Since it’s nigh on impossible to produce accurate figures of operating system usage, we have to make do with figures that provide a rough estimate, at best. One such set of statistics are the figures from Net Applications, which tracks the 160 million monthly visitors to its hosted websites. The latest figures from January 2009 have been published, and they show that the rise of Mac OS X continues, as well as that of the iPhone and iPod Touch. Unsurprisingly, Windows 7 did quite well too.
The Mac OS X consolidated its strong market position, now having 9.93% of the internet traffic as measured by Net Applications. The iPhone grew from 0.44% in December 2008 to 0.48% in January 2009. The iPod Touch did even better, growing 37.5% to 0.11%. With Apple doing so well, it’s no miracle others are hurting.
The biggest loser here is Linux, which dropped 3.53% from 0.85 to 0.82%. The other loser is Windows, which dropped to 88.27%, from 88.68% in December. Windows 7 made a decent impact here, rising from, well, nothing, to 0.23% on January 31st; not bad for a beta release.
The below schematic shows all the important data in one view:
As always, these figures come with a lot of ifs and buts, so take them with a grain of salt. Still, some figures are better than no figures at all.
Not seeing any figures for Windows 7 in particular in the linked page.
Oh, wait, here it is:
http://marketshare.hitslink.com/windows-7-market-share.aspx?qprid=4…
So 0.22% actually use Windows 7 _beta_? Wow.
And this is huge win for windows according to article because month ago it was 10 times less. There was no free windows 7 back then, but who cares windows win because their name begins with win that only matters.
Go read it again, sonny. There is no mention in the article ANYWHERE that “Windows wins”. All it says is that “Windows *7* wins”, which is *perfectly* accurate.
Microsoft distributed windows 7 for free and gained few more windows 7 users than it was in closed beta? What kind of win is this? EPIC FAIL. Almost everyone better buy MAC OS X than download windows for free.
p.s. Here I look into numbers as they are true not time-money waster idiot job as they really is and make my conclusions from them. Not windows 7 wins because I wrote so in other place. I do not accept this.
Edited 2009-02-02 18:32 UTC
And heavily seriously manipulative, If the title had said Windows Loses that would have been correct.
I like linux and it was my main OS before Os X.0 was released but certain thing I feel still stand in its way.
1. Migrating from a “competing” OS should be easy. Window users jumping to Mac have a far more seamless migration path IMHO. e.g. MS office is available (or you can use iWork, or OO). The user experience is more consistent across the platform. Other propriety software is also available for both. Whether its adobe, MS, or more niche products. Apple, it seems to me have spent a good portion of effort in removing barriers to adoption by making sure Apple plays nice with the Windows world and the external device world. I can’t remember the last time I had to install a driver for anything on OS x.
2. whiles linux is a great OS, it is not MARKETED. I said it before i’ll say it again, people are willing to pay what they perceive a product to be worth. If an OS that is FREE is not being used, the problem is either that people don’t know about it, or that switching to it is too painful (or both).
Needless to say this doesn’t apply to people on this forum or indeed their immediate families who get tech support!
True. You could also add “lack of time”. It is always possible to learn how to use Linux when you have enough time, but when you have a Windows computer at home, and you virtually have no time to google around, read documentation, discuss on forums, etc…You stick to what you have, even if you are not 100% satisfied. When you have a problem, you ask a friend, or take the computer to a local store, they fix it for you. Personally I have given up with Linux because I don’t have enough time to learn and troubleshoot the computer (because of my work, family, sport, charity, etc…).
Edited 2009-02-02 12:59 UTC
> I can’t remember the last time I had to install a
> driver for anything on OS x.
While this obviously stems from OS X being marketed with special (known) hardware, I am still puzzled why nobody tries that route with Linux. Choose known, working hardware; customize a working Linux distro like Ubuntu for exactly that hardware (and nothing else); slap a careful choice of applications on it; sell it.
Proprietary parts can be included since, unlike free-to-download distros, paying for licenses isn’t a problem with a product that is sold for money.
You just described the business plan of all those Linux HW startups a decade ago.
That’s basically what Apple did with BSD. Darwin is basically BSD with the Aqua GUI on top. Apple could have chosen to use Linux, but I think the licensing for BSD suited them better.
The technology suited them better, too. Can you imagine Apple trying to develop an OS like OS X on a kernel as volatile as Linux?
There is no reason why Apple couldnt have forked the Linux kernel and done their own thing. The only real reason why they didnt is that they wanted more control over what happens than the GPL allows.
I would think they based it on BSD as NeXT was based on BSD before Linux existed.
There actually IS at least one valid technical reason for Apple not choosing Linux: Linux kernel lacks stable API, so 3rd party binary drivers would be a huge problem!
Aren’t drivers done in a separate part of XNu from the FreeBSD bits? I thought DriverKit was part of the Mach side of things … but may be mis-remembering things.
Wrong. Darwin has no Aqua GUI on top. OS X is far more than Darwin with Aqua GUI on top.
The massive collection of Cocoa Frameworks, the various parts of BSD, non-BSD, and more that add to their custom kernel and much more is OS X.
The additions of OpenCL and its Dev APIs for Cocoa and more, plus other additions to CUPS specifically for OS X, etc., not in Darwin is another reason people haven’t just slapped X11 on-top of Darwin and thus presto! We’ve got OS X.
People who want Darwin with X11 also want Darwin to have all the QuickTime APIs, Xcode, Cocoa and more, but don’t want to pay for it.
That’s life.
“I am still puzzled why nobody tries that route with Linux. Choose known, working hardware; customize a working Linux distro like Ubuntu for exactly that hardware (and nothing else); slap a careful choice of applications on it; sell it.”
