Apparantly, it’s damn lies statistics time again in the Apple world, boys and girls. A few days ago, Net Applications published OS market share figures which showed that the market share of the Mac OS remained largely flat over the past year. However, today, the Switchtoamac website posted an article which looks at the same set of figures in a different way, only to conclude that when you compare last year’s figures to this year’s, the Mac OS has risen 25% in marketshare. We all know the saying: there are lies… Damn lies… Funny detail: while the debate rages on about half percentage points for the Mac OS, Windows XP increased its share by 8 percent the past 12 months. According to these figures, of course.
What the article points out is that the data from Market Share lists 2 different operating systems for Apple.
1. Mac OS
2. MacIntel
So, the other stories earlier in the week failed to scroll down to see “MacIntel” in the data. The article is very detailed and convincing.
I also looked at Windows XP data, although XP rose, all other Windows variants declined. The article over at switch to a mac demonstrates that Windows has actually lost market share so far in 2006.
The results from both articles can exist side-by-side, one does not exclude the other.
The first article (Inq) says that when looking at 2006, Apple’s market share did not increase month-over-month.
The second article (switchtoamac) says that when looking at both 2005 and 2006, Apple’s market share rose year over year.
Point: same figures, different results. I try to steer away from market share figures until I can combine multiple points of view into one newsitem, like this one. I hope you guys understand that, as it provides the whole picture, instead of a single-sided one, as is the case with various other webistes.
While I agree with the Mac OS and Mac Intel issue. I wouldn’t call this “great analysis” by a long shot. First, the author exaggerates the effect of this omission by starting his graphs at 3.25% instead of 0%. I understand the site is biased towards Macs but still. Second, though the “analysis” finds that YOY increase is there, if you look at the actual market share, all that increase in Sept-Dec 2005 and the share is pretty much flat throughout 2006. While this could mean that most Mac sales occur in the last calendar quarter (like any other business) and there may be another spike in the months to follow but the YOY analysis buries that result which a month to month or just the raw market share numbers do show.
I’m sceptical both of claims that show a dramatic reduction in MacOS market share, and of figures that show a dramatic upswing. Especially in an article that purports to show that “other OSes” collectively have experienced NO GROWTH over the same period: I just put FreeBSD on one of my machines, but because I didn’t buy it, it’s not going to figure in any OS market share statistics.
The point is not that my putting FBSD on anything is statistically significant, but that limiting marketshare statistics to sales of new OSes discounts both those that can be gotten for free, and retention of old versions of all OSes – as we know, for example, millions are still running Win98. Hell, maybe there are even a few millions still using DOS on desktops.
Put another way, if you gauged Linux marketshare by taking into into account all *installations* of Linux/BSD distributions (regardless of whether those installations actually stayed on the computers), and everyone who is using it installed as many systems as I have on as few computers, you would probably conclude that Linux/BSD *usage* outstrips Linux *sales* by a factor of something like 6 to one, and that consequently Linux marketshare (overall, not only on the desktop) is around 72% (six times its share measured by sales).
Edited 2006-09-21 15:46
The total number of computer does increase with the time, so how can you compare percentages cleanly withouth taking those in account?. F.e., If apple has 1000 computers in the world and sales increase a 20%, the final sales will be 1200 computers. However, if windows-based computers owns 100000 units and increases sales 1%, the increase will be 1000, which is highter than Apple and will look like a decrease in the graphics, despite of Apple sales being much more healty.
Also, look at those numbers. They’re web statistics. Newflash: Not all computers in the world are connected to internet, those statistics are completely unreliable. And I don’t care about how much windows XP itself increases sales, because most of them are users migrating from windows 9x and some 2k machines.
The real fact is that right now Apple sales have a highter percent increase compared to the last year; when compared with windows-based machines – according to noumerous sources. Will that last forever? probably not, who knows, it’s easy for apple to get big % numbers precisely because their current user base is small, and its much harder for windows-based machines because their user base is huge. Even windows zealots can udnerstand that. I just don’t understand what this article is about; I though osnews readers are quite used to statistics lies because of the benchmarks you usually see posted here…
Edited 2006-09-21 15:55
A combined Windows market share of 95.11%, sounds like a monopoly to me. The DOJ sure is blowing that one. At least the EU is stepping up to the plate. Will we ever have real competition again?
A combined Windows market share of 95.11%, sounds like a monopoly to me. The DOJ sure is blowing that one. At least the EU is stepping up to the plate. Will we ever have real competition again?
