The success of Ubuntu

Source

(final edit: okay, I posted this, had second thoughts about the prudence of posting it and hid it, but it seems like everyone saw it anyway, so what the hell, I'll unhide it again. It's maybe not the most diplomatic thing to post, but...erf.) Very important disclaimer: this post is utterly and entirely my own personal thinking and in no way whatsoever reflective or representative of my employer. Quick foreword, edited in later - I forgot to explain the reason for this post. When people talk about Ubuntu's shortcomings they often hedge it around by saying 'but hey, they're good at marketing', or 'hey, they make a successful desktop'. I'm arguing that this isn't so clear cut. I'm going to try and keep this as succint as I can, as this kind of thinking has gotten me in Big Trouble before, but commenting on Greg's blog post got me thinking. Has Ubuntu, really, been successful? Becoming the 'leading' Linux desktop distribution is a pretty paltry measure of success, after all. As I wrote on Greg's post, no-one's really taken SUSE seriously since Novell bought them (though that seems rather unfair, and probably *is* a problem of perception / marketing - from all I hear, it's still a very solid distro), Mandriva has been basically bankrupt since 2003 and has huge management problems (also, can't string together a press release in correct English, which makes appearing more professional than Mandriva a freaking low bar. Don't get me wrong, I love Mandriva, but the company has huge problems. I never said it didn't when I worked there. I just didn't *talk* about them. Publicly. :>), and Fedora and Debian aren't really Uncle Bob-type distributions. All the rest are derivatives of the others, or second-stringers. If you show up with a couple of graphic designers, anyone who's passed Media Relations 101, and a bit of cash, you can pretty much win by default, which is what Ubuntu did. So, let's not go on the kiddy pool stuff. After all, that's not how Ubuntu says it's measuring success. Remember Bug #1, filed in August 2004? It doesn't say 'Mandrake has the biggest share of the Linux market', it says 'Microsoft has a majority market share'. So, let's look at Ubuntu's record in *those* terms. Let's look at the W3Schools operating system share statistics. In July and September 2004 (so presumably also in August), Linux is at 3.1%. In June 2010, after nearly six years of Ubuntu as the generally-perceived Linux desktop standard bearer, Linux is at...4.8%. In March 2003, Linux was at 2.2%. So that's a rate of growth of 0.9% over 16 months to July 2004 - 0.05625 percentage points per month. The rate of growth from July 2004 to June 2010 is 1.7% over 71 months - 0.02394 percentage points per month. The margin of error in those numbers is likely huge, because we're playing with such small numbers, but even so, it sure doesn't look like Ubuntu has even managed to increase the rate of growth of Linux one iota over the 'leading desktop distributions' that preceded it (in the 2003-2004 range that was probably Mandriva; before there was Gentoo and Red Hat Linux, and SUSE was always there or thereabouts). It's hard to find stats from the other places that track operating system usage that go back as far, but going back as far as they do - to around 2007 or so, usually - they seem to tell much the same story. I can't find any which show really significant growth in general Linux adoption, or a significant increase of the rate of growth at any point in Ubuntu's tenure. So, is this success? Really? Discuss. :) Given what's discussed in this post, I'd say the top 3 damage-dealers to Microsoft since Canonical started up have been Mozilla, Google, and Microsoft itself (hello, Vista). Canonical probably ain't in the top 10. (To pre-empt the obvious point - yes, I was working for Mandriva and Red Hat during most of the period under discussion, and we haven't revolutionized the world in terms of Linux market share either. But Mandriva never exactly aimed to; it aimed to be a successful commercial Linux distributor, and was always more about effectively serving a relatively small market than trying to Take Over The World. I'd argue Mandriva's usually provided an excellent product to its mostly-loyal user base, and almost all the company's problems, which are significant, are in management and marketing. Red Hat, also, has never said we're aiming to make significant inroads into general user desktop market share. This is not the same as 'we don't care about the desktop', note, but the emphasis is slightly different. We maintain a great desktop distribution, which I work on, and do a large amount of work on the F/OSS desktop - see Greg's post - but not in exactly the way Ubuntu claims to be aiming for.)

Comments

Anon wrote on 2010-07-30 08:57:
Hmm. I think the issue is that Ubuntu has grown the edge of the Linux market in areas that the people who built a lot of the original work never really wanted to adopt. In the academic places you see students adopting Ubuntu (whereas labs may be SUSE/Fedora/Slackware etc.) if they want to use something non-Windows on their PC. The area that Ubuntu has made easiest is for things like retrieval of binary drivers and proprietary codecs - I'm not even thinking about the actual writing of said drivers/codecs. This is starting to play out in that you see binary things only being tested or packaged for RHEL and versions of Ubuntu (e.g. games). I imagine Chrome/Chromium is likely to be shipped in the core Ubuntu repositories before Fedora due to its build choices. Even getting Ubuntu feels easier than Fedora - there's one CD for the i386 version (far from optimal but simple). This combined auto installed binary drivers makes it easier to recommend to the few people who ask about Linux as it feels like they will have tried most combinations of open/closed in one go. But is it successful among the masses? Not really. You will get the occasional "high profile" switcher going from OSX to Ubuntu but most people who wanted a Unix that was easy went to OSX years ago. Maybe the odd techie organisation will tacitly base itself on Ubuntu internally (e.g. Google). I think a large number of Ubuntu users (including myself!) have been lured away existing Linux distributions. After years of seeing the problems, I'm not interested in trying to market Linux to those using non-Linux systems. I'm not making purchasing decisions so swaying me is pretty useless anyway :)
Peter Webb wrote on 2010-07-30 11:09:
Is this the most appropriate statistic to your hypothesis? Really? Discuss.
adamw wrote on 2010-07-30 15:23:
Peter: if you mean is this specific operating system share metric the best, possibly not, but as I wrote, it's the only one I could find with a history back to 2004 easily available. If you mean operating system usage share isn't the best statistic to use, what would you suggest is? Especially given the wording of Bug #1?