Exactly!
Isn’t this exactly what Asus and Acer are doing with their netbooks? Putting together a bundled system, working hardware and software, and using Linux as the os? Granted, I wouldn’t have used the distributions of Linux they did, but still, the concept is the same.
indeed it is. And it appears to be working!
I may be a mac fanatic but I want to see more linux machines out there. Long live the competition.
Just to add a point though, I would like to see someone throw out the rule book a bit more. Apple created a new layer(or adapted a combination of two existing ones) on top of BSD on selected hardware. It would be nice for someone to do the same with the linux Kernel (throw out KDE and Gnome??).
However, as a previous poster mentioned, the investment required to do so is substantial to put it mildly.
I would like to see Microsoft pull an Apple, throw away their NT Kernel and take a BSD or Linux Kernel, and rebuild their most recent APIs around that.
Staying with the UNIX principles of “everything is a file” and the Unix standard permission (rwxrwxrwx) scheme would be a good thing, as the windows permission system nowadays is not as easy to understand and as flexible.
For Backwards compatibility they should use virtualisation, not a big deal nowadays.
There have been dozens of companies that have tried that, few of them might even still be around. The basic flaw in the plan is that they are too small to compete on price and most people who want a dedicated Linux box have the necessary skills to install Linux on whatever hardware they happen to own. The vast majority of Linux users aren’t willing to pay a $200-300 premium simply to get Linux pre-installed.
Because the only people who have so far come up with this idea have been geeks and enthusiasts – neither one of them are willing to spend the time and money on the spit and polish required to get it to the same level (and exceed) that of Mac OS X.
What is required is an alternative operating system whose vendor maintains and integrates everything so strongly that one cannot view the separation between the different components; the whole operating system is viewed in its entirety rather than being viewed right now as a collection of things hobbled together and put in an iso image.
What it would require first and foremost are drop in replacements for applications; you will need to have an Office alternative, a Creative Suite alternative, a Photoshop Elements alternative (for basic photo touching up) plus many other alternatives (MYOB anyone?) All of these are epic failures in the open source world.
Now, I don’t blame the open source world because the basic foundation of open source is to scratch and itch. If the itch of that individual (or group of) developer(s) isn’t the same as an end user – can we really blame the open source developers for not meeting the needs of end users?
What is required is a for profit entity who has paid developers and are able to meet these requirements by paying developers to work on issues that might not necessarily be issues they find sratchable but because they are being paid – they’ll address them anyway.
Btw, as a side issue, a couple of years ago I tried to setup something similar to this; the ‘ministry of economic development’ was setup to foster these sorts of ideas. I went to talk to someone there to find if I could get VC funding or some sort of grant. Lets put it this way, they have thousands working in the ministry and the best they could do is refer me to a rack of pamphlets. Unless one can find long term investors who look at the long term gain rather than short term profit, ideas such as what you have suggested are never going to get off the ground.
Edited 2009-02-04 03:40 UTC
I believe it is only marketing. Apple got it right, so people change even it is the most difficult to use OS.
Yes in my opinion it is difficult. Mac is very different from Win and Linux in Look&Feel starting with keyboard layout. Some trivial task like editing text ASCII files are available only from terminal and in general there is not a lot software included with OS by default compared to Win and specially Linux.
The fact what Linux is free is not true from user point of view (my Linux netbook was more expensive with Linux compared with WinXP), or not relevant (I never saw Linux source code even I’m programmer myself and I use it for years now).
So the close BSD distribution (yes MacOSX) is much more successful.
Open TextEdit, select ‘Make Plain Text’ from the Format menu, or open an existing ASCII file itself and it will be in plain text mode. You can select the encoding when saving.
Is there really less software included with OS X than with Windows? Mac hardware comes bundled with whatever the latest version of iLife is at the time too (iPhoto, iMovie, Garage Band, iDVD, iWeb), but the OS by itself comes with a full IDE and associated development software, in addition to all the base applications you would want – eMail and contact manager, calendar, phone sync, browser, media player, media organiser, instant messenger, document reader that can handle PDFs, etc, as well as a bunch of the core Unix utilities.
Edited 2009-02-02 13:31 UTC
To some extent, yes. Keys to access areas of the screen such as the menu bar are different, but the majority of keystrokes used by average users are the same. The only habbit they really need to change is to use the command key rather than the ctrl key for a good number of tasks. Yes some are different (the return key in Finder comes to mind), but you get that across every os.
Um, in what reality are you living? I can open just about any file I want, either with the built-in Textedit application or with any number of third-party text editors (Smultron being my personal favorite). There’s no need to drop to the terminal for this if you don’t wish to.
It has an Email client, address book, Calendar, media player, Web browser, Rich text editor, and IM application, not to mention all the other little utilities. That’s not much different from Windows, which doesn’t come with all that many apps in its default configuration either. Most OEMs pre-install a lot of other apps ontop of Windows, but that doesn’t mean Windows is bundled with those apps. Even Microsoft is going this route, removing Windows Live applications from Windows 7 and having the end users install them if they want them. This actually leaves OS X with the better selection of bundled apps. I can’t argue with Linux distros coming bundled with the most software, however, that’s definitely true. Whether that’s a good or a bad thing, however, depends on your personal tastes.
Edited 2009-02-02 13:44 UTC
Well, perhaps it’s more consistant across osX that other platforms but MS Office is inconsistant across it’s own program suite let along across the entire Windows platform.
Though not the best, it is the new user favorite so I’ll offer Ubuntu as an example also; it’s mostly consistent and as a preinstalled OS it “just works” like any other OS preinstalled by the vendor who’s already smoothed out the edges.
