The second point: "Companies tend to depend on a myriad of applications to run their business, and just a couple of them not running under Linux would be enough to derail a transition to Linux desktops" is the real killer.
This is really the only reason that I can't recommend trying Ubuntu to more people. People site usability concerns etc and while there are some issues there I think it's mostly "good enough" now and we're long past the days of having to compile a .tar.gz full of .c files and fettle with vi in order to get sound to work.
So the issue is how to get third party developers interested. I think the best way is by including a really sexy app store. Ubuntu Software Centre is a start but it's still nowhere close to what Apple has achieved in this area. Nasty looking icons, inconsistent screen shots (some showing gnome2, others Unity) and thousands of free apps with weird names don't make it the most attractive place to shop.
In many ways though , I would consider desktop Linux a success regardless of marketshare for the simple reason that it is now possible to "use a computer in freedom". I think the software world would be a bleaker place if Torvalds , Stallman et all hadn't spent the hours pushing code. Imagine a world where the cheapest HTTP server license went into thousands of $.
>> In many ways though , I would consider desktop Linux a success regardless of marketshare for the simple reason that it is now possible to "use a computer in freedom". I think the software world would be a bleaker place if Torvalds , Stallman et all hadn't spent the hours pushing code. Imagine a world where the cheapest HTTP server license went into thousands of $
I completely, wholeheartedly, agree with this. I will always maintain at least one Linux machine for this very reason.
>> People site usability concerns etc and while there are some issues there I think it's mostly "good enough" now and we're long past the days of having to compile a .tar.gz full of .c files and fettle with vi in order to get sound to work.
IMHO usability is still a concern. I truly believe it scares off quite a few potential users.
Anecdote : I've watched as a room full of professional developers cursed at their screen while working past Unity, dealing with NVIDIA driver compatibility and playing around with Xorgs. All the while, mind you, the devs using OSX would occasionally chime in with casual snark such as "oh, installing tomcat on OSX is really just a drag and drop..."
Obviously, we all got our desktops working - and I'm actually very fond of my highly tailored Xubuntu setup. However, it wasn't exactly pain free. I have to think that many a casual user might be put off by such an experience.
> All the while, mind you, the devs using OSX would occasionally chime in with casual snark such as "oh, installing tomcat on OSX is really just a drag and drop..."
Dev using OS X here. I've been on OS X since 2003 and haven't looked back. I'm now doing iOS development.
Unless it's kidding between good friends, I have no need for such snark. (Though I find it useful as a symptom.)
Part of the problem is supporting more than one distro is not as easy as people think, even different versions of the same distro are a challenge.
Compare to Windows and Mac OS where they really take good care of developers in this respect.
I'm not saying its impossible just that's it's hard enough to make developers think twice, slow them down or target only one distro.
I think the best way is by including a really sexy app store.
This is most likely necessary, but if this were sufficient, we'd all be using Lindows. Naturally, you can adjust the definition of "sexy" to include Apple's app stores, while excluding everyone else, but that doesn't strike me as a useful definition of the word.
The only thing that was keeping my parents and some of my friends was that they needed Microsoft Office for their work/schoolwork. As of now MS Office runs like a piece of cake on Linux -- with PlayOnLinux (http://playonlinux.com/). Honestly, it's quite impressive how smoothly it runs (and how easy it is to install it.)
MS Office isn't the only application that PlayOnLinux supports - there's a ton of games and other software (Photoshop, Blender, Dreamweaver, Flash, etc.) that it supports. To top all that off, I feel like the desktop on has gotten better and better lately. I use KDE 4.9, and I will say it is quite nice. The level of integration KDE offers and the high quality of many of the standard apps that come with it will make a Windows user never turn back. Ubuntu too has a rather simple and straightforward UI (although it doesn't personally appeal to my taste).
