Khaotic Musings
Random musings on Linux and Free and Open Source Software.
Monday 4 May 2009, 2:17 AM
Why Users Aren't Moving to Linux
The usual response goes something like "it's the lack of drivers", or "lack of software application [X]" or "it's harder to install and fix problems on" or perhaps even "it doesn't have any glossy marketing".
In reality, the biggest single reason for the slower-than-expected uptake of desktop Linux is much more prosaic, and it has to do with human psychology, not inherent inadequacies within the operating system itself. The reason is covered in some detail in an article by my friend and colleague Steve Jenkin.
Steve, a long-time observer of the IT industry, reviewed various presentations made by former Xerox PARC luminary Alan Kay. Alan, perhaps one of the most innovative and influential computer scientists working in the 'human computing' space, has some illuminating ideas about how people acclimatise and work with personal computing environments.
Alan covers a lot of ground, but the pertinent nugget of knowledge which applies to the uptake of Linux (and free and open source software in general) is this clearly stated by Steve in his post, so I'll quote him here:
Kay claims that 95% of people are 'instrumental reasoners' and the remaining 5% 'are interested in ideas':
"an instrumental reasoner is a person who judges any new tool or idea by how well that tool or idea contributes to his or her current goal. He goes onto talk about reward/motivation and says that 85% of people are 'outer motivated' versus 15% 'inner motivated'."
Most people (~80%) fall into the 'outer motivated instrumental reasoners' group. These people won't pick up an idea if other people aren't doing it. Which seems like a very wise evolutionary group tactic - if a little safe. Kay, in his ACM talk, uses a contagion or forest fire model to demonstrate/claim that around 66% of a population is needed to achieve 'ignition'. To hit the tipping point where 'everyone is doing it' and the new idea takes over.
This notion has been sitting in the back of my mind for a while, but it's never really made it to the forefront as a cogent idea until I read Steve's review of Alan Kay, whereupon it suddenly all gels and makes sense - it "just feels right".
So there you have it. Most people wont adopt "the new thing", be it the Internet or desktop Linux, until most of their peers have already done so. But most of them wont do so unless their peers have also done so. In fact, there seems to be hostile antagonism from most corners for such a shift. The standard chicken and egg conundrum.
But how do we as Linux evangelists solve this problem? Pretty much the same way that you solve the chicken and egg problem; through step-wise refinement and gradual shifts, until you reach the point where "the thing that wasn't quite a chicken, produces an egg which is somewhat like the kind of egg which will produce something more like a chicken". We need to similarly grow the desktop Linux marketplace every year, in time, one by one of the phase shifts and tipping points will just happen.
PreviousComments on this post
I don't wish to sound insulting, but this seems like a massively over-complicated way of stating two obvious truths about human nature.
1) People don't like change
2) People are most comfortable in groups.
In the present discussion there is an important point that is completely overlooked.
In my (possibly limited) experience, those that do make a change to Linux, and persevere long enough to get over the 'hump' never want to go back, and find Windows frustratingly limiting when forced to use it.
I think Tezzer makes a good point. People - and I mean "average users" want the comfort factor of being in a group that uses a well known platform rather than a (some might call it) maverick new breed.
Linux users in my experience are very often developers or DBAs etc - and the usage patterns within these groups can be fascinating. I did a job for someone last month using Adobe InDesign to create some great page layouts that looked cool in Acrobat Reader. But my customer insisted upon making the pages compliant with the open source Evince reader which is not as powerful and can't handle all the rendering.
Why you might ask. Well, we simplified the pages, they didn't look as good and everyone was happy.
> But how do we as Linux evangelists solve this problem?... through step-wise refinement and gradual shifts...
Or how about Innovation? Linux and the Free Software movement has it's very _roots_ in copying what other people have done. Just look at GNU's name: GNU's Not Unix!! Is there anything you can do on Linux that you can't do on a Windows or Mac box?