This entire “my daddy OS can beat up your daddy OS” discussion is hot air being blown around yet again; but then, that’s a weekly occurrence (not picking on osnews, it’s a weekly thing everywhere).
That’s exactly what Novell is doing. Novell gets lots of bashing for its interoperability deal with MS but it’s not that differnent from the deal(s) Apple has with MS. The major difference is that Apple sometimes gets original source code from MS to port (Apple’s MS Office Documents Spotlight importers contain MS Office source code), while Novell mostly just gets documentation that allow Novell to release the self-written code under a FOSS license. In the end the whole FOSS world benefits, as seen with the adoption of Novell’s OpenOffice in Debian, Ubuntu, etc. Heck, thanks to Novell I was able to open and save OfficeXML files on Mac OS X through NeoOffice (an OO derivate) before MS Office 2008 was released.
I’m not a fanboy of MS technologies (like Miguel de Icaza), but there are users — especially in the enterprise segment — who require as much MS compatibility as possible, so it’s good that the guys and gals from Samba, OpenChange, Apache, Novell, WINE, etc. help to write the needed code.
Edited 2009-02-02 14:08 UTC
“I can’t remember the last time I had to install a driver for anything on OS x.”
Seriously, I can’t remember that I installed any driver in Linux, too, since about 2004. Before that, I remember that in 2002 I even had to recompile the kernel to get specific drivers I need.
If you don’t buy cheesy hardware, it will run out-of-the-box with the Linux kernel. It will even be auto-detected on boot.
So there has ben lots of progress made there. Still, you have to somewhat chose the hardware you buy (instead of Apple doing it for you..). But unlike some years ago, that’s not a big challenge anymore. There are also many really good laptops out there which are fully supported by Linux. Take Thinkpad machines for example.
I think it is interesting that in several usability aspects, Linux distributions are already far ahead Windows, comparable or even better than OS X. Mainly hardware and software installation, system transparency. I don’t want to say this makes a GNU/Linux based system the better choice for “the” regular user, but that there are a lot of things that would comfort these users, but they cannot see their value before fully getting into it. On OSX you still have to find&install application themselves and get updates for them as well.
Edited 2009-02-02 16:02 UTC
Software installation on Linux is only ahead of Windows/Mac if what you want is in your distro’s repository. Otherwise, it’s about as far behind as you can get, requiring more manual operations (usually on the command-line) and with each piece of software being just a bit different. Don’t even bother trying to uninstall software installed manually unless you kept track of exactly what files it placed in which directories, you’ll probably not find them all otherwise. It makes the uninstallation on OS X look like a snap by comparison. And don’t even get me started about driver installation if you need a driver that is not part of the mainline kernel tree, and your distro of choice has merged some patch into its kernel that renders the driver you need unable to compile. Yes, it happens, and while I’m bringing it up, why should we even need to compile drivers ourselves anymore if we’re not the ones developing them? This is what Windows, OS X, and Solaris have correct, none of this “driver must exactly match the running kernel version” nonsense.
Yeah yeah yeah, it’s the hardware vendors, blah blah blah. I fully support open standards, and I’m aware of the situation. But it doesn’t matter to 99% of the computer users out there, face it. All they want is for their brand new, shiny peripheral to work, possibly after using the provided driver disk. If it doesn’t, it’s broken, and the philosophical and political issues behind the reasons are irrelevant to them.
“But it doesn’t matter to 99% of the computer users out there, face it.”
I claim that 99% of the users out there don’t need any drivers not packaged by their distribution of choice. That’s exactly what my posting was about. Things have rapidly changed in the latest years.
Same goes for software. Even with ArchLinux, a really small distro compared to the big players, I get all the software I need through the two official channels (pacman and aurbuild, which is well automated, too).
Btw. just a hint in case you may need it later: All programs built with autotools don’t only have “make install”, but also “make uninstall”. But you have to keep the sources directory for that.
Who cares if there isn’t any money to be made.
Most people tend to forget that “Linux” isn’t a corporation. “Linux” goes wherever the companies that direct Linux want it to go. And if there’s millions to be made in the game Red Hat is playing, and virtually nothing “on the desktop”, then guess where Linux is going.
This has absolutely nothing to do with marketing. This has everything to do with creating a real product.
Linux on the server side is holding its own. Yes there are promoters like IBM, and Redhat, etc. But the fact is that Linux on the server side makes a whole lot of sense. Linux on the client side makes NO sense.
3 years ago I said Linux on the desktop was dead, and it did not get anywhere.
Though now I actually have some hope. Namely in the form of Nokia. Nokia is doing an LGPL with QT, which means that now people can develop closed source apps for KDE. That will make a huge dent.
KDE always was better than GNOME, but the stupid GPL licensing has always held it back. On top of that with Trolltech charging several thousand nobody was interested.
If there are more users in december 2008 than in january 2008, then it’s an increase, not a decrease. It seems the whole graph is wrong, inless years are swapped or something.
You’re right, typing error there. Shift+reload, fixed it.
Thanks!
Soo, you can^aEURTMt prove any wins, or loses, yet you^aEURTMll report news about it anyway?
Brilliant.
That’s not what I said. Statistics are aALWAYS debatable, but that doesn’t mean they don’t provide an insight. These figures clearly show that the general trend – the growth of Mac OS X/iPhone at the cost of others – continues even after the holiday season. Even when numbers have an error margin, they can still be useful.
But sir, if you want us to only report on matters that are 100% accurate, well, be my guest – it will mean I have a lot more free time on my hands, because I think in that case OSAlert would only be able to post 1 or 2 items per month – at best.