The question of ever getting real competition again should be decided by the marketplace and innovation…not by politicians.
Relying on politicians is why the monopoly conitnues.
The marketplace has already decided…People will go for an overhyped, overdue, overcomplicated system at the expense of decent software, because the decent software is either not marketed, or marketed poorly, whilst the poor software gets all the backing because the company behind it is unscrupulous.
MS’s problems are due to non-compliance with the law.
Edited 2006-09-21 16:47
The decent software if in your definition is Ubuntu or FreeSpire then thank god XP is there.
In last 3 days i have tried Ubuntu Desktop, Ubuntu Alternative and FreeSpire and all fail to install. Ubuntu Desktop Install hangs at 16%, alternative at 6% and FreeSpire on the very first screen…go figure…
Their forum…no reply in last 24 our…nice…it is costing me already way more than windows…
Once they’re up and running, Linux and *BSD are more stable than Windows, easier to fix, easier to maintain (unless you refuse to factor in the performance and private-data-loss factor) and less susceptible to viruses, &c. I can’t vouch for the quality of Ubuntu, but I can vouch for the quality of plenty other distributions, and their support (which you don’t have a right to complain about if you don’t pay for, and why should you?)
And in what way has it “already cost you more than Windows?”
Edited 2006-09-21 17:35
Once they’re up and running, Linux and *BSD are more stable than Windows, easier to fix, easier to maintain … and less susceptible to viruses,
I’ll agree with you about the virus drawback of Windows…but I don’t agree with Linux and *BSD being more stable, easier to fix & maintain.
That’s *your* opinion.
*My* opinion is that windows XP is rock solid. I can’t even remember the colour of the blue screen of death anymore
))
I will also say that I’ve been using and programming computers since the apple ][ came out, and not once have been the victim of a virus, trojan, malware, or anything that I didn’t know was on my Windows machine(s).
But that’s cuz I practice *informed* computing…and sadly, most Windows users do not. Now, who’s fault that is, is open to debate.
But, hey, park your BMW with the windows down, doors unlocked, in a bad area of town…don’t blame BMW if your car is stolen.
Edited 2006-09-21 17:47
My opinion is that we do not use the same version of Windows.
You must use a hidden version one which is “rock solid”.
My opinion is that we do not use the same version of Windows.
You must use a hidden version one which is “rock solid”.
Well then, that’s your problem, not mine…isn’t it.
I’m sorry if you’re not capable of installing and operating a stable Windows desktop.
Neither are my parents…that’s why I set their machines up for them *immediately* after they pulled them out of the box. They, too, have been crash and incident free.
I don’t know what your experience has been, but mine has been that Windows is quite stable for a while, but as you add more applications and spend time on your machine, it eventually becomes unstable. It usually takes about one year for me to see my Windows installation degrade to the point where I want a fresh install so that everything is nice and snappy again.
Last week, my laptop’s screensaver was really slow, the wireless wouldn’t share the router with any other machines, and the machine takes too long to start up. So I reinstalled, spent about three hours getting all my updates, then spent another four hours setting up my application set again. The machine is fast and stable now, with all the same apps as before when it was slow.
I expect in a year to have to do this again, so I’m taking this installation and backing it up as a tarball that I’ll just drop in next time.
Edited 2006-09-21 18:24
I don’t know what your experience has been, but mine has been that Windows is quite stable for a while, but as you add more applications and spend time on your machine, it eventually becomes unstable. It usually takes about one year for me to see my Windows installation degrade to the point where I want a fresh install so that everything is nice and snappy again.
Agreed. I’m beginning to have this issue on a notebook that has had the same install of win2k since Feb 2001.
But having the problem after only 1 year sounds a bit suspect. I avoid 3rd party software that has let me down in the past by adversely affecting the OS.
It’s getting to the point where I’m going to put FreeBSD or XP on it…depending on my mood at the time.
But the xp pro desktop on which I’m typing now is 2 years old is still as slick as the day it was given to me…but start up and shut down times are agonizingly slow.
By the way, I do in fact defrag
I just want to spare myself the pain of hearing someone think that was my problem.
I like Windows, it just doesn’t stay stable indefinitely. One year may sound like a brief time, but I’m on my laptop nearly all day every day.
I haven’t been using Linux on the desktop long enough yet to know if it’s better. But at least with it, I can dig down to troubleshoot my issues.