Adam Miller wrote on 2010-07-30 13:17:
Amazing post. Yet another reason why you're on my list of heroes! :)
Peter Webb wrote on 2010-07-30 16:01:
Are you satisfied that you have an accurate measure of OS usage statistics? The problems in quantifying OS share as related to FOSS has been discussed number of times before, with no answer. Hard numbers are so difficult to get. I can offer you another set of numbers. http://www.google.com/trends?q=ubuntu,+fedora&ctab=0&geo=all&date=all&sort=0 But what conclusions can I draw from this? Do you believe your conclusions are justified, given that your starting point is dubious?
adamw wrote on 2010-07-30 16:15:
peter: I don't think there's any particular difficulty with measuring OS share in those visiting a given website, it's a simple operation. The difficulty is in positively *counting* the total number of installs of a F/OSS operating system, which is slightly different. You can argue all day about the biases that will be caused by the type of people most likely to visit a site, but it's worth noting that w3schools is, by a considerable margin, the most Linux-friendly of the major trackers; all the other most widely-cited operating system usage-tracking stats have Linux on substantially lower numbers (like, often under 2%). And as I said, they all showed about the same trend as W3Schools - very slow growth, which does not appear to be at all accelerated by Ubuntu. I suppose the easiest way to look at it is to ask yourself: if Ubuntu were really succeeding as Bug #1 suggests it is trying to, would the number W3Schools report be going up much faster than it is? I can't see any reason to answer that question with a 'no'. Look at W3Schools' browser numbers over the same period, for instance, and see the growth of Firefox. Clearly, when an F/OSS project is actually wildly successful, W3Schools sees it.
adamw wrote on 2010-07-30 16:17:
By the way, the conclusion you can draw from your number is that lots more people do Google searches for 'Ubuntu' than for 'Fedora'. That's really all. And remember, I already wrote that comparing Ubuntu to other distros is really not the point here. Fedora isn't trying to 'beat' Ubuntu.
Gabriel wrote on 2010-07-30 20:49:
In few months, Windows 7 RC usage statistics surpassed all desktop Linux distributiones combined. I'll let the Ubuntu fans draw their own conclusions on Ubuntu's so called "success"
Benjamin Otte wrote on 2010-07-30 21:59:
You are aware of http://stats.wikimedia.org/wikimedia/squids/SquidReportOperatingSystems.htm ? It only goes back to April 2009 - http://stats.wikimedia.org/archive/squid_reports/ - but it differentiates between different Linux distros. And it seems Ubuntu has been doing a good job of attracting former Debian and Feora users.
adamw wrote on 2010-07-30 22:12:
Um. I *did* explain, right up there in the post, why 'how many users does Ubuntu take away from other distributions' is not a valid measure of Ubuntu's success, by Ubuntu's own standards. Just as I wrote up there, Bug #1 is not about becoming more popular than other Linux distributions, it is about making Linux more popular than Windows. That's supposed to be Ubuntu's ultimate raison d'etre. Go back to 2004 and you'll find lots of interviews with Mark happily talking this up, though it doesn't seem to be so prominent these days.
LinuxLover wrote on 2010-07-30 22:38:
Adam, you always say what I seem to want to say, but can't get the words, or the numbers, to say! I've been saying for quite sometime that Ubuntu is trying hard to simply replace the word Linux with the word Ubuntu, as if it stands on its own. They also seem to take the credit for everything. I've seen what both Novell and Red Hat have contributed to Open Source, and to a lesser degree, Mandriva. However, Ubuntu has never seemed to be much more than a huge marketing movement to lay claim to the Linux market. Shuttleworth has made big rants about user interfact problems within FOSS, but I haven't seen him sink one single US dime into fixing them. If he's so concerned, and really wants to move Ubuntu forward, why doesn't he invest some of his resources to help Gnome out more? To help GNU out more? To help KDE out more? Fact is, he's relying on someone else to do it for him, so he can market it and take credit. Shuttleworth could simply come clean if it wasn't all true. He doesn't. The hypnotized masses that follow him just believe everything he says/writes, without question. This seem eerily similar to a certain fruit-flavored computer company.
adamw wrote on 2010-07-30 22:47:
Well, Ubuntu has been trying to address UI issues lately, with their new notification system and Unity and things like that. We can talk about whether they're managing these in the best way, but they're genuine efforts. I'm not exactly talking about how much work Ubuntu puts in, with this post, more about how much of a result it's having...
Peter Webb wrote on 2010-07-30 22:48:
Just like the google trends page, the only inference you can draw from w3schools data is that the number of visits made to w3schools from linux based systems as a proportion of total visits has increased slowly. The rigour you apply to the inference of google data, you do not apply to the w3schools data. You are trying to minimise the effect of sampling bias. Linux friendly is not the criteria for choosing a sample. The sample is chosen to be as representative as possible on the population on which you wish to derive an extrapolation. There is an additional problem in interpreting rates; both the numerator and the denominator are important. Small percentage changes in a very large denominator translates into large absolute changes. I also can't see how you can infer "specific" ubuntu numbers from "generic" linux numbers. As an aside, I think Bug #1 is just a tongue in cheek remark. I do not think it reflects the ideals of Ubuntu or Canonical, but that is only my opinion.