Edited 2009-02-02 13:24 UTC
The iPhone growing in the mobile phone market would be fairly accurate. It’s the hot sexy toy that all other mobile phones are benchmarked against right now. That doesn’t relate to the rest of the market shares unless we can determine how the overall market has grown at the same time though.
The other problem is that it does not measure use; only retail success. retail and marketing success measures the companies ability to spin and spend money neglecting the actual product’s qualities and application. Windows is not technically the best OS available nor is it applicable to many of the places it’s used but we can agree that Microsoft is an increadable marketing and business strategy machine based on results from shady deals and “mind share”.
I am grateful for the deluge of OS related articles on the site though. My pessimism is just expecting this to be the usual “I like blue” vs “I like red” arguments between posters.
Self-fulfilling prophecy, sir.
The more you mention “Oh this will surely devolve into […]”, the more likely it is to happen.
Edited 2009-02-02 13:54 UTC
That is the conundrum. The topic itself causes the discussion to be an argument of opinions versus facts but I figured mention it and hope it helps guide the rest of the conversations in a productive direction even if my own threads devolve versus say nothing and watch as all the threads devolve away from any productive information sharing.
We’ll see how it goes though. Either way, I’m sure it’ll be a high post count by the time everybody moves on.
You take a trend from a month not a year. Without accounting for seasonal changes etc etc
At best, the figures from webservers present what the browser chooses to report. At worst, they are a complete waste.
When comparing retail only OS, the best you have is what the company chooses to tell you; only Apple knows what osX actual usage is and only Microsoft knows what Windows actual usage is.
– how do you account for activation and ongoing updates to measure actual use of the OS since only Apple and Microsoft know what those figures are.
– how do you account for OS licenses imposed on consumers but replaced before or soon after first boot.
– how do you account for multi-boot systems or does that use get counted under each brand name?
But it gets more complicated; when comparing all OS in the greater software world, you have to separate the various Linux based distributions and BSDs since each one is a separate OS that happens to use similar lego pieces. Red Hat != Mandriva != Debian; they are distinctly different though very interoperable.
– how do you account for freely downloaded installs
– how do you account for freely shared .ISO and disks among friends
– how do you account for multiple distributions on the same machine (I have two on the hardware plus a long list of VMs.. does that mean my three Debian VMs adds three to that count?)
As your using webserver log statistics:
– how do you account for browsers not reporting correct brand, version or OS? (lots of IE only sites work fine with Firefox if FF tells the webserver it’s IE)
– how do you account for multiple browsers open on the same system
– how do you account for browsers open across VMs (again, Adobe’s incompetence with 64bit Flash means I have to use a win32 VM to see the crappity animations and youtube videos; I’m unwilling to dirty my 64bit install with a 32bit browser)
– how do you account for false reports from proxies and routers along the network path
– how do you account for multiple varous OS running through NAT’d gateways
– how do you account for html browsing wrappers used to get around firewall restrictions
– how do you account for multiple hits since each url request shows as a separate “visit”; a page with three images means four “hits” counted against one user/machine
Then let’s expand the discussion a little; are we counting all versions and instances of Windows, osX, BSD, Linux and other more custom kernels? osX gets the Apple servers, desktops and notebooks plus mobile phones plus ipods. Windows gets servers, desktops and mobile phones plus a few embedded devices like car “computers”. BSD gets servers, desktops, routers, gateways, switches… Linux gets servers, desktops, routers, gateways, switches, RC radios, car computers, mobile phones… Or are we just cherry-picking a select few usages to insure that the figures tell our pre-decided story?
If the article only compared retail only OS usage statistics then within that comparison, it may be a closer to real estimate but comparing usage statistics from something as untrustworthy as webserver hit logs.
Since accurate figures from any platform are not forthcoming, this is just another fluff piece about how one colour of spoon is somehow better than another colour of spoon.
I run a pure 64bit Fedora 10 on my March ’08 17″MBP. The 64bit Flash 10 plugin from Adobe works perfectly well with 64bit Fireffox 3.0.5.
http://download.macromedia.com/pub/labs/flashplayer10/libflashplaye…
RE: The Thom’s article, I think a slightly paraphrased quote from the genius that was Robert Nesta Marley sums it up nicely.
“You can please some people sometimes, but you can’t please all the people all the time”
Cheers,
If adobe has finally released the 64bit Flash player then I can pull that back off my list of limitations. I’m off to check into that now. Thanks for the head’s up.
Cheers and Happy Monday
+1 for quoting a quote (correctly) that was used in a dylan song
Try using a 64-bit version of gnash or swfdec (I use swfdec) and can see all Youtube-vids and with the lastest version also the HD-ones.
Does it not bother anyone, someone trots out these “stats” every so often…. yet never buys or publishes the more detail stats?
like a google global or for that mater even publishes there own local stats?
Lack of Linux refinement. We’re just getting now to where Microsoft has been for years. nouveau, open ati, hal/dbus are all brilliant projects. when nouveau/all ati cards have open drivers IN kernel with 3d support and randr 1.3 the desktop will be a lot easir to setup. There are still some application holes remaining but they are getting closed. The big problem with convincing someone to go Linux is why would a person want to run Linux? The average person doesn’t give a crap about half the issues we think make linux awesome. They want to throw a bluray disc in the drive and have it play back menus and all. Linux can’t do that at the moment. (I actually have a bluray drive hacked the firmware and run movies in linux it’s a PAIN IN THE ASS have to decrypt and feed the vid into mplayer with a pipe to play off the disc, else I have to dump to my hdd.) anyway the point is that it has to just work. with EVERYTHING. Every little detail. We have to aim for Mac as a minimum in terms of ease of use. Really we need to aim higher again. We need maximum interoperability. More wine work (those guys are brilliant and the current progress of wine is amazing.) native apps are great, we need more of them and in different fields. I’d like to see more full featured 3d modellers than just blender (I am coding one at the moment so don’t ask me to put my money where my mouth is) Game developers need to stop writing engines. We’ve got 50+ engines now that are high quality such
as Ogre and Vegastrike and Crystal Space.