Edited 2006-09-21 18:29
For a desktop workstation, it’s probably possible to make a WinXP machine as stable as a Linux or BSD machine used for the same purpose, with more hard work. (Nevertheless, I have indeed seen viruses, BSOD’s, data loss and (bucket loads of) spyware on PC’s used for nothing much beyond internet access and short Word documents.) However, try using it (that is, a “server” version of Windows) for the same number of tasks as an equivalent Linux server, and you will quickly see why server admins love Linux.
There have been allegations (in a poorly researched article on NetworkWorld) that Google is considering switching from Linux to Solaris; however, it’s notable that whilst many would consider Windows as the “path of least resistance,” not even that article purports to claim a switch to Windows. The reason? In order to maintain and extend its position as the no. 1 search engine on the Web, Google needs to maintain 24/7 service, and Windows just doesn’t cut it.
Windows just doesn’t cut it…you say.
Well, writely.com, google’s acquired web word processing program, runs on Microsoft’s ASP.net…and runs well, it’s my wp of choice.
I do agree, though, when it comes to *server* administration…nothing beats *nix. Windows is not even close in this regard.
Windows just doesn’t cut it…you say.
Well, writely.com, google’s acquired web word processing program, runs on Microsoft’s ASP.net…and runs well, it’s my wp of choice.
I do agree, though, when it comes to *server* administration…nothing beats *nix. Windows is not even close in this regard.
When I said “Windows just doesn’t cut it”, I *was* referring to use as a server operating system, specifically. As for web-based apps, they are much the same position as OpenOffice: They’re perfectly good, if you don’t need the extra features of MS Office; the same, for different users, could be said of Abiword or even vi (with or without TeX or troff). With web-based applications, of course, you also have to factor in the additional risks of hosting your data on somebody else’s servers, which (among other things) may or may not be accessible 24/7. (I’m pretty sure that Google runs 24/7, but also pretty sure that my internet access doesn’t.)
When I said “Windows just doesn’t cut it”, I *was* referring to use as a server operating system, specifically.
Huh? I just showed you a perfectly good example of a *google* .net web application running on a *windows* server.
I’d say windows was cutting it in this case as a ‘server operating system’.
That would depend on its uptime, measured against usage patterns, in absolute numbers and in comparison with others. And of course, many people do run Windows as a server operating system. Just not for the things they need Unix and VMS for.
That would depend on its uptime, measured against usage patterns, in absolute numbers and in comparison with others. And of course, many people do run Windows as a server operating system. Just not for the things they need Unix and VMS for.
If it’s good enough for google, it’s good enough for me.
I think Google is likely to port the Writely .Net stuff to their J2EE architecture for scalability reasons.
My company has had to do this itself. This has been really helpful:
http://dev.mainsoft.com/
.Net is pretty cool, but it still doesn’t scale like running an appserver does. You can get .Net to scale, but it’s a lot more expensive than J2EE after a certain threshhold.
Edited 2006-09-21 18:40
I think Google is likely to port the Writely .Net stuff to their J2EE architecture for scalability reasons.
Oh please…
If they do port, it will be so that it jives with the rest of their enterprise architecture.
If they don’t port, Writely is, and will continue to scale as a .net app just fine, thank you very much.
Just pick the right tool for the job, I say. We do all our in-house IT apps in .Net. For our app that’s being used on a national scale by the UK healthcare system, .Net didn’t scale in our trials. J2EE did. That’s one reason why the financial industry goes with it too.
EJBs have been a lifesaver for us. We have to be able to perform transactions across a huge constellation of powerful UNIX machines. .Net simply doesn’t have capability!
Now, as far as serving up web pages goes, .Net and PHP are both good candidates. But Google is very amibitious, and I doubt they want to start forking over enough Windows licenses for the huge clusters they maintain to give Writely the level of availability they want.
Edited 2006-09-21 18:54
I’m sure there are instances where the converse is true.
Like you said: “Just pick the right tool for the job”
Oh, and RE the financial industry…I was on a team that built a large derivatives trading system…using .net and an all MS software stack.
A 2 year project used by offshore banks in the Caribbean. At the end, worked like a champ. It’s what sold me on .net…after coming off a J2EE project.
Hmm, if you think .Net can scale as high as J2EE, you need to do more research. Beneath the threshhold where that matters, though, I think .Net apps can be developed faster than J2EE apps. If time isn’t such a factor, though, I think Java or PHP is best because you can play vendors off each other if it comes to that or even port from UNIX to Linux when it’s time to start shaving off costs five years down the road.
Edited 2006-09-21 19:01
But Google is very amibitious, and I doubt they want to start forking over enough Windows licenses for the huge clusters they maintain to give Writely the level of availability they want.