Peter Webb wrote on 2010-07-30 22:57:
Another point I forgot to mention is nicely demonstrated by the wikimedia link posted by Benjamin Otte. The counts do not map to unique IPs, so there is a very large issue of over-counting. For example are we to infer that there are more iPad users than Fedora users?
adamw wrote on 2010-07-30 23:02:
Sure, w3schools isn't a perfect metric. Few metrics are. But it's a far, far better one than a comparison of search terms (and not even *relevant* search terms, since as I was at pains to point out, this is not about Ubuntu vs. Fedora, which isn't even a fight that makes any sense. No, don't bother with a Google Trends for Ubuntu vs. Windows, because people don't search for Windows the way they search for Ubuntu. Very few people need to look up Windows to find out what it is. Sadly.) As I said, I went out and found all the operating system usage metrics I could, and none of them contradict the w3schools numbers. Everything I've seen in the real world confirms the trend w3schools indicates: Linux usage is tiny, and is rising as slowly as ever it did before Ubuntu was around. Are you actually disputing that these things are true, or just picking holes in the metric I used to indicate the trend to no particular purpose? Do you actually believe that Ubuntu has increased Linux usage significantly more or at a faster pace than was the case before Ubuntu was around? w3schools being Linux friendly is in fact important, because it means the Linux numbers are at least large enough that it's vaguely possible to draw conclusions from them. In all the other surveys I could have used, the Linux usage numbers are so low that it would be almost impossible to draw any certain conclusions from any change in those numbers, because they'd be way within the overall survey's margin of error. "I also can’t see how you can infer “specific” ubuntu numbers from “generic” linux numbers." The great thing is, I don't need to. If Ubuntu was being successful in widening the appeal of Linux, the number of Linux users would be increasing much more rapidly. You don't need to count the number of people who use Ubuntu specifically to test this, because everyone who is gained as an additional user by the efforts of Ubuntu would of course show up in the 'Linux' number. The only times the distinction would become important would be if you were trying to compare the impact of different distributions on the growth of Linux, which I'm not, or if you were to theorize that other distros did such a disastrous job that they lost a large enough number of users - who, inexplicably, didn't switch to Ubuntu or some other Linux distro, but switched (back) to Windows or OS X - that it was covering up growth caused by Ubuntu. I really don't think that's plausible.
adamw wrote on 2010-07-30 23:06:
"For example are we to infer that there are more iPad users than Fedora users?" Honestly? It wouldn't particularly surprise me if there were more visiting Wikimedia, yeah. Apple's sold, what, well north of 3 million iPads by now? Since it's so new, lots of those are likely to be in very active daily use. We claim some ballpark number of 12 million or so Fedora installs, but lots of those are probably headless machines that would never go to Wikimedia, and it's probably something of an over-count anyway. It honestly wouldn't surprise me much if more people are using iPads for typical every day web surfing than are using Fedora for it. Or, for that matter, are using Ubuntu for it.
Jono Bacon wrote on 2010-07-30 23:22:
"All the rest are derivatives of the others, or second-stringers. If you show up with a couple of graphic designers, anyone who’s passed Media Relations 101, and a bit of cash, you can pretty much win by default, which is what Ubuntu did." I find this incredibly disrespectful and insulting to the global community of Ubuntu contributors. Do you really think that being so insulting and derogatory is helping? Disappointed.
adamw wrote on 2010-07-30 23:29:
Jono: That's not to say that's all Ubuntu is *now*, but it's all Ubuntu needed to start out with. Look back at the first Ubuntu release - it's Debian with a package selection, a coat of paint, and a press release, and that was all it needed. (If anything, it reflects terribly badly on Mandrake, which I was heavily involved with at the time. Look on it as self-deprecatory. :>)
adamw wrote on 2010-07-30 23:33:
Jono: I'd be rather more interested in your take on the wider argument, really. Whether or not you think it's a significant achievement to become the 'leading' desktop Linux distribution, that's professedly not what Ubuntu is about; in fact, I recall one of the responses to my old 'Ubuntu ate the desktop distro market' post was that it's fine, because there was a purpose behind Ubuntu doing that, and that purpose was to effectively take on Windows. In that context, do you think Ubuntu is succeeding? Do you think the trend I cite is reasonably accurate? Do you - both you personally, and Canonical - view this very slow growth as satisfactory, or is it something you're not happy with?
Peter Webb wrote on 2010-07-30 23:26:
A bad metric may be better than a worse metric, but the conclusions will still be bad. In fact, you clearly articulate this in the context of iPads, but you don't consistently interrogate the assumptions elsewhere. My point about the w3schools metric is that you are assuming the w3schools user base is an adequate representation of the ubuntu user base. Considering that ubuntu has been promoted among non technical users and w3schools user base is more likely to be technical users, the metric you use is biased towards a small subset of ubuntu users. There is no reasonable estimate for this proportion. And there is the problem of over-counting, which brings me back to my first sentence. As far as I can see, there is no objective way for anyone to claim anything about OS statistics.
adamw wrote on 2010-07-30 23:37:
Peter: All I can say is I disagree. As I've kept saying, although I only cited the W3Schools numbers in the post, I did in fact look at a lot more sites, and none of them told a different story. I'm not aware of any evidence to support a theory that Ubuntu is driving a significant increase in the Linux user base. My theory is that Ubuntu is not doing this. If you actually have any evidence to bring to bear that would contradict my theory, could you please present it? What is there to prove that Ubuntu is making a significant splash in the operating system world?