Really though, the only real reason desktop Linux never caught on was the Windows default. That is it. If brick and mortar PC realtors sold Ubuntu laptops and desktops next to Windows PCs with the same specs and a price cut of the Windows licensing costs the MS monopoly would have ended a long time ago.
Wow. Office is literally the only thing that's preventing me from switching to linux. I tried Wine (a year or two ago) and came off less than impressed. I'm going to try PlayOnLinux now and if it's even "good enough" I'm making the switch.
The challenge of Desktop Linux is that it's not a product, it's a project.
Ubuntu is a very successful desktop Linux distro. It's pleasant to use and very modern. Nerds might hate it because Unity doesn't fall in line with Linux "the project" so much as its there to make Ubuntu "the product" better.
Overall, desktop linux as an overarching product failed, but so did mobile linux pre-android, but Android isn't so much mobile linux as it is Android.
Open source is a bit like herding cats and if you don't have a real product you are trying to ship, devs will scratch their own itch.
I agree with everything except for the past tense "failed". I would use the (pause while I look this up) present perfect. Linux has failed on the desktop so far, but with influence from Android platforms and the availability of cross-platform dev tools like Java, Qt, Tk, and crap-in-browsers, it may yet succeed. I work on a desktop application for Windows and OSX and at least 80% of my development I can do on my Linux machine.
EDIT: meant to stick this in - really to capture the desktop market, Linux only needs to emulate or simulate WinXP functionality. For most users, Win7 only adds gimcrackery.
> at least 80% of my development I can do on my Linux machine.
That's the whole problem right there. The remaining 20% are all the desktop-integration features and UI polish that is not cross-platform, but which makes the difference between a run-o'-the mill app and a great one. And OSX does raise the bar quite high for the latter.
The desktop is being left behind by mobile. Recognize that Android/GNU/Linux has won mobile. Despite an early lead, Microsoft has been crushed worldwide in mobile by a Free Software platform.
Android provides a platform where both Free libraries and closed source apps proliferate - and are very inexpensive. All the failures of desktop GNU/Linux have been solved, or are not relevant, in mobile GNU/Linux.
Regarding the tablet segment of mobile, one could argue that GNU/Linux will seize the low end and gradually gain market share at the expense of iPad, leaving no room for Microsoft.
The only reason I care about desktop OS at all is to develop for mobile or back-end server.
Android is specifically not "GNU/Linux". It's a Linux kernel with an almost entirely custom user land, from the C library up. Sure, some of it is Open Source, some of it is even non-Google-created Open Source, but it's knowingly and specifically not "GNU" and/or GPL.
Adding insult to injury, its likely that once-dominant Microsoft will not only be crushed by Android/GNU/Linux in mobile, but also by other upcoming mobile ecosystems built on GNU/Linux - Amazon's Android fork and the Tizen mobile OS.
As of June, Google said that Android/GNU/Linux was on 400 million devices and that each day 1 million new devices were activated by users worldwide.
Clearly, no one is worrying about why Linux failed on mobile!
I don't believe that. I think mobile augments people's computing, but I haven't seen any data to indicate that people are actually replacing desktop with mobile.
I concede that Google locks up the code for upcoming releases. No need really to tell you that most mature Free Software projects encourage bug reports and patch submissions, but limit source code commit privileges in order to maintain stability.
Evidently its more useful to develop/maintain a Free Java library for app developers. Or submit/maintain a kernel patch for a mobile device driver. Understand that Google does not control GNU/Linux. And even its control of Android is not enough to prevent forks.
UI/API tinkering and fragmentation hindered desktop GNU/Linux adoption. Why go there unless you want to fork?
That gives me source for old, outdated versions of Android, not what I'm looking for. You can get the source for old idTech as well, does that make idTech open source?
> Can't, only Google employees have access to the current version of Android.
Wrong. Google employees have access to source of binaries you don't have access to. If you happen to have the binaries of an unreleased version of Android, you are entitled to the source. At least, all of the parts that are under GPL licenses. Google may or may not release sources for other versions.