Fair point.
I always thought writely was a strang fit considering the way Google operates.
I think Google is likely to port the Writely .Net stuff to their J2EE architecture for scalability reasons.
MySpace is build on .Net and has scaled pretty well:
http://weblogs.asp.net/scottgu/archive/2006/03/25/Handling-1.5-Bill…
MySpace is build on .Net and has scaled pretty well:
Wow, I did not know that…thanks for pointing it out and linking to the associated blog:
MySpace.com is now processing 1.5 Billion page views per day
MySpace.com handles 2.3 million concurrent users during the day
MySpace.com’s average server CPU utilization went from 85% to 27% after moving (from another technology) to ASP.NET 2.0
MySpace is build on .Net and has scaled pretty well:
More from the site you posted:
“Totaljobs.com also uses asp.net, it’s the biggest recruitment site in the uk: well impressed by it.”
.Net scales, but it scales more expensively than J2EE at Google-sized porportions.
Don’t get me wrong: Microsoft makes great stuff, but at the end of the day, their technology is meant to benefit Microsoft first, and you second. At least the money goes towards making a better product for you, though, so it’s a symbiotic relationship. However, it’s doubtful Microsoft would continue to advance its technology if it weren’t for its open source competitors in the form of J2EE and LAMP (this doesn’t make MS “evil”, just a reality of business). But the advancement of these open source solutions are driven partially by market forces, and partially by science. With Microsoft, science doesn’t enter the equation at all.
J2EE is definitely headed toward niche, “superproject” status, though. The only reason my company had to use it was because of EJBs and the neccessity of good transaction management across many AIX/Oracle boxes.
LAMP would be superior to MS if only they had a better RAD environment and integration. Ruby on Rails is a great language & framework, but sorely needs an IDE with true breakpointing (not the breakpointer hack) and better scalability. PHP with MySQL and Eclipse is a good start, but still not as integrated as the MS world.
Edited 2006-09-21 20:51
“When I said “Windows just doesn’t cut it”, I *was* referring to use as a server operating system, specifically.”
Then the whole “monopoly” rhetoric is irrelevant. Microsoft doesn’t have a monopoly in server operating systems, and has never been ruled to have such. Jackson found that MS had a monopoly in desktop OSes running on intel-compatible CPUs. (I don’t think that’s even the case anymore with Macs now running on intel CPUs as well.)
Then the whole “monopoly” rhetoric is irrelevant…Jackson found that MS had a monopoly in desktop OSes running on intel-compatible CPUs.
Thanks for pointing out the flaw in your own argument.
(I don’t think that’s even the case anymore with Macs now running on intel CPUs as well.)
MacOS still doesn’t run on just any old Intel chips… Even if there’s a way of getting it to run (natively) on generic Intel hardware, I doubt it works on as many different hardware configurations as any of the free Unices.
The DoJ is supposed to be judicial and not legislative: It’s not supposed to be bound to politics. Although, the unfortunate appointment mechanism has made it political.
Anyway, they don’t regulate success, they regulate anti-competitive behaviour. Having a 100% market share is a tell-tale sign of anti-competitive behaviour, but it’s not anti-competitive in and of itself: You’ve got to figure out why they run the market. Are they just that great, or are they stepping on their fledgling competition.
This is why the EU goes after things like Windows Media player, and not just the prevalence of Windows itself.
In the words of everyone under 30: It’s not the market share that counts, it’s how you use it.
And relying on politicians they will. http://www.windowsitpro.com/Article/ArticleID/93590/93590.html
The circus must go on!
“The question of ever getting real competition again should be decided by the marketplace and innovation…not by politicians.”
I think you may need to do a little more research into the the history of monopoly and it’s detriment on innovation. I’m all for the free market but it is not perfect in of itself. It is governed by laws to protect the consumer and businesses alike from unfair and anticompetitive business practices which eventually lead to monopoly. Microsoft is guilty of these practices and have been ruled many times to be a monopoly. They don’t sustain dominance through innovation, they sustain it through monopoly. The best way to break that cycle is to hold them to the law. Too bad our corrupt government can’t do that anymore.
“The question of ever getting real competition again should be decided by the marketplace and innovation…not by politicians.”
Big disagreement on that one. For one, marketplace and innovation do not necessarily go hand in hand. Second, history tells us the exact opposite. Without monopoly busting by the government, we’d still be buying crappy phone service from Ma Bell along with their overpriced long-distance service. There would be no MCI, Sprint, or any other long-distance company you can think of.