Peter Webb wrote on 2010-07-30 23:52:
You are still missing my point. I am not disputing your hypothesis. I am disputing your evidence. I am not claiming the opposite of your hypothesis, because I have no evidence. I am agnostic to the whole issue. All I am saying is that there is no evidence either way. What you claim as your theory is just your opinion (or wishful thinking) backed up by incomplete or inappropriate data. For all we know ubuntu might be increasing its share by cannibalising other distros.
adamw wrote on 2010-07-31 00:02:
Then, honestly, you're just nitpicking. It's a blog, not New Scientist, I didn't feel the need to come up with 50,000 whizzy graphs saying the same thing. Like I said, I looked at all the numbers you can find, and they all say the same thing. I look around the real world, I see the same thing. I see people all over the place using Firefox. I never see anyone using Ubuntu (or, for that matter, any other Linux distribution). I really think the trend is pretty obvious if you're willing to consider it. Why would that be wishful thinking? Why would I want things to be that way? Certainly I don't. Whatever I think about Ubuntu, I'd much rather people were using it than Windows or OS X. Sadly, it seems they aren't.
Jono Bacon wrote on 2010-07-31 00:17:
Adam, Frankly, I am not interested in getting wrapped into a debate around metrics. The inaccuracy and unspecific nature of Open Source metrics has been well discussed in the past and I am more interested in bringing my skills and contributions to Ubuntu to help it succeed. This is what our global community seeks to do and like any other community, puts in long hours and a significant investment of time and energy. It is with this reason why I found the quote I highlighted earlier so disrespectful.
adamw wrote on 2010-07-31 00:27:
Jono: it's not about 'open source metrics'. As I wrote above, indeed, counting a number of installs of a given distro is very difficult (this is something Fedora goes to great lengths to point out, but Canonical never seems to, instead throwing impressively large yet very vague numbers at pet journalists). But this isn't an 'open source metric', it's a simple 'what OS are the people coming to this website using?' metric, which is pretty tough to screw up. The point I'm making isn't fundamentally about specific numbers, but about a trend: Linux usage is very small, and is not getting larger very fast at all.
Jono Bacon wrote on 2010-07-31 00:37:
Adam, ...and there we have our disconnect, you have a belief that the figure is small and not experience much growth, but my feeling is that we are growing across the board. Ultimately though, it is all just hypothesis as the figures are...well...hypothetical. :-) I am out, have a great weekend. :-)
Actual Scientist wrote on 2010-07-31 02:57:
I'm sticking up for Adam here. In the world of hypothesis testing, a null hypothesis is tested against an alternate. Invariably, the null hypothesis is some version of "there is no difference in _Y_ because of phenomenon _X_" (hence its catchy name--the *null* hypothesis). In this "study," the alternate hypothesis might be phrased as "there is an increase in the rate of Linux adoption subsequent to the advent Ubuntu." The null would be "there is no difference in the rate of Linux adoption subsequent to the advent of Ubuntu." Science is inherently conservative, and favors the null hypothesis by default--that is, science prefers a false negative to a false positive. Absent strong evidence to the contrary, we "fail to reject" the null hypothesis. (It's important to note that one never *"accepts"* the null--one merely "fails to reject" it.) Adam's point: that we fail to reject the null, is the only scientifically responsible conclusion supported by the available data. Which is not to say that the null is true in this (or any other) case, but rather only that the available data fail to refute it. It's worth noting that Adam's post is not addressing the question, "Is Ubuntu taking 'market share' away from other Linux distros?" That may be a fascinating question, but it's not the question under scrutiny. "Do people prefer wheat pancakes to buttermilk?" may also be a fascinating question, but it's just not the focus of this particular blog post.
Mahesh wrote on 2010-07-31 05:37:
Really? Mandriva never claimed to be a leader? http://www.mandriva.com/enterprise/en/company/press/mandriva-appoints-herve-yahi-as-chairman-chief-executive-officer "Mandriva is a market leader with tremendous assets.." Leader! http://www.mandrivaindia.com/aboutus.html "he company’s Linux distribution quickly became a leader in many major markets." Leader! Leader of the 3% market share that you claim is very miniscule to be called success. So, how come you didn't do the right thing and fix it all in your stint at Mandriva? One can argue anything. Recognize the efforts of people, whatever they do. If you can't just move on. Don't rant and create the world a sad place with negativity.
adamw wrote on 2010-07-31 06:46:
Mahesh: those are talking about the Linux market, not the operating system market. It can't really be otherwise, as there's no way Mandriva could be a leader in the operating system market. =) As I said, Mandriva's goals were usually within the Linux market. Your quotations simply support my point.
Techrights » The Ubuntu-GNOME Debate Carries On wrote on 2010-07-31 05:59:
[...] Williamson (Fedora): “The success of Ubuntu” In July and September 2004 (so presumably also in August), Linux is at [...]
Madsen wrote on 2010-07-31 06:33:
Ubuntu has indeed helped people move to Linux. The first Linux-distro i tried 3 years ago was indeed Ubuntu. However, nowadays i and my famile we are using mostly Mint, Fedora and Mandriva. My personal fav is Fedora with Gnome and Mandriva with KDE. So perhaps Ubuntu could nowadays be a port to Linux. Majority of Linux-users however don't want to use only Ubuntu. They wanna take a look what Linux is all about.
Peter Webb wrote on 2010-07-31 07:58:
@adamw > It’s a blog, not New Scientist Therefore, it is an opinion, but you parade it as a fact. You are entitled to your opinion, and others are entitled to the opinion that you are wrong. Since you have invited people to comment in your blog, I think it is fair to expect it be critically examined. In no scientific endeavor would a critical examination of assumptions and data would be considered nitpicking. @ Actual Scientist You are right about hypothesis testing and default positions. I am not disagreeing with the hypothesis. But is the hypothesis falsifiable with the metric chosen? What I don't understand is that similar evidence from wikimedia can be explained away, but that from w3schools can be accepted without question. Seems like confirmation bias to me.