Ten years ago we started writing vegastrike and ten years later we’ve got a good engine and some good art but we’re stalling due to a lack of tools for telling stories with.
Most game developers don’t want to learn python. They want to click to create a waypoint and click on a ship and select what they want it to do. Fred/fred2 (freespace editor) are good lessons in design for budding game developers. What the games really lack now are tools like level editors and mission/star system editors. Make Linux games have point/click mission creation like freespace 2 had on windows and I guarantee a large gaming community will form around them.
It’s all about making people’s lives easier. One of Linux’ big strengths is it can leverage it’s various parts to create things other platforms can’t do as easily or as well. Things like XBMC and mythtv are brilliant. No one installs XBMC thinking I want to run XBMC. They install it thinking I want to watch a LOT of movies and tv shows and I want it all managed for me in an easy to view format.
People think about results not how they get them. At the moment Linux contains a bunch of well engineered tools but now they need to be given purpose and focus. Take compiz. It looks great, but can you really justify using it? how does it make your life better? Now if you use compiz effects to bring user attention to notifications and help them focus on their primary work window that’s a good use. More work needs to be done to figure out howto exploit existing tech we already have to be useful.
We need to push R&D and development of the platform beyond what mac and windows can do otherwise we’ll always be stuck in their wash. Apple created OpenCL.
What’s Linux doing to promote an open audio api standard among vendors? Where’s the Linux tech to do advanced processing on multicore/cpu/gpu systems? At the moment we’re letting Microsoft and Apple lead the market, and that’s why they’re leading the marketshare.
They’ll start making inroads into traditional Linux strongholds soon if we’re not careful. Stuff like OpenCL makes Apple into a very legitimate platform for multicore parallel scientific work which is a traditional Linux area. Microsoft is standardising their BDA driver architecture which will cut into using linux as a media center (Media portal/windows media center is already doing this). Why have a linux media center that supports most formats when a windows one supports all the formats as well as bluray AND plays all your games.
My background is gaming and game development as well as media center pcs. I’ve got 3 hdtv tuner cards I run linux and I have about 500+gb of games installed Most of which are Windows based (run in wine, I have not had windows installed on my pc in about 8 years although I do keep a copy around on my laptop). I do have at least 50gb of Linux games installed. The platform in my opinion is not as bad off for gaming as many people like to moan and whine about. Almost all of the Lucasarts’ titles run. Most EA stuff I’ve thrown at it loads. Dos is 99% working via dosbox.
Things like there is still 10+ years on no tool for configuring surround sound and speakers are big hinders to adoption. Although Ubuntu looks like they’ll finally fix it. For that matter in the last 3 years the only distro that seems to have done anything useful for the desktop has been Ubuntu. I don’t use Ubuntu but the applications that have flowed out of that project have been high quality and numerous. More people need to be taking a page out of their book. It’s not that Ubuntu/freedesktop.org is the only group working on making Linux easier to use but they are the only group that seems to be creating standard open tools to edit system configuration.There are too many distro specific tools to setup video and audio and it’s a mess to figure them all out.
(btw is anyone interested in writing libbluray? the encryption has been totally cracked now it’s just all the code for it is in java
doom9.org forums have the sourcecode).
Edited 2009-02-02 14:44 UTC
what does that have to do with those questionable “surface” stats”
MS spends 1B marketing & 6B R&D a year?
And yet in that 1080p TV/monitor is either Tron or Linux?
the point being it is the same 1’s & 0’s
just packaged & marketed (or not) differently?
and like many/most of the salacious headlines & intro copy,there is much more or different info…. if one looks?
Just saying ?
Did not know nouveau did anything with ATI/AMD?
But works well with the nVidia cards I have?
But have you worked/played with Neuros stuff?
there is even a bounty? see
http://www.thinkgeek.com/electronics/home-entertainment/b4ba/
and/or http://open.neurostechnology.com/content/first-link-bounties
anyway just saying ?
why no netbooks,MIDs or gphones or whatever?
I’m all for praising Ubuntu/Debian, but you shouldn’t forget Fedora, a lot of what is in Ubuntu also comes from Fedora.
“some figures are better than no figures at all”
A nonsense saying. Do you want to actually _use_ something or do you want to _sell_ something. No figures are better then lots of fantasy.
You can’t really count PlayStation and iPhones in the stats.
Do you honestly think these visitors use these devices as their main computer?
In our house we have Macs/PCs, but the kids also browse with Wii, PSP, iPod Touch.
I think so, especially among the youngest internet users. Around here few kids have ‘real’ computers, but nearly every one of them has a cell phone and/or game system.
Those users’ stats were so small as to be irrelevant anyway.
Edited 2009-02-02 15:55 UTC
Man open eyes “The biggest loser” is Linux it lost 3!!! users because they used their Iphone at counting moment.
That Apple writes a little code that triggers the Net Applications counter every so often? Everything “Apple” is doing well, this would be grate PR. Not sure how Net Applications has it setup but I can write a little code that would trigger the logs on Apache every so often.
I’m not an expert in statistics, but could it be that the sheer amount of internet users has grown?