People won't change unless there's a good reason for them to do so, particularly as they've invested time and money into learning one way of doing things. FWIW I wrote a small blog entry about my take on the barriers to desktop Linux:
http://www.zdnet.co.uk/trackback/0,1000001387,10012459o-2000332346b,00.htm
Tezzer,
it's not merely the case that many people don't like change and that most people are comfortable in groups. Alan Kay's perspective delves deeper into human adoption of technology than this.
According to the view espoused by Kay, most people will change, even if they don't like to, as long as they're following the bulk of the crowd.
And yes, most people are comfortable in groups, but It's not enough to just state this truism. Over the past decade, most towns and cities in the world have had viable, friendly, Linux user groups, often establishing high standards of community; but this is still not enough to cause the majority of people to move to Linux.
Based on the Kay model, most people will only move across when a clear majority have already moved. Yes, it's a tautology of sorts, but there you go.
Jake Rayson says "People won't change unless there's a good reason for them to do so, particularly as they've invested time and money into learning one way of doing things."
This sounds reasonable, and in an open market, where the consumer reigns, this is what you would expect. However, we don't live in an open market when it comes to desktops, and the consumer has had to do what Microsoft has told them to do, for the better part of two decades.
Whether they wanted to or not, users have had to acclimatise to each successive Windows environment since Windows became popular (Windows 3.0, 1989). And in most circumstances, one version of Windows didn't deliver particularly high levels of innovation, as far as the users where concerned, over its predecessor; that didn't stop the successor version of Windows taking over the market.
Jake Rayson also says "Or how about Innovation? Linux and the Free Software movement has it's very _roots_ in copying what other people have done. Just look at GNU's name: GNU's Not Unix!! Is there anything you can do on Linux that you can't do on a Windows or Mac box?"
Jake, let me give you a non-exhaustive overview of innovation, Linux/free and open source software style; specifically on technologies that Linux/FOSS had (and in some instances, still has) before Microsoft's Windows.
- Real multitasking
- Remote desktop capability (network enabled GUI)
- Multi-user capabilities (ie, more than one person logged in to Linux at once)
- Separation of system binaries from end-user data, to reduce virus attack profile
- Virtual Desktops
- Desktop compositing effects (Compiz et al)
- Competent package management
- Library dependency checking
- Digital signatures on installable packages
- Competent utilities and scripting tools bundled with the OS
- Autoplay on mouse-over of media files in the file browser
Do the operating systems copy off each other nowdays? You bet. The mainstream OSes (Windows, MacOS, Linux) are resembling each othere more as time goes on, much like the automotive industry's cars have become lookalikes.
The important point is that Linux innovates as much as the others do; to whit: most everyone who's used both systems, states that the Windows 7 desktop is a near-clone of KDE 4.
In the end, innovation is not what leads to one platform snaring the bulk of the market. To understand why this is, read this
Adrian Bridgwater says "Linux users in my experience are very often developers or DBAs etc - and the usage patterns within these groups can be fascinating."
Adrian, this is a valid point. Consider the following however.
When I started using the Internet in 1989, pretty much ever one on it at the time was a developer, physicist or otherwise highly-technical. That didn't stop the Internet from morphing, in time, into something which 1.5 billion people now regularly use.
And also note. The Internet didn't change completely to make itself available to all those users; they too, in turn, had to change (or adapt) in order to use the Internet.
And just as it was feasible for most users to eventually find the Internet viable and useful, so too, in time, they may find Linux viable and useful.
What a lot of discussion around this topic (and there is SO much discussion of it!) seems to miss is that it's the apps that matter.
Who cares about the OS? Users in general don't they care about getting their work done. See my blog here for a slightly more in-depth discussion of this.
One day and maybe soon there will be a choice of software and hardwired to the motherboard OSs much in the same way that we can choose vehicles.
They will all have wheels and be capable of a unified minimum speed i.e they will all be able to run open office and handle any file type as standard, network with each other and draw power in a way we have not yet imagined. He says with all the hope he can muster.