Gabriel wrote on 2010-07-31 09:24:
- OEMs are ignoring Ubuntu - With the same exceptions that before, there's no commercial software available for Ubuntu - Again with a few exceptions, no official hardware support for Ubuntu Do you guys really need graphs and figures to find out Ubuntu is going nowhere? And what should that even happen? All the things that I read about Ubuntu finally making Linux accessible to the average Joe I read already ten years ago, when Corel Linux was released. Not much has changed since then.
Canonical, Red Hat und das Trittbrettfahrerproblem « Volker Radke wrote on 2010-07-31 11:34:
[...] Red-Hat-Mitarbeiter Adam Williamson hat gezeigt, dass der relative Erfolg von Ubuntu bisher eher zu Lasten anderer Linux-Mitbewerber als zu Lasten [...]
Peter Webb wrote on 2010-07-31 12:02:
@ Gabriel I have corrected your post for you. Hope this clarifies things :) - OEMS are ignoring Linux (with a few exceptions) - With the same exceptions that before, there’s no commercial software available for Linux - Again with a few exceptions, no official hardware support for Linux All these things I hear about Linux being user-friendly or useful at all in any sense, and the year of the Linux desktop I have heard for the last ten years. Let us just ditch Linux and go with Microsoft and Apple. Do you actually have an argument to make?
Gabriel wrote on 2010-07-31 13:22:
@ Peter Webb Yeah, what you say is even more accurate. So how exactly can Ubuntu be defined as a success, then?
adamw wrote on 2010-07-31 15:28:
peter: er, 'explained away'? You proposed the argument that more people are using iPads than are using Fedora (implicitly meaning 'as a desktop'), expecting me to angrily deny it. I didn't. I said you're probably right, and there probably *are* more people using iPads for typical personal computing than using Fedora for it. I didn't explain anything away...
adamw wrote on 2010-07-31 15:29:
Peter: Gabriel's exactly right. As I keep saying, I'm not arguing that any distribution is doing any better than Ubuntu in this way. But Ubuntu is the distribution which claimed it was going to finally succeed where others had failed in doing all those things.
GreyGeek wrote on 2010-07-31 16:42:
The Linux desktop market share is at 4.5% ???? Steve Ballmer disagrees with your statistics. He could be lying. Or not. I suspect that he has sources of information that are more accurate and reliable, if he was going to run his business using them, but regardless, he stated 18 months ago that the Linux DESKTOP market share was larger than Apple's and Linux was Microsoft's greatest competitor. At the time, Apple's market share was around 10%. VISTA only helped people to move away from Microsoft, and Win7 hasn't done that much to bring them back. Win7's growth is more from cannibalizing VISTA and XP users. While Apple is picking up some of the migration, most go to Linux because they can't afford an Apple. Concerning W3Schools statistics, it is common knowledge that many Linux users spoof their browser OS to avoid incompatibilities with Windows centric websites, even on websites that claim, or are, browser neutral. NetApplication's "statistics", in which they claim the Linux desktop market share is less than 1%, a favorite citation of Windows fanboys, is even more flawed when one considers that it is a firm which specializes in selling rebranded/renamed Windows executables. No chance they'd want to tilt the field. Not! My own statistic is that in the last two years I have had nearly two dozen Windows users SEEK ME OUT to ask me to install Linux on their computers. The most recent was a week ago. The most common reasons for asking: they've lost patience with Windows malware problem, and with the costs of keeping Windows running. Since I also used to support most of them when they were running Windows (what are friends for?) I find that the number of calls for support help have dropped to nearly zero. Now it is "What kind of printer should I get?" and the like.
adamw wrote on 2010-07-31 18:03:
"Steve Ballmer disagrees with your statistics. He could be lying. Or not." Er, does he? I don't see any numbers on that graph, but the Linux slice sure looks like 5% or less to me. I doubt the Linux share of desktop use is near 5%, actually. The w3schools numbers are, as I've mentioned, substantially more favourable to Linux than any other site that tracks such numbers. Most sites give a number around 1-2% for Linux. "Concerning W3Schools statistics, it is common knowledge that many Linux users spoof their browser OS to avoid incompatibilities with Windows centric websites, even on websites that claim, or are, browser neutral. " It's a common excuse. I doubt it's actually true, particularly these days, when non-IE browsers have a huge share of the browser market. I don't actually know anyone who spoofs their browser agent. Let's face it, if we're assuming a significant number of Linux users are people technically proficient enough to know what 'browser agent spoofing' is and how to do it, that rather supports my argument. =) "My own statistic is that in the last two years I have had nearly two dozen Windows users SEEK ME OUT to ask me to install Linux on their computers. The most recent was a week ago. The most common reasons for asking: they’ve lost patience with Windows malware problem, and with the costs of keeping Windows running. Since I also used to support most of them when they were running Windows (what are friends for?) I find that the number of calls for support help have dropped to nearly zero. Now it is “What kind of printer should I get?” and the like." Yeah, lots of people have anecdotes like that. But there you go - people seek you out to install Linux on their computers. That ain't the way to market dominance. There's only one way to do that, and that's to be preloaded, and that frankly isn't going to happen. The desktop wars are pretty much over, anyway. I'm much more worried about getting real Linux on phones and tablets.
Gabriel wrote on 2010-07-31 23:10:
"Steve Ballmer disagrees with your statistics. He could be lying. Or not.”" Well, according to him, you should be using Windows anyway :) "Concerning W3Schools statistics, it is common knowledge that many Linux users spoof their browser OS to avoid incompatibilities with Windows centric websites, even on websites that claim, or are, browser neutral." Well, besides the fact that no one spoofs the UA, let me get this straight. If Linux usage has grown, that's because average Joe is now using Linux. Do you really believe average Joe is going to do something like that? Please. "My own statistic is that in the last two years I have had nearly two dozen Windows users SEEK ME OUT to ask me to install Linux on their computers. " Right. Also been hearing that for a decade, at least. Do you remember Summer of Worms (circa 2003) and all the evil malware out there was gonna push everyone to use Linux? Denial is not helping Linux get anywhere. Get real.