That’d account for the “decrease” in Linux users, for example, since a 3% decrease is way too significant over a year (we’re talking big numbers here, more than the people in my own country. And a 3% rise in unemployment rate over a year would be quite noticeable).
Edited 2009-02-02 15:04 UTC
I wasn’t even able to find the degrees in the statistics. Actually I found something else:
http://osnews.com/permalink?346638
is that 160 monthly visitors the correct sample size and not a typo? Because if it is correct, 160 is not an adequate sample size for something like this.
Only if it is truly random, in which case it would be ok. Its the randomness that is critical, raising the numbers will not help if it is non-random.
Non random sample reports are amusing starts to a conversation, but they are not data.
Yes but 160 means some fuuny things:
And HOW FreeBSD came up with 0.01. One Free BSD user come up there and generated those nubers. Maybe there is one lucky FreeBSD user in entire world?
I had statistic (biometrics) lectures in university and based on them I can say that these numbers mean NOTHING.
> Only if it is truly random, in which case it would
> be ok. Its the randomness that is critical…
The sample size is also important. If your sample size is too small, you won’t be able to get a meaningful P value, and this means you cannot determine whether the change is statistically significant.
Edited 2009-02-02 17:29 UTC
And from other side if sample size is 160 000 000 you will not be able even to input this number into computer statistics programs.
And if you do so huge sample size it means you are stupid and wasting time and money.
Statistics invented in XIX century just because we need method to find information about huge population in fast and cheap way. I looks like some big companies find XIX century math too difficult for them even in computer age.
Uhm, I’ve actually studied statistics rather vigorously at university, and what on earth makes you think a statistics program can’t handle 160m? My EUR 3 calculator can even handle that number!
Do you have any background in this subject at all?
you funny and know nothing more than how to find % value based on real numbers. How much RAM your calculator have if you can input 160 000 000 lines of data into it?
Ok calculate dispersion of 1, 2, 3, 1, 2, 3 …(160 000 000 times) with your calculator.
Supercomputer is needed for this stupid time wasting calculation. Not talk about this idiotic data harvest of so huge size even.
Edited 2009-02-02 18:10 UTC
You’re joking, right?
160,000,000 users visit some websites. 140,000,000 of those people used Windows. That means that 88% of the visitors were using Windows.
C’mon, a $1 calculator from WalMart could compute that.
OK lets go now to basics of statistics. Really i do not know where to start.
You must understand that data is random. Like gambling.
Some users visited site last month but not this month.
Some users sow nice article about win7 on same site and downloaded it after reading and changed their os for fow days.
Some users wanted to try how that site looks on their brand new PS3, but they never use PS3 for net really.
Randomness change results of statistics? OMFG So it is useless? NO! Mathematics 200 hundreds years ago developed methods how to interpret those results.
And for sure reasonable interpreting of data need them and they much more complicated that just dividing 2 numbers one from other.
And without this it is time wasting to say that I waked up this month 1% later so I become lazy. Maybe just one day alarm clock failed to work.
I hope you understand now.
You are free to ask.
Edited 2009-02-02 19:24 UTC
Actually its more complicated than that these figures are not accurately reported, they were massaged a few months ago and now Linux is about half the size.
That’s supposed to be 160 millions visitors
At first, I was surprised to read that too : who would care about statistics for 160 users per month??
And from where you decided that there is 160 000 000 users questioned? Just added some 0?
who would care about statistics for 160 users per month??
The ones who want fake results for their speculations. With 160 you can make “catastrophic” 10% change in month from 3-4 people opinions/actions. and if you do not like results you can repeat same with other 160 and get other results.
Edited 2009-02-02 16:41 UTC
If you cared to read that article, you would know. But if you feel more comfortable with 160 because of those many 0s which makes your eyes go round, hey… no problem.
I do not read articles full of lies.
“Windows wins” but gain is negative in numbers?
Pure manipulation of someone and quite stupid one.
Pure lies and speculations -> reading it will only prevent from thinking critically.
The headline says, “Windows 7” wins.
As you can see from the linked chart, Windows 7 DID make gains, just as the headline says.
http://marketshare.hitslink.com/windows-7-market-share.aspx?qprid=4…
Windows beta is distributed for free first time ever. Windows cares not only about hardware that will become released two years later.
Windows is trying to stop Linux while it is not too late.
But windows can not stop unstoppable. They trapped themselves inside lack of IE7 and MSoffice standards. Following standards now wold be MS fans back knifing.
Do not following standards ends up in angry of even more other open thinking people.
They trapped themselves in Vista API. They can not change it and broke things, they can not stay with same API 10 years.
All they do is muping smuping with GUI and selling same crap second time. EPIC FAIL.
Edited 2009-02-02 15:32 UTC
Well, I wouldn’t say this is an indication of operation system market share… rather it can be considered a statistics about Internet-connected operation systems. Not very indicative but fairly informative as 160 millions users is quite a good number.
Unfortunately, this doesn’t account for lots of never-connected or quite-rarely-connected systems which won’t show up in such list.
Also, it would be better to have more information about geographic location and market targets. For example, 9.something % is a good number for OS X if measured globally, not-so-exciting if measured for US market as we all know OS X is by far more relevant in US market than it is in other markets.
Same consideration can be done for Linux, which is known to be more relevant in areas outside US than it is in US itself.
Taken as aggregate and without more information, it’s hard to get a trend out of those numbers. Nothing new under the sun, I’d say…
http://marketshare.hitslink.com/os-market-share.aspx?qprid=9
For example: March, 2008 Linux: 0.61%, December, 2008 0.85%, January, 2009 0.83%. That would make a lot more sense.