> But there is a process for getting commit privileges that doesn't involve job interviews.
Being able to contribute to the upstream is not what defines free software. Being able to fork released software and work on it is. Android passes this test. Also, accepting patches from third parties could open Android to all sorts of litigation. What if a former Microsoft employee contributes leap year handling code from their products? The fact Android phones will crash on every February 29th will be the least of Google's problems.
> That gives me source for old, outdated versions of Android
No. It gives you access to the sources of the Android versions publicly available. Being open source doesn't mean you get access to pre-release versions or to every single commit to the codebase. It means you can have the source code for the binary you have.
A few crucial differences,unlike the other projects you mentioned, Android development happens behind closed doors, no outside patches are ever accepted, and an OEM is specially picked for exclusive early access to a new release and the others and CM are left scrambling to update their devices if they fork too much. Not to mention, Google apps and the marketplace is neither free or Free.
Even Windows betas are distributed to all OEMs equally with no special treatment.
> unlike the other projects you mentioned, Android development happens behind closed doors,
That doesn't make it any less free than any other project. This is a governance issue, nothing more. If you get an Android handset you're able to download all parts of Android that went into it.
Software freedom is not about getting access to unreleased software.
>Android provides a platform where both Free libraries and closed source apps proliferate - and are very inexpensive. All the failures of desktop GNU/Linux have been solved, or are not relevant, in mobile GNU/Linux.
Well, mobile is not for productivity, or even for typing a long email. How many people you know that haven't got a laptop or desktop, except grandmas? What kind of programs do typical people use on millions of Android phones?
>Regarding the tablet segment of mobile, one could argue that GNU/Linux will seize the low end and gradually gain market share at the expense of iPad, leaving no room for Microsoft.
What? Android tablets are not selling very well and are getting crushed by the iPad. Meanwhile, Surface seems to be the only device ready to take on the iPad by the horns, the rest are cheap imitations of the iPad but really cheap since Apple has a tight supply chain. Not to mention that a 7" iPad will take away some of the low end market you're talking about. So that's a rather silly argument to make.
>The only reason I care about desktop OS at all is to develop for mobile or back-end server.
What about laptops? Are they "mobile"? You're pretty much ignoring them.
In my crusade of boycotting apple production I switched from OS X to ubuntu. And you know what, despite high disregard from HN auditory, I've found Unity to be surprisingly good. For example, unity dock still worse than os x dock, but it's good enough. Single menu for all apps is familiar from os x, although I hit it when attempting to move windows more often than I want to. Application switcher can be navigated with arrows by default, which is good, although I would also like it to be navigatable via mouse too. 'Spotlight-like' menu named Dash Home takes way too much screen estate, but again, it's at least usable. Hotkeys are terrible, though, and first thing I did was disable alt and super keys calling dock/dash-home.
Overall, I like Unity way more than current Gnome, KDE, XFCE and LXDE. Your mileage may vary.
I think, the real linux desktop problems are when something goes wrong. Sometimes updates are unsafe. Sometimes you find a bug in a software. I had 10 or 20 crashes and error report windows in my first day. Commercial software is terrible too. Skype is buggy, crashes often and is just bad. Nvidia binary drivers suck and nouveau crashes on my card (560ti). Twinview can only VSync one screen. Your other screen is doomed to lag on renders. Xinerama has a bug with cursor randomly jumping over to another screen. Whenever something bad happens you resort to google and waste 10 minutes+ for fixing it.
I also think that applications not being made for linux is not a very big deal. 80% use case includes browser, music player and office package. All of which are included by default in most distributions.
I have similar issues myself. It always seems that there are glitches all over the place. I really wish the desktop was as stable as windows (or osx). Don't get me wrong - I like KDE, GNOME, and Unity ... the problem is that they are not that stable. There always seems something that is off.
I use desktop linux quite a bit (though only through a VM these days). What's interesting is many of the applications that always seemed to be missing or worse than their Windows competitors have been absolutely destroyed with web-based software. Both Windows and desktop Linux have been losing that battle.