El Nerdo wrote on 2010-07-31 23:55:
@Adam I'm convinced about your argument that Ubuntu has NOT increased total Linux usage. You say then they have failed, because their purpose was to challenge Microsoft. Have you ever considered that you may be mistaken on this last point? What if their purpose was not what they stated on their Bug#1 but rather something else entirely different? What if what they set out to do was to destroy other Linux desktop distributions, which at the time they started was mainly Mandrake (Mandriva now)? In this they were almost successful. Canonical, not necessarily Shuttleworth, poured a lot of money essentially in PR promoting Ubuntu, sending CD's to anyone who asked for them, influencing the press, which very willingly took the bait, and generally sounding the marketing horn very loudly. I say that the money should be ascribed to Canonical and not necessarily to Shuttleworth, because the Charter given to it in the Isle of Man, where it is registered, does NOT require that the chartered company specify the sources of its capital. So we now have the situation that we don't really know who is behind the $ 100 000 000 supposedly invested by Canonical in Ubuntu. This amount of money would be indeed a very deep pocketful for a former financial advisor and broker like Shuttleworth used to be before he turned on Ubuntu. So, who in the world might want to dominate the Linux scene first on the desktop and now on the server *without increasing its market share* compared to MS and Apple while losing money steadily year after year? And for what reasons? Any thoughts about this, Adam? And well, other people too. If you answer with the one word "Microsoft", I'd think you would be, untipically, making things far too easy for yourself.
adamw wrote on 2010-08-01 00:06:
Sorry, I don't go for the conspiracy theories. I think Canonical hasn't been as successful with Ubuntu as they'd like or as they claim, and I think they've gone about a lot of things the wrong way, but I don't think there's anything malicious about their *intentions*. I also think they've done some things well and can certainly be talked into doing other things better in future, or else I wouldn't be trying. =)
Randy wrote on 2010-08-01 02:38:
Adam I think you have a strong case. Getting Linux numbers is like measuring global warming. There are small changes in noisy data. But taking your sources together they appear to confirm a slow modest rise. Although Ubuntu has not put Linux over the top yet, and they may or may not be where they would like to be, they help. Even if they may have the effect of displacing Mandriva in the process. I think Chromium OS can have a similar effect. Raise the total Linux numbers another percent or two and place additional pressure on other distributions. What may make more difference though is if Google succeeds in establishing greater expectations for security than MicroSoft provides.
Telic wrote on 2010-08-01 05:05:
Adam Williams' coworkers should keep in mind that, when he eventually departs Red Hat, he'll feel free to vomit on Fedora -- just as he now feels free to vomit on Mandriva. Adam is a leopard of sour-grape spots -- at least until Canonical offers him a sweet salary. As one Nathaniel Smith has said, "In the end, marketing Linux is a huge, scary, and danger-fraught undertaking. I absolutely celebrate Canonical for what they do to get Linux on the desktop."
adamw wrote on 2010-08-01 06:31:
Telic: you, on the other hand, could at least do me the courtesy of figuring out my name before you lay into me. I think Mandriva's a great distribution. I still run it on four machines. However, the company has lots of problems, some of them of its own making, and some not (as I've pointed out before, it's more or less impossible to make money the way Mandriva tries/tried to make money, thanks in no small part to Canonical). Marketing desktop Linux is indeed somewhat tricky. But why should Canonical get lauded for trying and not doing terribly well, when everyone else who did the same got mocked?
adamw wrote on 2010-08-01 06:34:
Oh, and contrary to your implication that I left Mandriva in order to get a 'sweet salary' at Red Hat, Mandriva chose to end my contract. I had no intention at all of leaving.
El Nerdo wrote on 2010-08-01 07:03:
I'm scrambling as hard as I can to get the *correct* reaction in on time. So, ¡Viva Shuttleworth! ¡Viva Gnome! ¡Viva Mono! ¡Viva Miguel de Icaza! Sorry, sorry, sorry, my bad: ¡VIVA UBUNTU! Conspiracy theories? No, no, no. Not me. Never. Just in case I just put on a BIG garlic garland on my tin-foil hat.
Telic wrote on 2010-08-01 07:17:
Adam Williamson huffs, "Marketing desktop Linux is indeed somewhat tricky. But why should Canonical get lauded for trying and not doing terribly well ..." You're in denial of Ubuntu's real-world effect on Linux awareness since 2004: http://tinyurl.com/2cxo4ze Also, the Linux that Dell preloads onto desktop PCs: Ubuntu.
adamw wrote on 2010-08-01 07:27:
Telic: Not in denial of anything. Look up, and you'll see my take on Dell (other distributors have had OEM deals before, and Canonical/Dell is going exactly the same way: starts out with a couple of systems and claimed plans for many many more, several years later, they're still selling two systems. Maybe three. Whatever.) Google Trends graphs tell you very little of any significance - I like how they don't even have a label on the Y axis - but assuming it's not logarithmic, try "suse, fedora, debian, ubuntu, mandrake" and note that the total of suse, fedora, debian and mandrake in 2004 comfortably outstrips any point in the Ubuntu line (or, for that matter, any subsequent total). Also note that the Ubuntu line appears to peak in 2008 and seems to be trending gradually downwards since.