Woops, it was indeed 160 million, instead of just 160, of course. It’s Monday for me too, people
.
So many more question came up then.
And who can confirm that really millions not some “ups we better add some 0 then”?
160 millions is huge numbers. Who paid for this tracking? It is not part time job to track hundreds of millions.
Well, uhm, gosh, I don’t know… Maybe Net Applications which gets mentioned all over the article and the linked website?
Just a guess, though.
And huge 30% gain in some kind of Misc. What that misc is? Looks like Linux is losing its popularity to totally unknown evil. lol How misc can gain something if there is even no public information what that misc is? Maybe aliens connecting to internet.
Or, you actually check out the link and find out that “misc” contains all sorts of small operating systems.
I want that list of those small OS much more popular than FreeBSD. Please post link here.
Edited 2009-02-02 17:37 UTC
Well, uhm, gosh, I don’t know… Maybe Net Applications which gets mentioned all over the article and the linked website?
Maybe they deserved world record for so huge tracking of people.
I also wonder about their methods how they claim they tracked Linux if Linux user is allowed to change his Linux as much as he wants.
Indeed and since they apparently does not document the method by which they collected the data and came to their conclusions the data is utterly worthless.
You count hits…?
Seriously though, the methodology is clearly listed on their frontpage [1]. Do I have to pre-chew everything in here?
[1] http://marketshare.hitslink.com/
is there a list of the web site hosted?
i would like to know the kind of web site analysed….
if a lot of web site is oriented windows… (tutorial, news, tips…) it’s sure that have an impact on number….
does google have os statistic?
because the number are a lot different then gartner group…
http://digg.com/linux_unix/GNU_Linux_Desktop_Market_Share_is_4_Gart…
That’s a good point. If there were even a single popular OS-specific website included in the survey, the total user percentages could certainly be skewed.
Similarly, since operating system popularity varies significantly between geographical regions, nation-specific or language-specific websites could also affect the results.
So IMO the most relevant aspect is the change trends.
Darkmage above makes some very good points.
I’m wary of surveys like these as predictors. I think they say more about where we are right now. It might throw more light on where we may be in the future if a survey tracked hardware rather than OS usage. What’s been noticeble in the past couple of years is the growth of laptops at the expense of conventional desktops, the sudden growth of netbooks, and the growth of smartphones on the web. If – a big if – these new developments in hardware presage the future, then the OS that will do best is the OS or OSes that do best on them. Joker in the pack not mentioned in this survey: Android and any forthcoming Google web/files appliance.
Anyone can adjust the parameters to make any stats say what they want to. Today, there is another page showing Win market share failing…
http://www.theregister.co.uk/2009/02/02/mac_marketshare_up/
Which is exactly the same as the item reported above. You did read it didn’t you?
Mac gaining ground, (possibly as a result of the wow! it’s terrible Vista) Linux, worrying (hmmm..) not.
Clearly you didn’t read the original or the 2nd articles, they are two separate articles with two separate outcomes, different results but the same conclusion.
The 2nd article has no mention of Linux at all, so why did you feel the need to troll ?
Poor article….it is shameful to read something like this in this website. I guess Apple and MS are paying the bills.
But if you check the monthly trend since march, windows has been declining and OSX and Linux have been rising.
http://marketshare.hitslink.com/os-market-share.aspx?qprid=9
Ok I’ll finally bite and I you can mod me down as you wish. It seems that anytime Linux is mentioned in a negative light here comes the all out defenders. However I want some honest responses, not the I love Linux and defend it responses.
This is just my view so do with it as you want;,
1. Linux is not really advancing the industry in terms of inventive technologies. they continue to make alternatives however they are providing little in terms of technological advances such as OpenCL.
2. Linux is a great platform however in terms of ease of use enterprise management with great tools. The community has vehemently torn down in my opinon the one company able to bring enterprise tools and linux to small buisness and enterprises with the ease of Microsoft which is Novell. In terms of enterprise tools there is not another Linux vendor that has the complete enterprise platform of email, identity management, linux, virtualization(platespin), patch management(both Windows and linux), reporting services, directory management.
Redhat is great however there are not a cohesive integrated easily managed platform in comparison to Novell’s offering and neither is ubuntu.
From an enterprise perspective even OS X’s opendirectory, server admin tools provide a better view than ubuntu and redhat. if OS’ X was to run on more platforms than apple we would not really even be talking about Linux especially if ZFS, grand central and OpenCl are what there are expected to be.
Why would we use OS X? You can always use Opensolaris for free.
However, I think the amount of control over the list of demands you had, patches/updates etc can easily be met by Opensuse 11.
Opensuse cannot meet all the demands, we are talking enterprise not hacks. Groupwise server, Zenworks server, edirectory server, non of these are supported on Opensuse. Opensuse offers no support for enterprises nor does it offer integration of enterprise services.
In regard to Solaris you are still missing the ease of use enterprise directory system and management capabilities. We are talking GUIi tool or web based tools that enterprises demand to train staff. Time machine was not inventive it was merely marketed with the right tools that a user could easily be trained to use.
Active directory is not inventive but provides the tools to train admins to use. Apple, and Microsoft succeed partly from being inventive the other from producing intuitive tools to allow users to manage networks.
Even Apple’s OS X server can allow office admins the ability to easily teach them how to use the tools in a method that they can understand from creating users and email accounts, restarting services and creating shares. The fluidness of Linux tools, often lack of consistent documentation makes it a moving target. That was were Novell would have succeeded if the community did not start to rip them to shreds. I sometimes view the community as crabs in a barrel committed more to distros than the common Linux goal.