Manek says "What a lot of discussion around this topic (and there is SO much discussion of it!) seems to miss is that it's the apps that matter. Who cares about the OS? Users in general don't they care about getting their work done."
In general, yes, it's the applications which matter.
However, your blog's argument can be broken down into two parts, and I can show that Linux can account for viable solutions in part now, and in whole eventually.
One, says that users might know and rely on an application which isn't available on Linux, and has no close facsimile. For example, you mention Cool Edit 2000 and how Audacity isn't quite the same.
The other is that users might want access to applications in future, for example new games, which may never become available on Linux; ie, it's the loss of application potential which matters.
To respond to these in turn; with a combination of tool like Wine or VirtualBox, you can run pretty much any business-type Windows application under Linux, today. Microsoft has had to rely on the same strategy with Windows 7, so that it can run Windows XP apps.
This facilitates the viability of Linux as a desktop replacement for Windows today. As Linux reaches milestones in popularity, it becomes a worthwhile platform for those vendors who write software which isn't viably run under Wine or VirtualBox, eg high-end games, to also target-build those applications for Linux.
This is what happened with games vendors when the eventually target-built their applications for Windows, rather than DOS. It takes time, but it will happen.
Linux isn't going away, unlike Microsoft's other competitors, and as each year brings cheaper computer hardware, Linux's competitive advantage grows and grows.
I have been using Linux on and off for 10 years and use it as my main personal OS and to administer a Windows environment in my job, ho ho. Boy is it better today as a desktop OS.
There is one major point that I think everybody has missed regarding the Linux v Windows debate is the fact that Windows comes pre-installed. I bet 90% of Windows users have never installed an OS. Nor would they want to or lets face it should have to. Pre-installed Linux has only ever been a small dalliance for major hardware manufacturers. Netbooks, well thats another story.
The other fact is that im sorry to say, as much as I love Ubuntu, out of the box it is ugly. Linux mint is a step forward, but is still missing something.
Microsoft have designed a new font for Vista / Windows 7, Segoe UI and OSX has its own, they are pretty nice. Ubuntus default font (I dont even know the name, as its the first thing I change) is horrid, please Mark Shuttleworth find somebody to design a unique Ubuntu Sans Serif font.
Ubuntu 9.04 is making inroads towards OSX but it feels like an embarrassing catch up, especially with the nearly-identical growl notifications. However projects such as Gnome-do are breathing new life into Gnome for sure.
Its certainly not just about Windows v Linux anymore as OSX has made serious inroads, the amount of 20 somethings I deal with in my job have got macs and love them with a passion. I hate to admit it, they look smooth and if you have used Linux, its not unfamiliar territory.
Whats the future? Microsoft will stay as Enterprise desktop OS for a while yet, with hackers and the like preferring linux and OSX for your style conscious digitalier.
I'll always stick with Linux as I love it, what it stands for and because it is so configurable. The desktop can change to suit whatever style fad I am going through. I have to live in hope for such basics as a decent flash player that doesn't hog the CPU and is choppy on fullscreen, but I will persevere.
This comment has been deleted at the users request
BlackHoleSun says "There is one major point that I think everybody has missed regarding the Linux v Windows debate is the fact that Windows comes pre-installed. I bet 90% of Windows users have never installed an OS. Nor would they want to or lets face it should have to. Pre-installed Linux has only ever been a small dalliance for major hardware manufacturers. Netbooks, well thats another story."
I completely agree. Windows has its position of dominance because of the complete stranglehold Microsoft has in OEM pre-installs. It achieved this market dominance illegally through highly anti-competitive means. The only way to break this nexus to to ensure Windows is unbundled from all PCs
BlackHoleSun also says "The other fact is that im sorry to say, as much as I love Ubuntu, out of the box it is ugly. Linux mint is a step forward, but is still missing something."
There will always be areas where Windows or MacOS does a better job than Linux, and other areas where Linux does a better job than Windows or MacOS. The existence of these areas aren't enough to account for the disparity in marketshare however.