Telic wrote on 2010-08-01 07:50:
@adamw About the Google search trend: How much spoon-feeding to you need? Click on their "Learn more" link. With Fedora as the "1.0" traffic reference, Ubuntu hits "2.55". According to Dell, fully 30% of their Inspiron Mini have sold with preloaded Ubuntu. That's a hefty percentage of a top brand. Your desperation chooses to scoff about the number of different models.
Gabriel wrote on 2010-08-01 10:25:
http://tinyurl.com/367m8sx Oops, I was wrong. Someone call Mr. Shuttleworth, tell him to close bug#1!
Fausto R. wrote on 2010-08-01 12:59:
I can tell you that at least in my country DR. the increase of adoptions of linux specially Ubuntu, have been from a couple of hundreds from 2006 - 2010 to several thousands. Specially because Ubuntu have made it easier for the masses, i've personally use gentoo/ubuntu/centos. But i can tell you that ubuntu, doesn't matter the reasons, has made live easier for the majority if not all of the new comers. Ps: It seems that you have some issues about it.. I'm just happy they are spreading the word FREE..
adamw wrote on 2010-08-01 16:05:
I know it's hard, but you really need to get some kind of sense of scale :). Hundreds, thousands, even tens of thousands - these numbers are *all* tiny. The context you need to consider is that there are over ten million PCs sold per *month*. Even by its own, likely inflated, numbers, it's taken Ubuntu six years to get ten million installs. Telic: I just clicked through the US and Canadian Dell sites, looking at the Mini 10 and Mini 10v Nickelodeon Edition systems. AFAICT, no default configuration of either is offered with Ubuntu. The number you cite is about the Mini 9, which they don't sell any more, and came around early 2009, when far more netbooks were being sold with Linux. It seems pretty indicative to me of Microsoft's success in killing pre-loaded Linux on netbooks that Dell don't offer it in any default configuration on the Mini 9's replacement...
adamw wrote on 2010-08-01 16:22:
By the way, if you search 'Ubuntu' on Dell's site, the third result that comes up is the Mini 10, with a list of operating systems that includes Ubuntu (8.04...), but if you click on it and click customize and run through the customization process, you never get offered a choice of operating system. The fourth link is an almost identical one, to the 'Inspiron 10', which comes up as a 404. The fifth result is the Inspiron 14, which again has Ubuntu (8.10) listed in the search results, but if you go to the page and customize the base configuration, you get a choice of operating systems which is all Windows, not Ubuntu. After the first result (a so-so desktop), I had to go to the seventh to find a system which actually comes with Ubuntu - a chunky, overpriced Latitude (business) branded netbook (the Latitude 2110 - $420, thankyouverymuch). Clearly, Dell puts a high value on its Ubuntu relationship!
Telic wrote on 2010-08-02 17:02:
Adamw, it takes a toxic case of sour grapes for you to scrounge for ways to fault a world-class brand, Dell, who can claim to have sold a whopping 30% of a mainstream consumer PC with preloaded Ubuntu -- and who continue to find ways to make the relatively tough sell for Linux (versus Windows). Your pettiness is a PR disaster for Linux's potential in the bigger PC landscape. Methinks Mandriva is lucky to be rid of you. Methinks Red Hat could one day feel the same.
adamw wrote on 2010-08-02 17:43:
telic: you seem to have a highly inflated sense of both my and your importance. As it says right up at the top of the post, this post has nothing to do with Red Hat. I'm just a dude on the internet with some opinions.
Telic wrote on 2010-08-02 19:44:
Yet, in your blog, you use "we" when speaking of Red Hat agendas: "Red Hat, also, has never said we're aiming to make significant inroads into general user desktop market share ... We maintain a great desktop distribution, which I work on." You thus leverage yourself as a brand-name distro insider, associating Red Hat credibility with your personal publicity.
devski wrote on 2010-08-03 15:29:
i'm not suprised about the results, and this will stay that way for years to come. i mean for every one converted linux user, another linux user will leave/return for windows or mac. and in that small pool of linux users most will be distro switching.
PeteF wrote on 2010-08-03 15:38:
The fact remains that almost all computers come bundled with some version of windows. If someone wants to run Linux, it's usually install-it-yourself, and the vast majority of computer users don't have the knowledge/skill/cojones/whatever to do this. Perhaps it's possible to fault Canonical for not getting more preloads, but it's more realistic to point to microsoft's monopoly power. I had hoped years ago that the antitrust litigation would break this stranglehold, but it didn't happen. I don't know what the answer is.
adamw wrote on 2010-08-03 17:20:
Actually I agree that it's a very tough nut to crack, and it's hard to blame Canonical for not managing it, exactly. But the thing is that even though it's a high standard, it's the standard Canonical set itself, with all its talk about succeeding where other distributors failed in breaking the Windows stranglehold. You're exactly right that the best way to do that is to crack the OEM market, so if Canonical isn't managing that, surely it's reasonable to point it out, even though it is indeed a very difficult thing to do?