The funny thing is that desktop Linux apps have always been trying to match Windows apps feature-to-feature, but web developers haven't. Turns out I didn't need every feature from Excel, I needed something faster, more convenient, and easier to use.
Speaking as someone who uses plenty of the more obscure features of Excel, I hate the "lets do everything in our browsers!" paradigm.
That being said, what really matters for the question "why isn't linux successful" is only: "How easy is it to get the software that does what the user wants?" For an average home user, I think Linux could have been ahead of the curve. All a distro had to do was slap a pretty GUI on their package management system and they would have had an app store where everything was free and easily accessible. The free office replacements are more than sufficient for home users, as are the chat clients, web browsers, and media players.
I agree you've got to make it easy for people to find and install software.
It's the same for all OS users though. I know enough Apple and Windows users that have never installed any software on their machines whatsoever.
My brother called last week asking how he could share some digital photos with someone, he had over 100MB worth of photos. I suggested he needed to resize them and possibly archive them. He was absolutely clueless.
People just want to be able to do something with the least bit of fuss. The goal surely is to make it easier for people to carry out said somethings.
Package management could and should be better than it is. I don't believe that an app store is the answer to these issues. I've yet to try an app store that I like (haven't tried Apple's.) Normally it's tricky trying to look for software, and find solid recommendations.
Ubuntu's software center offers me little extra. I still resort to Synaptic!
If you can't install software easily, then you better offer a good set of defaults.
My argument is more that GUI software available for linux was not revolutionary in the way web apps have been, rather it was full of "lower-quality" windows applications. A better package manager wouldn't have solved that problem in my opinion.
Well the complaint is that mainstream Linux desktop usage has not been adopted. I would say Google docs is a vastly better solution for most people than OpenOffice. OO tried to copy MS Office feature-to-feature. Docs tried to just make something fast and easy to use. Docs won.
I don't think there needed to be so many bullet points to explain why the adoption of Linux as a desktop operating system has been extraordinarily slow and generally unsuccessful.
In my opinion, it would fall under these two reasons
1) The difference between a Linux desktop & a Windows/Mac desktop is negligible, or worse. GNOME/KDE don't really add any compelling features that make them better than Windows or Mac anymore. I remember a few years ago, I loved putting Ubuntu on my system because drivers would be downloaded automatically and I could easily access all Linux packages from one simple package manager. The folder browser was pretty familiar, and the GNOME 2 bar was a nice hybrid between Mac & Windows, but nothing too special.
However, the driver installation & central package directory have been long a competitive advantage with the era of Windows 7, etc. Granted, these weren't "defining" features of Linux, but when I personally used it, these were thinks that struck me then, but are no longer relevant now.
2) The ecosystem. I think it goes without saying that the Linux software ecosystem is much more fragmented, and is often found in the "underground".
It's not an issue of whether or not there are substitutes to things like Office, Adobe Creative Suite products, iPhoto, and other essential apps that normal people/working people use on a daily basis, etc. (although I do think that there aren't adequate substitutes for these and that the friction of trying to get these actual products to work through things like WINE, etc is too much).
However, because things are so much more fragmented on Linux, it being an open system, it's harder for there to be de facto software (unless you lurk in some Linux community, which again, normal people aren't generally interested in).
The user has to make so many choices, which are often arbitrary and needless, and in doing so becomes frustrated and confused. There is too much stuff to explain that isn't necessary to explain, and too much detail to go into that again, is not practical.
The trade-off with Linux is that you get an enormous amount of power and responsibility. The benefit of this is that you get an enormous amount of power & responsibility. The cost is...the same, some people just don't want to bother.
I see a much simpler explanation. Linux has, for a very long time, followed the path opened by (and thus replacing) technical Unix workstations. Those were specialized machines built for people who would put up with cryptic interfaces in order to run the powerful software they needed. As any specialized tool, it has appeal to the few professionals who have actual use for them.