We’d be paying even more for gasoline than we do today (US vs. Standard Oil), and price fixing between competitors would be legal.
Regulating monopolies might have pitfalls, but laissez-faire capitalism isn’t the answer. Advocating the complete absence of legislation in favor of laissez-faire relies on the same assumption that Communism did—that everyone is as high-minded and morally incorruptible as you are, and that is never ever the case. The best products don’t always prevail under that system, if the company with the largest marketshare can successfully force everyone to buy their inferior product by OEM deals and taking a loss on pricing to increase marketshare.
All good points, nalf38.
And you’re right on all accounts.
I guess I’m just expressing my frustration that too many folks have thrown in the towel and thinking the DOJ and EU are the only avenues of recourse.
You are making a mistake here.
You think the companies can force us due to large marketshare… that’s not the case.
They get the large market share because they can force us through law. In a system laizzes-faire capitalism there would be no patents and no limitations at all in competition, and therefore no way for companies to become big – unless it is due to people not caring. Powerful companies are powerful ONLY because they can buy power and therefore change the laws according to their needs.
That’s why we’ve had increasingly tightened copyright laws the last century.
Ironically, the “baby bells” have again begun to “remerge”. I wonder how long it is before we once again have a “ma bell.”
IBM is restarting OS/2? Apple is going to license OSX to OEMs?
Actually, I’d like to see both.. I’d like to see Apple recognizing that because I bought an iBook and a Powerbook doesn’t mean I wouldn’t like to run Mac OS X on a 3rd party desktop…
I’ve loved OS/2 for years but people don’t seem to get it, so It’s being discontinued and all I’m left with, arguably, is Windows, Linux, FreeBSD, NetBDSD, OpenBSD the latter of which aren’t necessarily very suited for desktop usage… i’d like for a strong commercial altenative to Windows… Not because I dislike it but rather since I’d like for others like me to have viable alternatives…. Linux isn’t the only one… We need all the fire power we can get
In the US it is NOT illegal to be a monopoly. It IS illegal to use your monopoly to compete unfairly in the marketplace, however. In a free economy where competition is not stifled and innovation is fostered, it is almost impossible to maintain a monopoly. The key thing to do is not to prevent monopolies or “break them up” as was done to AT&T but to punish them very very severely (in the case of MS that would require Billions of $US) when they illegaly squash competition.
Actual statistical studies involve control groups, a rafter of validity tests analyzing both the validity of the sample set and the significance of the results, as well as detailed error analysis and uncertainty estimation.
I can post a marketshare listing just as valid as this one:
Windows’s market share is cake +- ice cream.
Linux’s market share is pie +- whip cream.
OS X’s market share is eclairs +- fudge.
Mmmm, eclairs…
Edited 2006-09-21 15:53
Yes, Rayiner, you are right, these aren’t statistics in the purest sense of the word. However, when I get my Dictionary of English (a real paper one, people), it says:
“Branch of mathematics dealing with the collection, analysis, intepretation, and presentation of masses of numerical data.”
As well as:
“A collection of quantative data.”
It’s statistics alright.
There is a distinction between “quantative data” and “numbers”. What seperates “data” and “numbers”? Methodology. This article displays no methodology, and its results are therefore not “data”, but mere “numbers”. If statistics are a collection of data, this does not qualify.
“There is a distinction between “quantative data” and “numbers”. What seperates “data” and “numbers”? Methodology. This article displays no methodology, and its results are therefore not “data”, but mere “numbers”. If statistics are a collection of data, this does not qualify.”
Like I said: a collection of quantitative data. Quantative data means data you get by counting stuff, simply put. This is a collection of quantitative data, hence, according to the dictionary, the term statistics is well suited.
Of course, Rayiner, like I said, you are right that strictly speaking, the analysis of quantitative data produces statistics. However, in day-to-day speech, any set of numerical (=quantitative) data is referred to as statistics.
There is a distinction between “numbers” and “quantative data”. This is a distinction that is carried over into everyday speech. A reporter might say “the sheriff estimates that 50% of day workers are illegal”, while he might also say “the county labor board statistics show that 50% of day workers are illegal”. I’ll grant you that statistics are sometimes conflated with numbers, but the distinction between the two is maintained even in scenarios where one is trying to maintain a modicum of precision (ie: in the reporter example).
Thom, in .it we call this “scrambling on mirrors”.