Edited 2009-02-02 23:51 UTC
It is clear from your posts that
a: you have never worked in an enterprise
b: you have never fully evaluated Linux, especially Opensuse
OSX is fine for a personal computer, or as a server for something like a youth group, but using it in a data centre ? c’mon
Unfortunately it is clear you are a bad judge of character because;
A: I have worked in many enterprises and continue to work in them.
B: I have not only evaluated Opensuse but contributed, as well. Also I have deployed Linux farms both Red Hat as well as SLES and frequently serve as a consultant for Solaris to Linux migrations.
However I will elaborate, why don’t you educate me and show me an ENTERPRISE that is using Opensuse as the center of their environment. Directory, identity management, virtualization, email, monitoring tool not one part but in all cases. Again I said show me OpenSuse not SLES with the Novell add-ons since that is what you brought up. PLEASE show me the SOX and HIPAA organizations using Opensuse enterprise wide.
You are taking my comments and placing them in your perspective of discounting OpenSUSE when that is not the case. However center of the ENTERPRISE from a support standpoint it lacks the backing as well as the tools. Where is the competing product to Zenworks, where is the comparable Groupwise product. ( I am talking feature for feature) OpenLDAP is no edirectory, where is the identity management ENTERPRISE identity management. I suggest you brush up on ENTERPRISE tools. There is a difference between tools that can be used in an ENTERPRISE and ENTERPRISE class tools.
VirtualBOX can be used in a ENTERPRISE that does not make it VMWARE. Yes I know of Xen but I am being sarcastic.
See what you failed to realize is that I was speaking from an enterprise wide solution which Novell provides over support on their SLES platform. In fact they are the ONLY Linux vendor one with a comprehensive solution comparable to Microsoft on Linux.
I also suggest you search all my other posts here over the years and see that I have always been a Linux backer, as well as I am very proficient in Opensuse. However I am very honest in it’s short comings.
Edited 2009-02-03 16:26 UTC
I think the other person was winding you up because of what I said about OSX. Pity some people dont have their sarcasm detectors switched on before they post. I wouls have modded him down, but I have posted already.
While these little figures are an interesting conversation piece they serve no other purpose. I have been using Linux for over six years now and I like Linux better than Windows and Mac. Linux serves my computing needs just fine and it’s free and open. Microsoft Windows and Apple Macintosh are not open, or free so I won’t touch them unless I’m forced to do so for work related purposes. For my own machines, it’s Linux baby all the way. I also try to use only open source applications like OpenOffice.org. I really don’t care what my fellow man is using on their computers so statistics regarding that are meaningless to me. If there’s a silver lining here it’s that virus authors probably won’t be targeting Linux anytime soon.
Someone already mentioned it, but I am really surprised that there’s not a single company, that actually creates its own, great looking quality-hardware, takes coreboot, a Linux-distribution and ties them to the hardware.
Hell, for what it’s worth, create its own OpenGL-based window server and load it as a binary, so the open source-crowd can’t just recompile the source you have to publish in order to make it work on every machine. A few core-components of OS X are open source too.
At least it’ll have a future, which I can’t say about sold hackintosh-boxes.
> Hell, for what it’s worth, create its own OpenGL-based window
> server and load it as a binary, so the open source-crowd can’t just
> recompile the source you have to publish in order to make it work
> on every machine. A few core-components of OS X are open source
> too.
I can only agree in part to this. If you start including major proprietary parts in the OS, then you basically start fighting Apple on their own ground, and that is bound to fail because 1) they are doing a good job and 2) they have a huge head start.
Don’t forget that the product you’re trying to sell is a fully integrated system, not a single piece of magic software. Better keep it open, let others use and improve it, but protect your own trade marks at all cost such that rip-offs must create their own distro or product and not pull a Hackintosh.
Of course, some parts may have to be closed because of licensing requirements (e.g. third-party drivers), and this might even force you to use the BSD kernel instead of Linux. However, such closed parts would be a compromise, not a concept.
Well, KDE 4.0 did not help things.
A couple of friends of mine who were die hard KDE 3.5 fans couldn’t wait for KDE 4.0, and when it was released just went nuts.
They told me personally, “These idiots have no idea how to produce a desktop OS, just a kernel. We have what we need on the Mac, so come with us!”
I stayed behind, but 3 of the people I use to know as KDE 3.5 linux users, are now Macd users.
I think KDE 4.0 scared a lot of people.
It is going to be several years, before KDE 4.0 is as nice and as stable as 3.5. I am willing to wait, but my friends were not.
-Hack
What a misleading title! Seriously Windows has such acommanding lead and monopoly that 3% loss is like nada.
What’s a 3% loss of 0.85% amounts to? Nada.
I see nothing has changed it’s still the same old status-quo.
say only 5% of linux users used net apps…
what if only 50% of windows users used net apps?
its food for thought yes, but conclusions of how good an OS is based on stats like this are just silly.
I think people think that just because lots of people dont use linux, that our linux desktop is all of a sudden less usable than we found it yesterday.
until linux adoption starts declining, there is no problem. linux doesn’t have marketing as so many have pointed out – so it’ll only ever grow slowly. there will be no massive increase. Plus, at the moment, it still requires above-average computer skills to use it effectively.
That doe not matter much tracking 160 000 000!!! is even more idiotic time and money wasting.
—
Statistics are like cake they are super-good (almost future knowing sometimes) if professional made them and interpreted them deeply understanding how he can be totally wrong because everything is based on probability theory.
AND
Here we see pure opposite. All possible big mistakes in statistics are done.
Edited 2009-02-03 13:05 UTC