DH wrote on 2010-08-03 16:20:
From my perspective, there are a couple of things standing in the way of a widespread adoption of Linux on the desktop, and neither of these reasons is that people "like" windows. 1) The typical user is so technically inept that they will just use "whatever it comes with", 2) Related to point (1) is that it *comes* with windows. Even for hardcore linux users, you want the machine you want, you pay the windows tax, and wipe it clean after you bring it home. The big question to ask isn't "what is the windows:linux proportion?", but rather "what is the windows:linux proportion among those with an average to high degree of technical aptitude?" Now in some cases, the windows tax is refundable -- I know that ACER has this policy. In most cases, this is simply not an option -- you either take it how it comes or don't get what you want. Of those cases where there IS the option, what proportion of those who DO ditch windows ACTUALLY APPLY for the refund? --- sales numbers of new computers vs number of windows licenses is often a metric used to count the proportion of windows, but for aforementioned reasoning, this metric is totally invalid, as, it happens, is every other metric used to make this measurement for various other reasons. As it stands, both MS and the hardware vendors are singing this song about it being too difficult for the common user to install an operating system on their own, hence it has to come pre-installed for their limited intellect. This is something I ****VERY STRONGLY**** disagree with. An operating system can be installed by putting the disk in the drive, turning it on, and continuously clicking the "next" button until its done. Well at least this is the case for most major Linux distros -- I will grant that the process of installing windows is extremely confusing, even for an engineer -- and good luck with getting all the drivers sorted out. If you can't install the OS, then you won't be able to figure out how to launch a web browser (or whatever) and the machine will be useless to you either way. It is pretty obvious that the hardware vendors are getting a significant kickback/discount for (a) preinstalling windows, and (b) ONLY offering windows. Given the choice between paying retail for a copy of the latest windows for $300 US or paying $0.20 (the cost of the blank DVD) for the latest Linux, I think its pretty safe to assume that MOST people would opt for the $0.20 option given that they could always go back to the store and drop the $300 later on. I expect at some point that the EU will eventually mandate that general purpose computers be shipped in a "virgin" state, or at least that the buyer be given the option to choose at purchase time -- in the same manner as they are required to be given the option to select a web browser to install. Sadly, the US (and as a result Canada as well since up here we copy everything the US ever says/does) is unlikely to ever implement any kind of restrictions on MS since the governments are more interested in maintaining the strength of monopolies than to allow any kind of actual competition. It is, of course, the "average american" who is to blame for this since they seem to be WANT to have someone else THINK for them as long as they get a 12-pack on the weekend and can watch sports in high definition (infested with DRM, of course).
adamw wrote on 2010-08-03 17:21:
"Given the choice between paying retail for a copy of the latest windows for $300 US or paying $0.20 (the cost of the blank DVD) for the latest Linux, I think its pretty safe to assume that MOST people would opt for the $0.20 option given that they could always go back to the store and drop the $300 later on." False choice. The choice is between $0.20 and $0.20 - the cost of Linux and the cost of pirating Windows. (You can bet that if the EU did do such a thing, the anti-piracy controls on Windows would suddenly, mysteriously, become much easier to break, again). Aside from that, though, you make some good points indeed.
Gabriel wrote on 2010-08-03 17:43:
"From my perspective, there are a couple of things standing in the way of a widespread adoption of Linux on the desktop, and neither of these reasons is that people “like” windows." Of course. Few Linux apologists seem able to accept that lack of success may have to do with deficiencies in Linux. Your whole argument seems to be that Windows succeed because users are stupid. That lefts unexplained why technically minded people capable of installing an OS on their own find more attractive a RC of Windows 7 than any freely available Linux distribution, but who cares... at least denial gives peace of mind.
Telic wrote on 2010-08-04 03:27:
Adamw said, "it's the standard Canonical set itself, with all its talk about succeeding where other distributors failed in breaking the Windows stranglehold." Really? Post a hyperlink to where Mark Shuttleworth or other Canonical executive makes such an assertion. If your irrational angst is incapable of seeing *whimsy* in Ubuntu's Bug #1 -- "Microsoft has a majority market share" -- then try to recognise that the bug status continues to be listed by Shuttleworth as "In Progress" and "Critical."
Telic wrote on 2010-08-04 07:41:
Adamw guffs, "It seems pretty indicative to me of Microsoft's success in killing pre-loaded Linux on netbooks that Dell don't offer it in any default configuration on the Mini 9's replacement..." That's either a boldface lie or uncommon incompetence on your part, since the Mini 10 is indeed available from Dell with preloaded Ubuntu... http://tinyurl.com/ya7rkb4 For someone who's trusted with QA at Red Hat, there's nothing of either Quality or Assurance in the poor example that you set for personal and professional integrity in public forums.
adamw wrote on 2010-08-04 07:58:
It says right up there in the comment what I did. I went to the main dell.com site, I went to the Mini section, and I searched for 'Ubuntu'. This seems like a reasonable way to go about things. It seems odd that the systems available through that page wouldn't show up through any of those methods.
adamw wrote on 2010-08-04 08:00:
BTW, the *only* way you seem to be able to get the 10n is to click the 'Personalize' link on that page. Even if you click on the system picture, it takes you to the regular 'Mini' section of dell.com, where the 10n is not listed or available. I have no idea why Dell would set it up that way, but there it is.
Telic wrote on 2010-08-04 08:59:
Adamw weasels, "It says right up there in the comment what I did." Your failed procedure obviously wasn't effective in averting your faulty conclusion and subsequent accusation. Good thing that no one needs to trust you for competent "Plan, Do, Check, and Act." Oh, wait, Red Hat hired you for QA. The horror! :O
Gabriel wrote on 2010-08-04 09:08:
And when arguments fail, resort to insults and personal attacks. Funny, coming from people who are allegedly fighting for "freedom".
phillip wrote on 2010-08-04 11:39:
Telic, please stop trolling here. If you make your own blog, you can troll there without annoying to many people.
Telic wrote on 2010-08-04 14:14:
Meanwhile, Adam Williamson's "arguments" misrepresent Canonical as insidious hucksters, simply because they don't disclose exact sales and fiscal figures. Spontaneous mud-slinging for one and all! Where's the hyperlink to substantiate Adam's accusation of Canonical claiming to have succeeded "where other distributors failed in breaking the Windows stranglehold"?
adamw wrote on 2010-08-04 15:28:
And, as Greg put it, scene.