This comment has been deleted at the users request
Yes, up to a point. I agree with you about the illegal monopoly and all that -- I remember reporting all those legal stories in the 1990s...
However, it's never going to be about the functionality. All the OSes we're talking about do way more than most users will want to do. They're hugely capable. It's to do with the familiarity.
Imagine if all PCs came without an OS, and buyers had to pick one. Which would they pick? Windows (mostly). Why? Because they know it. And why is that important? Because (and I return to my earlier trope), because it's the apps that matter...most people don't want new stuff, they want familiar stuff.
I mean, (depressingly) the most used theme in advertising is "follow the herd, "do what everyone else is doing", wisdom of the crowd, all that stuff. That's because it works. Sadly.
Manek
anonymous says "Imagine if all PCs came without an OS, and buyers had to pick one. Which would they pick? Windows (mostly). Why? Because they know it."
This is indeed complete conjecture, however, as we've never been in such a scenario. We don't know what proportion of users would select a non-Windows option, if they were given a choice.
Well, actually, that's not quiet true...
We do know what happens when consumers are offered the option to buy an alternate OS instead of Windows, on the same client-side computer - Linux snares one-third of the market share:
One Third of Dell Inspiron Mini 9s Sold Run Linux
Contrast this to the few percent that Linux gets when it has to compete with Windows pre-installs on desktops and notebooks, and you see that there are 10 times as many people who would opt for an alternative to Windows than current competition practices enshrined by Microsoft allow.
I do agree with the broad thrust of your argument. I said that they would pick Windows (mostly) - and two-thirds is mostly.
And just to be clear, I'm absolutely NOT saying that this is how things should be.
[Not sure why I'm "anonymous" - I'm logged in]
Manek
"I mean, (depressingly) the most used theme in advertising is "follow the herd, "do what everyone else is doing", wisdom of the crowd, all that stuff. That's because it works. Sadly."
Actually the theme of modern advertising is stand out from the crowd. Buy this brand to look cool and distinguish yourself as a person "in the know".
Thats the tactic Mac uses to keep its OS alive. It appeals to the cool people (graphic designers etc.).
Maybe thats what linux need to be adopted. Just be cool. Get it mentioned in hip hop lyrics and sponsored by orange.
Unfortunately for linux, its developed by the very antithesis of cool. On a positively note though, the open source "movement" as a political ideal is gaining in coolness(Radiohead etc.). Todays alternative is tomorrows mainstream, so perhaps linux adoption will follow once the mundanes realize it is open source software. Basically you need cool linux users and the young will follow.
1000238123 mentioned "Thats the tactic Mac uses to keep its OS alive. It appeals to the cool people (graphic designers etc.). Maybe thats what linux need to be adopted. Just be cool. Get it mentioned in hip hop lyrics and sponsored by orange."
It's true, the Mac has high-gloss marketing appeal and branding at present, something that Linux can't match.
However, keep in mind that Microsoft's own polling, as shown in this slide by Steve Ballmer just the other month, indicates that desktop Linux has more marketshare than Apple.
More importantly, while it is easy for Linux to acquire many more users, if not adherents, for example through systems such as this one it's not as easy for Apple to do the same, with PCs and laptops which cost five times more.
Hmm. "Be cool." "Pay more".
Hardly a tactic for driving mass market adoption is it - especially now? I stand by my earlier assertion, which is that, broadly speaking, there's no greater used tactic in advertising than 'follow the crowd'. As a strategy, it has some solid anthropology behind it too. And if desktop Linux is to succeed it needs to reach beyond those who care about cool towards those who want fucntionality.
And right now, Linux offers less functionality because it doesn't easily run all the apps people use.
Do I sound like I'm repeating myself...?
Manek
As a relatively non-technical computer user (I do a bit of web design and have a Linux VPS with a control panel) I agree with comments made about familiarity and compatability.