Only recently (3 years?) Linux distros started being friendly to "mere mortals". This change coincided with the acceleration of the demise of the PC. There may never be a year of the Linux desktop.
Yet, Linux is everywhere. As soon as you fire up your network connection, you are using Linux. Every time I look up the time on my phone, I'm using Linux. Most internet-connected TVs run Linux, as do most set-top boxes and e-readers. I amuse myself thinking the convoluted things Steve Ballmer is compelled to do just to be able to claim he doesn't use it.
To me, 'Desktop' Linux has never worked because the lofty aim of projects like GNOME to secure 'everyday' users is completely at odds with many of the basic design decisions of the Linux/Unix platform that underpin it. Things like storing data in string-based-files, chaining small command-line-tools to do everything, a historically eccentric file structure, and pluralism in every-way (eg. custom built binaries on a per-machine basis) are never going to work for the 'everyday' user. You'd need to have an even bigger project than GNOME to change these design decisions (like redesigning the file structure and standardizing the hardware/software platform on which it runs, as OSX did).
That said, many of the components of GNOME (GTK, GStreamer, etc.) are great and we shouldn't forget their usefulness to other projects whilst the Gnome Desktop is coming up against these existential questions...
"Things like storing data in string-based-files, chaining small command-line-tools to do everything, a historically eccentric file structure, and pluralism in every-way (eg. custom built binaries on a per-machine basis) are never going to work for the 'everyday' user."
I'm fairly everyday user most days (LibreOffice, R via R-studio, a bit of graphics). I never see those things on Ubuntu (12.04, Unity on the laptop).
On Sundays I puggle about a bit with some python, and bodge up a few bash scripts, but that is out of interest not workflow.
Yeah, Ubuntu is much closer to what I was trying to say needs to happen to overcome these problems (as a whole distribution, not just a 'desktop environment' like GNOME), and theres a lot of hard work in there, but the issues mentioned will still cause some problems. For instance, difficulty in installing less-common software (see the giant page https://help.ubuntu.com/community/InstallingSoftware).
Gnome 3 completely borked my system of 8 years running on an Thinkpad R50e I was happy with. There really is no excuse for what happened. There was no cure either. I am done. Writing from a mac now.
I read that blog post and all I can do is shake my head. They seem to be fighting the same old monsters for the last 12 years. The list of reasons on the post sickens me.
What do you mean by borked? Did it actually cause damage somehow or did it just not run well? I wouldn't expect something so resource intensive to work well on that computer just as I would be amazed if you could smoothly run mountain lion on an 8 year old mac. It's actually a major strength of linux that you can run modern software versions on even extremely outdated hardware; you just need to choose an appropriate desktop environment.
Borked: Unable to boot into the GUI, unable to downgrade properly or go back to the earlier version of Gnome. Tried doing stuff manually, ran into a bunch of dependency issues, gave up. So then I loaded up XFCE and after a while of using it, saw a mac, liked it, and moved. Also I guess I had moved on from the time I was more of a hobbyist and just wanted a dependable system. I had backed up all my config files just in case as well.
XFCE continues to be the only usable UI for laptops IMHO. XFCE actually rescued said R50e, and yes, I am amazed by the fact that it is usable. But given that I had been using Gnome for years and years, I personally was not comfortable with XFCE. The mac gives me what I want, a unixy system and a stable GUI without the hoop jumping tricks.
Personally I don't see these points as being that important when looking at casual home users. I've recently setup a couple of older laptops with Ubuntu for some family members who just want to use them for watching movies and browsing the web ... and for that, it excels.
The GNOME project has failed because they have hacked at things that don't matter. For example, what the world wanted and needed was a decent word processor. It got OpenOffice. It needed an easy to use flow chart creator. It got dia. It needed project management software. It got... Heck, what did it get?!?
Cheese and Gnome Shell are all very nice, but so what?