He has a point, simply put… even if with a provocative style
Statistics can mean statistical analysis or information systematically collected. All data collections imply certain types of bias and problems of many kinds. But these data are actually systematically collected. They aren’t unbiased, but it’s likely that the bias persists over time hence an adequate analysis can draw conclusions from their analysis.
I’d say that the second analysis (the one adding macintel+macppc) is elemental, but hits on the nail. The data are from the same source, the comparison year-by-year helps address potential seasonality issues. It doesn’t address what type of information this is, but at least compares the trends in the netapplications source with the trends in computer sales. It’s not material for a statistics journal, but it’s OK.
I’d be very interested, of course, in having better metrics and a better defined problem — it’s unlikely however that we’ll ever have the chance to see a study like that in a public forum like this.
Actual statistical studies involve control groups…
I’ve been compelled to teach statistics (thankfully, those days are over), and according to every textbook I saw, the data given in the article are… statistics! So if I’m wrong, you’ll have to blame the poor quality of textbooks on elementary statistics in circulation.
They may not be particularly useful for drawing certain conclusions, and for drawing such conclusions, one may need to do some hypothesis testing, but that’s another matter.
BTW, what would be a control group in this situation?
To get a true sense of Apple’s market share, one should look at hardware sales, not software sales. Apple is above all a hardware company and makes most of their money off of selling computers (and electronic appliances), not boxed sets of OSX. I think it would make a lot more sense to compare their PC and especially Laptop market share (which has much higher margis) with all the other PC vendors: Dell, HP, Lenovo, Sony, Acer, Asus, Gateway, Alienware, etc etc. Then you really get a sense of how Apple stands in its market. It may only have 3-4% of the OS market, but when you consider that all its competitors sell the same OS (or flavors of it), it’s hardware PC and especially Laptop share is considerably larger. I don’t know if anyone has reliable figures on that. A quick Google searched turned up a report that Apple’s laptop market share in the US market is around 12%, but that’s only Retail, so it’s very incomplete.
Funny detail: while the debate rages on about half percentage points for the Mac OS, Windows XP increased its share by 8 percent the past 12 months.
Seems the gain is mostly at the expense of older Windows operating systems.
That’s not unusual, it’s to be expected.
Nice graphics and numbers but I prefer to quote someone this time:
“Don’t trust any statistics that weren’t falsified by yourself.” (Winston Churchill)
Enough said.
Edited 2006-09-21 16:37
these are just browsing statistics no? could help explain the poor showing of linux in the “other” category, not many servers out there browsing. but anyway, to figure out market share is way easier than looking at some web stats, companies publish their hardware sales, and that will give you market share right there, problem solved. (though i believe you would likely see the same results, as apple sales for this last quarter were only up a hair, but i think we’ll see some pretty good growth this quarter.)
The unfortunate problem with these stats is that they are relying on metrics from Net Application usage. The problem is that these numbers have some serious flaws in usage. If you think about it logically (and you look at the partners list), you’ll realize that the metrics are being pulled together from many sites, but that many of the sites they use aren’t ‘general cross section’ sites, which, in my opinion alters the accuracy of the results. Further, looking at the rise and fall of the numbers in the month to month categories, the Mac numbers actually rise when they should fall, right before the Tiger launch, around the Intel announcements etc.
Notice that one of the partners is Alexa, which doesn’t offer any Mac support.
I think if you want a better cross section, you need to be using metrics off of ESPN.com, CNN.com and others. Google I feel is unfairly biased towards the Mac due to all the browsers having google as the default built in search tool, while the dominant Windows browser remains tied to MSN.com.
Either way, these numbers do confirm one thing, Mac OS marketshare is not increasing dramatically in the workplace yet, because the other place this kind of statistical analysis doesn’t include is Porn.
You want accurate home usage number? Log Porn sites.
This is just silly, statistics like that cant be wrong. I can tell you for sure. If mac has 4% it means they have sold 100 million computers. and all linuxes together comes up to a measly 0,4% wich means only 10 million. That is just ludicrous. There are more than 10 million linux computers than 10 million. Just count the servers sold by ibm, hp and redhat they alone gets that number. So what about the rest of us? do the desktop linuxes not count?
ok my dear friends, once again sarcasm =)
Sorry for the tripple postings but look at their browser vote poll…. http://www.surveyware.com/chartfx62/temp/CFT0921_0127111A5BD.png
0.2 has voted on non of the alternatives =P haha there goes their credibillity. As the last one isnt even possible to vote for…..
that’s why I set their machines up for them *immediately* after they pulled them out of the box. They, too, have been crash and incident free.