I've been thinking about trying Linux for sometime and finally installed it on a partition so I had the dual boot option. I'm yet to try Gimp and Kompozer for my web design and have so far stuck to trying to getting set-up to my liking. I'm impressed with many aspects and the performance is much better than Vista! However, trying to work out how to do somethings is a bit trying at times. A Google search here and there and I find a solution that involves editing files or using command line. I'm used to editing files, so no problem there, but for the average computer user its going to be very scary. I've rarely used command line before and find it awkward. Your average user isn't going to be too bothered about making the effort and that is the problem Linux will face.
I'm sticking with it and hope to eventually move completely to using it once I can get things like my camera connection working and even minor things like being able to program the buttons on my mouse; you know, those little things that makes using a computer more comfortable.
Until users can get by completely without having to edit files and use command line I'm afraid Linux will remain largely the domain of computer professionals and enthusiasts. That said, if more corporate IT departments can be convinced to switch and invest in user training it will make the leap for switching at home that much easier and less forbidding...
Mike
Actually, users are moving to Linux, and once they do, they seldom go back. So what we have is the slow but inexorable shift.
The strategy I find most effective is to first convert users to OpenOffice, and then to convert one of the workstations to Linux, usually as a replacement for an old computer.
Once people experience the success of that, they are then willing to migrate entirely to Linux as long as you can ensure that all of the apps they need will either run or have viable equivalents.
Switching people to web oriented apps is still Microsoft's biggest nightmare, at that point only the browser is relevant the OS doesn't even matter.
It's easy to sell people on the reliability and cost of Linux when Microsoft keeps shipping krap like Vista.
One of the biggest arguments against converting to Linux is empty. The argument is that the cost of user retraining and conversion is too high. Well guess how high is the cost of user retraining for MS Office 2007 and Vista? Converting to those has been a huge cost, and so suddenly Linux makes great economic sense. Vista is so different from XP that it is less effort to switch people to Linux.
as mentioned elsewhere, I've been testing linux on my wife.
using an 'ubuntu 9.04 netbook remix' usb livedisk, she's been more open to trying it out rather than having XP replaced on her netbook, and so far likes what she sees, even going so far as to say it's "cute"!
she, like many others, though, does not like change... just getting her to try the livedisk was a challenge in itself, but given how many times I have to help her out with even the simplest of tasks in XP, I have been pleasantly surprised how little intervention I've been required to give with ubuntu so far...
but that, I think, it the crux of the problem. i noted that a few people pointed to the fact that the average user doesn't install their own OS, and given the choice, most (ok, 2/3rds) will opt for a Windows variant, simply because they "don't like change."
for my own part, I have little option at the moment but to run a dual-boot system with WIndows & ubuntu. some programs are built for a windows environment only, and being a relative newb to linux (as well as being horrendously busy in a school), i haven't gotten around to trying wine or virtualbox yet.
when i get the chance, i suspect that i'll shift over to linux for good in the near future...
I have a friend who's pc has been playing up again - i suspect that some of the fault may lie with her teenaged sons... maybe i'll set up ubuntu on their pc, just to see...
I note that the following article, posted in the Wall Street Journal, covers much the same phsychological terrain that is alluded to in the original blog post heading this thread:
Quote:
Psychologist Stanley Milgram demonstrated that people pop over even if everyone is looking at nothing, by sending a group of experimenters into the street to stare upwards. With a large enough group of gapers, passersby stopped to stare, too.
A more-recent study demonstrates that popularity in the music world, even unearned, breeds more popularity. Researchers enlisted more than 12,000 volunteers to rate and download songs from among 48 chosen for their relative obscurity. Some of these volunteers were lied to: At a certain stage in the experiment, popularity rankings for this group were reversed, so the least-downloaded songs were made to appear most-downloaded.
Suddenly, everything changed. The prior No. 1 began making a comeback on the new top dog, but the former No. 47 maintained its comfortable lead on the old No. 2, buoyed by its apparent popularity. Overall, the study showed that popularity is both unstable and malleable.