The "Gnome Office" was Abiword+Gnumeric. Unfortunately, Abiword did not try to clone MS Office, but tried to make their own thing. That meant missing/bad support for opening doc files.
Gnome Planner is the official project management software. Not a serious solution, though.
His 1st bullet point says it all. Ken Thompson said the same thing about Linux in an interview years ago. He said that is Linux's major problem.
I find preoccupation with some company's metaphor to be a sign of lack of creativity. And the people behind Linux distributions are obsessed with Microsoft and the "desktop".
The "desktop" is only one metaphor.
Does iOS have a "desktop"?
To speed up Vista when it was first released, I used to disable the desktop in Windows by changing the registry key that specifies "explorer.exe". I would just boot to msconfig or task manager.
The system ran much faster that way. Applications can still be minimised. It worked so well, I never went back to the aero nonsense.
Obsession with a "desktop", and trying to look like Microsoft's version of it, is one of Linux's major flaws.
To all people complaining about unstable apps, poor performance and buggy drivers: I purchased my last laptop from Dell because I was aware it was well supported on Ubuntu.
This alone makes a huge difference, trying to install Mac OS X on a frankenmac will result in an troublesome experience too.
Doing your own transition to free software is also a key for success, and if you're not tied to a proprietary technology, you should try it. I've been a full-time Ubuntu user for 5+ years, working as a developer using Java, Ruby or Python, and I'm very satisfied. Even further, when I had to use Windows 7 or even Mac OS X it feels odd.
One of thing which I believe hindered in Linux’s popularity right from the early days, was its not so impressive User Interface, compared to Win95/98. I think this is where/when Windows took off and nobody was able to catch up comparably for a very long time to come.
Another aspect was the lack of availability of compatible software. Software back in the days were in the form of CDs-on-the-shelf, and I don’t think anybody made any of them Linux compatible.
So the challenge was just not there from the beginning to get a hold of average desktop user.
A dream-wish: The people who developed the Harmattan UI (Nokia N9) design something new and refreshing for Desktop Linux. I know, that's probably improbable, but if it did happen...
This was the first (and only) UI on top of linux that I really found beautiful. Android does not have the same polish, unfortunately..
They are perfectionists who are doing a good job with all their software. But they have been working on it for more than 18 months, but have not released yet. It is tough to make stable and consistently behaving software, but I can't imagine that they will get too much traction by delaying their release so much.
It definately is a very promising distro, but not right now.
I installed this distro a few weeks back and found out it was Ubuntu 10.10 running an older 2.6.x kernel.
To keep things short, it just didn't work very well on my i7 Sandy Bridge laptop with SSD. I had to tinker with some drivers to get Wifi working. I had a lot of graphical glitches that I could not get fixed. The OS also performed like it was running from a USB 2.0 dongle, while in fact it was running on a SSD. CPU constantly hitting 100% fan usage. Random lockups were frequent as well.
Altough I really liked the look and feel of this distro, I i was dissapointed and frustrated not being able to get it to run decently on modern hardware.
Afterwards I installed Pinguy OS built on Ubuntu 12.04 and that worked right out of the box.
I migth get back at Elementary OS when they release a stable distro built upon newer Ubuntu builds.
I'll agree it may not be a traditional GNU/Linux with an X-based GUI, but it certainly is as much Linux as, say, OSX is Unix (despite the lack of teletypes attached to most Macs).
If you don't think Microsoft is directly responsible for this, you are an absolute idiot.
Every manufacturer has to pay for Windows Mobile for every Android phone they sell. No company has stood up to Microsoft. This is a real threat.
You want to talk about Gnome 3? Fuck you. Why would anyone invest a cent in a WM if you can't distribute Linux installed on a laptop?
People like talking about Microsoft and Apple as though they are different teams. Nope, they are on the same team: fuck people who think they can get by without them.
I think there are better words you can use to make your point.