While I won’t say it’s impossible to set a Windows machine up to be stable, I will ask if you consider the machine to be ‘fully functional’?
I mean seriously, if it’s connected to the internet, I’ll assume you are using a hardware firewall to protect the Windows environment. I’ll further assume that there is no Wireless LAN enabled on the box, to prevent compromise.
Then there is the question’s of how your parents install software, since to protect the environment, I would assume you’ve made them ordinary users and revoked admin privileges, as well as renamed or disabled the Administrator account. Of course, you’ve also installed protections against malicious installers, and Anti-Virus software to prevent the obvious email and general virus issues.
I would also guess that you’ve removed Windows Media Player, and Internet Explorer, or at the very least have both set up run in high security mode, with JavaScript, ActiveX and Java all disabled. So you have probably also switched them to use Thunderbird and Firefox, so that you don’t have the problematic Outlook or Outlook Express products in use, and thus avoid to more notorious sources of malware, viruses.
Of course all of this is predicated on a little education, unless you are filtering all attachments off the incoming emails. It’s still awfully tempting to double click that attachment from Bob about your ‘Viagra Order Invoice’.
What I’m getting at, is that it’s not that hard to make Windows stable and protected, but it is hard to keep it that way and to use it to it’s full potential. Seriously, some of the most powerful features of Windows are so crippled by confirmation dialogs now that they have become useless features (Outlook’s robust and excellent scriptability via COM and JavaScript is basically useless since you can’t automate any task that tries to read the email address off of a mail message without confirming the access). That’s just one example where great ideas and great implementation within Windows has become hopelessly compromised by the security issues that boil down to stupid users.
Ultimately, finding a balance between usability and Security is going to have to be reached, and Vista doesn’t fix these problems, and in many ways makes them worse, because Windows has inherited a fundamentally flawed security design from a time of single user operations. Apple doesn’t have that issue (they have others, the above example is perfectly applicable to Mail.app, in that the protection isn’t there, but the potential for social engineered evil is).
That doesn’t matter though, there is ALWAYS the potential for user stupidity to make a mess of any platform, and as an administrator (what you become when you set that computer up to be ‘stable’ desktop) you have to find the balance between functionality and security. After all Windows is only C-2 certified when it is not plugged into a network in any way…
dru_satori, you assume and guess correctly. Good post.
Where I contract, the data centre is based on AIX and BEA Weblogic.
I’m consistently amazed at the AIX admins here, who would never consider doing anything on the heavily protected data centre boxes using the *root* account…then go home and share files & surf porn on their Windows machine using the *Administrator* account.
Then they come to work the next day and say windows sucks cuz they got a virus :rolling eyes:
Indeed…dumb or ill-informed users will always be out there. And they will almost always blame anybody but themselves when they get compromised.
If Aches is anything like other Unixes, they make it (relatively) “painful” to install stuff, and relatively easy to run it as a non-root user. Windows programs are not designed with that kind of security in mind, so running them as a limited user can often cause problems; thus everyone runs as an Administrator instead, even when indulging in unguarded browsing on the ‘net.
Doing everything as root in Unix would be as silly as is doing everything as the Administrator in Windows, but Windows forces you to do it anyway. That’ll be the source of their complaints.
If the rumours are true (I haven’t and don’t intend to try Vista unless and until I’m more or less forced to buy or use a computer with it preinstalled), Vista makes it impossible to do something as simple as put an icon on the desktop without administrator privileges. If true, this shows that MS either haven’t got a clue about how to implement proper security in an OS, or just haven’t got a clue about how to do it in Windows. Not wanting to break old apps can’t be the issue, as apparently this and other changes break plenty stuff; nor can it be definitively said that it will improve by release time, especially since one source purports to quote an MS employee saying MS has no plans to modify the way UAC works at present.
If Aches is anything like other Unixes, they make it (relatively) “painful” to install stuff, and relatively easy to run it as a non-root user. Windows programs are not designed with that kind of security in mind, so running them as a limited user can often cause problems; thus everyone runs as an Administrator instead, even when indulging in unguarded browsing on the ‘net.
Doing everything as root in Unix would be as silly as is doing everything as the Administrator in Windows, but Windows forces you to do it anyway. That’ll be the source of their complaints.
Ahhhhh!!! But is that the fault of the programs…or Windows?
It is my experience that, whenever application developers write software that works great…they take all the credit.
When their software doesn’t work well…they blame everything on Windows.
