented as an image gallery. :-/
The dearth of mainstream Linux laptops certainly has nothing to do with the quality of the software--it is simply a result of unholy market pressure from Microsoft coupled with a ton of marketing.
No big manufacturer is going to go all-out on Linux: too much risk and uncertainty; big corporations rarely endorse such risk-taking. Instead, they want to sell some Linux computers but still maintain Windows. This gives Microsoft a lot of leverage in the market.
A perfect example would be netbooks--a while back, there were Linux netbooks available and they were threatening to become successful. I think Microsoft first lowered the price of XP and extended support for it and then threatened and cajoled OEMs into stopping selling these netbooks. A cursory Google search revealed, among other things, this[1] article about Microsoft's tactics.
The other thing is that compared to the obscene amount of marketing behind Windows and Mac OS, Linux is driven almost entirely by word of mouth. Since there is no really heavy backing, Linux is at a disadvantage even ignoring vaguely anti-competitive practices and bullying on the part of the established players.
Edit: Here[2] is another article with more details that explains exactly what I was getting at about Linux netbooks.
I agree proprietary $$$ (marketing and otherwise) are a big part of the story. I'm reticent to say it is the entire story. That's very disempowering. Even if it is the entire story, the inability to successfully counter MSFT/AAPL/et al $$$ (not necessarily with more of the same of course) still points to some kind of gap in the FLOSS ecosystem.
The Open-PC is a PC for everyday use built by the Linux community for the Linux community. We use only free software. Your help will be welcomed, to make our project even better.
The main reason, by far, that I am not happy using Ubuntu on the desktop is that unity works for almost everything except the things it doesn't work for, notably libre office. If the number two most important app after the browser fails to behave like everything else, like wtf?
I will note that I have used tons of hard to use obscure window managers, but if I'm using the "usable" option, it damn well better be consistently usable.
Well, who cares, pretty much the only market where Linux lost was the desktop, which is the most boring, stale market to be in right now. Linux is everywhere from my company's cloud provider to my phone and my TV.
The issue is that some people take adoption as a proxy for quality and assume Linux is then obviously inferior. I'm just pointing out that Linux's adoption rate does not actually reflect on its quality--it is a product of a wildly imperfect market.
There are, of course, many other reasons why number of users is at best vaguely correlated to quality; I'm just showing that in the case of Linux it's even less about quality than in other products.
So the issue is that random Linux detractors care--the sort of people who try to convince all their friends away from Linux and towards the proprietary technology du jour. And, regardless of how "stale" the market may be, I would definitely like to see more Linux on laptops and desktops.
Right now, in the room I am, there are 3 linux machines: 2 phones, one tablet. In the rest of the house, there are other 7 machines running linux, only two of them laptops. There are 4 OSX machines and 5 iOS thingies.
There is one Windows box - a virtual machine on my work netbook - and it hasn't been booted in months.
That's one of the problems you face only because you're installing Linux yourself.
If you were a company that was going to sell a laptop with Linux already, you would not sell it with driver problems; you would either be sure to choose hardware that supported Linux or you would write your own driver. Thus, if Linux laptops were easily available, people would not have this sort of problem.
Except vendors have shipped preinstalled linux with driver issues (a high profile early netbook had broken wifi, for example).
It generally takes a lot more distro coordination to support Linux on particular hardware, but the root issue is that the PC OEMs just don't really care to put the effort in.
Also, welcome to Linux advocacy, if you look around I'm sure you can find these types of "Linux would be more popular if X" arguments going back for years.
Yes, if the OEM decides to supply poor hardware and worse support, the experience is bound to be sub-par. However, this still supports my fundamental thesis: Linux's popularity is wildly decoupled from its innate quality.
If an OEM shipped broken drivers with Windows it would be at least as inconvenient. In fact, on my last laptop, this is exactly what happened: the audio system under Windows did not support one of the headphone jacks (there were two for some odd and unknown reason). From a tinkerer's point of view, Linux even has an advantage here: if there is some way to improve the situation, it's much easier to find under Linux. I have literally spent days searching for 64bit drivers for an external wireless network adapter on Windows; finding the same for Linux was trivial.
And to me, of course, the most important bit is that once I get everything working--it's gotten easier with each new computer :)--Linux provides by far the best experience.
How are the Dells that come with Ubuntu, coincidentally? I haven't seen them in retail anywhere, but I have seen them for sale online. If the support there is good, it could be a nice stepping stone going forward.
It's not how popular the machine is, but how well supported its parts are. If you select carefully, you may well find linux-proof machines in every brand.
While the current version of Ubuntu is trashed by many, its usability is no longer an issue---as near as I can tell the argument is what it looks like with a few quibbles over changes to some standard utilities---the app installer used to allow you to check off the ones you wanted and then install all at once. No longer (or if it does, it is cleverly hidden!) I'd suggest if a user has problems with this particular flavor of Linux, then their chances of success with Win7 or Win8 seem unlikely.
Definitely. As I mentioned, I installed Ubuntu on my roommates laptop. He is completely inexperienced in computers, and not only has he had no problems with Ubuntu, but he has also become more productive than when he was on Windows 7.
(long time linux user, also Microsoft employee. I take credit for the joke, though :P)
[edit: I know there are games for linux. In fact, I applaud companies that go out of their way to make linux versions of their games. e.g. Unreal Tournament 2k4]
Hah, you do have a point though--even if there are great games for Linux, most of the popular ones (at least for my demographic ;) are Windows-only, and gaming is as much about playing with friends as it is about the quality of the games.
However, my roommate, like a bunch of my other friends, is entirely an XBox person, so the games aren't the issue.
While you have a point - many games are Windows only - I think it has more to do with focusing on the work rather than on the tool. Windows is intrusive. Ubuntu, much like OSX allow me to just let it flow.
On my case, Ubuntu has an edge because it resembles my servers more closely.
OpenStack might be really important conceptually, and it might have an impressive number of backers, but they really, really need to get something usable in production environments out, and soon.
At the moment I think it is hard to justify it being the 5th most important..
Kind of a long way to go to slam Android. :-) (not being snarky per se, but would have preferred the author separate the last bit into a different blog post 'Why Android didn't make the top 10 list'
I pretty much agree with most of what is on this list. I might add Asterisk or Gcc. Not sure what I would delete though.
I do agree with what he said, Android isn't as open as it seems. Google didn't release the Honeycomb code because it "wasn't ready", yet allowed manufacturers to ship tablets with a version of it on there. And he also made a point when he said that Android is more "source open". Nobody can really influence the code unless they work for the company or a company close to Google.
And Gcc isn't really a "new" software so it's not for this list, but it's definitely an important one.
This is not true. I know people that have contributed code to Android that don't work for Google (etc.) and that have their code on released devices now. Sure, someone may have already fixed the bug by the time you get to see the source code, but if the bug is obscure enough, your code is as good as anyone else's. (Also, features are probably always welcome.)
I'd argue that that is the big omission. Wherever you come down on Android itself, Cyanogenmod clearly meets the author's criteria for openness, is a technical standout in a new area and it has had a very big 2011.
Oye. Let's not start the "X Most Y of 2011" stuff. Every year we are bombarded by these badly written articles. Can we just skip this nonsense this year?
Nothing had an impact in the course of 2011 like Android... now a cheap 100$ phones (often provided to customers for free by phone companies) can do 99% of what an iPhone 4s can do, and this is changing the world.