Also note that Microsoft failed to stop Android even though they have been bullying all Android manufacturers too with their patent portfolio. The reason Desktop Linux never became a success cannot simply be attributed to Microsoft's bullyings. To them, it's probably just another source of income, not necessarily a means to stop free software or Linux.
What? Desktop Linux never became a success because nobody profits from it being awesome. That can't happen until you can distribute Linux on a PC without making some back room deal with Microsoft.
I'm sorry, but this issue makes me very angry, and I feel like most people are childish fools who pretend that it's the fault of hobby hackers making software that's difficult to use. No, it's because no matter how awesome Linux is, you can't distribute it without paying.
Also: Microsoft may have not stopped Android, but that doesn't mean they haven't hampered it. Without that stupid license every phone would be able to be sold $8-$15 cheaper.
I do get your point. But unfortunately that's the way the world works. Those who have patents bully others and make money of them. Microsoft is playing by the rules. I still don't think there is any point in blaming them for the lack of success of Desktop Linux.
The reason that Linux PCs didn't take off is manifold. First, they're unfamiliar to the user so the return rate is high when they find after taking it home that it won't run some app they need. Second, the price is the same or even more because the OEMs get paid a pretty penny for the preinstalled adware and bloatware. And all the support and the separate software tooling make it tough for the OEM to turn a buck. Not to mention that they're wary of showing it prominently on their site because in the past that lead to a lot of uninformed people buying them and returning them.
>No, it's because no matter how awesome Linux is, you can't distribute it without paying.
You're needlessly getting angry for nothing. What's stopping anyone from selling PCs with Linux? You can start a company today that does that, and Microsoft won't come near you until you're making hundreds of millions of dollars.
Are they paying Microsoft? Linux on the desktop is suffering precisely because of the points outlined in the article. If Linux was awesome, less people would've switched to OS X.
I agree with you, Linux as it stands is crap for the desktop. However, if you think no corporation wants to distribute it you are crazy.
If you want to make a coherent argument, use something besides System76. There will always be a niche market that panders, the problem is that no large corporation currently does with a small subset of their offerings.
>Does anyone remember netbooks? How they used to have Linux? That was when Linux had a real chance.
>If you don't think Microsoft is directly responsible for this, you are an absolute idiot.
What about netbooks? Vista was too bloated to run on netbooks, so MS noticed they were selling well and sold XP for a low price to OEMs, and consumers preferred XP to Linux in droves and thus Linux on netbooks died. So MS is as directly responsible for this as is Starbucks for killing my local coffee shop or OS X is responsible for taking away many Linux users.
>Why would anyone invest a cent in a WM if you can't distribute Linux installed on a laptop?
Why on earth would Microsoft care about System76? They pose no threat; nobody cares.
Secondly, about Dell, that doesn't mean anything. Dell doesn't care about Linux for consumer laptops, they are probably just making a threat to force Microsoft to give them more lenient pricing for the massive # of licenses they buy. I mean, it's coming out on the top of the line XPS series, so who cares? The problem happens when the <$600 laptops start using Linux, because those sell like hotcakes and Microsoft will die before giving up that market
This is really the only reason that I can't recommend trying Ubuntu to more people. People site usability concerns etc and while there are some issues there I think it's mostly "good enough" now and we're long past the days of having to compile a .tar.gz full of .c files and fettle with vi in order to get sound to work.
So the issue is how to get third party developers interested. I think the best way is by including a really sexy app store. Ubuntu Software Centre is a start but it's still nowhere close to what Apple has achieved in this area. Nasty looking icons, inconsistent screen shots (some showing gnome2, others Unity) and thousands of free apps with weird names don't make it the most attractive place to shop.
In many ways though , I would consider desktop Linux a success regardless of marketshare for the simple reason that it is now possible to "use a computer in freedom". I think the software world would be a bleaker place if Torvalds , Stallman et all hadn't spent the hours pushing code. Imagine a world where the cheapest HTTP server license went into thousands of $.