Having to run as admin is a poor design of the application…not windows.
Hell, if I can build web apps with VS.NET using a restricted account…there is no reason other than pure laziness to put release a Windows application that can run only as Admin.
That may be true, however, it’s the same argument that sometimes prevents people moving to Linux. If the application you need to use isn’t written for Linux/isn’t written to be run as a limited user, you’re still stuck with it.
I think the .47% market share for Linux is one of the lowest estimates I have seen.
This is interesting because other sources (
http://www.w3schools.com/browsers/browsers_stats.asp ) put it inline with Mac OS.
I also remember Google removing OS reporting from Zeitgeist because of the controversy the numbers created. I think their estimates places Linux in the ball park of 1.2% of something.
It is interesting to see a 3% margin of error for an OS that only has about a 3% share.
If they are using browser statistics to generate the numbers the difference could be attributed to the different user bases of the web sites they are profiling.
http://marketshare.hitslink.com/report.aspx?qprid=5&qpcustom=Linux
They show Linux going from .29% in january to .47% in August. So on the positive side there is growth.
Here is a simple method I have seen on another site for calculating Apple market share.
You take the total shipments over (say) 5 years, and the Apple shipments. Add them up. This gives you the percent of the market shipped that is Apple.
Next you make some assumption about the average life of systems. I see no hard evidence that Apple systems have any longer life than any others. Universities for instance renew Windows and Mac machines at the same intervals. Be that as it may, you can allow for different estimates. The number it seems reasonable to use for average life is 4 years. But you can take a different one if you like.
Here are some numbers to play with, from the Ars articles and the pegasus site. I have used 5 years, no special reason.
total shipments millions
2000 138
2001 128
2002 132
2003 151
2004 177
2005 197
total 923
Apple shipments
2000 3.8
2001 3.2
2002 3.1
2003 3.1
2004 3.3
2005 4.7
total 21.2
Total share for Apple: 2.3%
Suppose you think Apple systems stay in use longer. Then you would take the next couple of years back, say 1997, 1998, 1999 and add them to the Apple numbers, and recalculate the share. It would give you another 9.3 million or so systems, so it would raise the share appreciably. You have to raise the total market by 9.3 to correspond of course: the share is now of a higher number of machines.
Its not completely accurate as a method, because its only the average. To do it right you would really have to know more about aging and proportion of machines being replaced by year cohort. But it cannot be too far off, and globally, I doubt whether Apples share of installed base much exceeds its share of shipments over the last five years.
I don’t know of any equally simple and reliable way to estimate Linux share. The only reason this works, is because Apple is one of the majors, so we have all the numbers we need from Gartner and IDC.
Edited 2006-09-21 20:42
Sorry to reply to my own post. I have also seen this approach used to calculate what the average life expectancy of a PC would have to be for Apple to have some larger percentage share of installed base – eg 10%. Leave it to readers to run some examples.
What you do is figure the total Apple shipments you think are still in use. Then take your percentage share and find out how many machines the total base has to be, for this to be (say) 10 percent of it. Then deduct the Apple machines. Then figure back to find how many years old the Windows machines would have to be to ship that many.
That is the assumed average life expectancy of Windows machines. You find that for Apple to have 15% share of the base, the average life of a Windows machine has to be in months.
Have fun!
alcibiades, properly attribute your sources, ok? For this one time, I’ll forget it.
;).
http://www.pegasus3d.com/total_share.html
There’s a link to the Ars article. I don’t recall where I saw the method. It might have been on slashdot, or there have been periodic controversies about this topic on OSViews, could have been there.
Edited 2006-09-21 21:19
http://www.osviews.com/article/show/20247
By the way, I personally think Apple is doing very well with the new Mactels… My survey is very detailed and scientific. At work, 3 of the last 4 computers bought were Macs (mine being one of them), thats 75%. There are few other people at work who are about to buy a new computer and some are talking Macs too.
What does this mean… Well, OK, nothing other than in no other point in last 10 years have people talked about buying Macs like they are now, people not just in the computer industry like me, but non IT too. Outside of work, I know quite a few people thinking of getting one whereas once they wouldn’t have.
So I suspect since the intro of the Intel Macs, sales have increased a bit, and as a result, maybe Market share too…
Hardly,
What with Datagrids crashing from large queries.
Poor Memory Management
and Automatic “Recycling” to cover up all those memory leaks
you’ve really got to be stretching the truth to call this Second Class solution Scalable.
There’s no way Linux is .5% of market share.
What do you think is a more likely figure?