I thought it was weird that "other" had such a high percentage 4.98% - but then I clicked on the more broadly grouped chart (https://www.netmarketshare.com/operating-system-market-share...) and it dropped to 0% - "other" meant other versions than those listed.
This is a much better chart than the one linked. The other chart, due to the way it breaks down versions, is a little misleading because of how it splits up market share of older OS versions (mostly with OSX, where only two versions are accounted for in the graph itself).
It really seemed like Linux and Mac were much higher up these days. A more interesting statistic would be "OS usage times" i.e. what are people looking at for the most hours. Probably nigh possible to get. Wait, advertisers probably know.
I think the perception of Linux and Mac being used a ton is only in some developer circles. Unless the do a bunch of programming outside of work petty much all the developers I know use Windows at home.
Linux has about half of the market share of all OS X versions combined. Not too bad considering there's not exactly many systems that come with Linux preinstalled. Still minuscule compared to the Windows market share but I'd gladly announce that the year of the Linux desktop has finally arrived if it gets to OS X level...which isn't all that far to go.
But remember that Chromebooks are a US only story. There are very few of them outside the USA. One good year in the USA doesn't gain many percentage points in a worldwide stat.
I love my chromebooks (I have two), plus I bought one for my mother for whom it is absolutely perfect - a laptop (i.e. with keyboard) that is fast, light and trouble-free.
My dictionary definition of widely is a bit different than yours.
Once in a while, in a blue moon, I see one single device at Saturn megastores in Germany. Usually it stays at the store with increasing "product of the week" discounts until someone finally takes it away.
In Portugal, Italy and Greece I am yet to bump into a store selling them.
I think the actual Linux market share is higher. I know a lot of enterprises that have switched to Linux (for reason completely unrelated to the os itself).
> I think the actual Linux market share is higher. I know a lot of enterprises that have switched to Linux (for reason completely unrelated to the os itself).
So they switched to Linux because well, not because it's Linux. Funny :-) You probably meant : they switched to Linux for none of the reasons HN's readers like Linux (open, free, people-like-us, etc)
I've used Linux and FreeBSD (desktop|laptop @work/home) for everything since 2002. Within the last five years it really has been a very simple and feature rich environment with acceptable stability for most distributions.
Think if I hadn't started with KDE at that time I would be much more fond of OS X.
Something must be very wrong with their methodology. I don't believe the total installed base is so volatile that the Mac share jumps from 7% to 9.57% then down to 6% within a single year.
> Something must be very wrong with their methodology. I don't believe the total installed base is so volatile that the Mac share jumps from 7% to 9.57% then down to 6% within a single year.
Agreed. And it's by no means the first such example.
However, the numbers are based on unique visitors to a range of websites. The variation might reflect a wide variation in web accesses from different operating systems (though I doubt it), but it doesn't directly measure the installed base.
They weight users from different countries by CIA's total numbers of internet users in these countries. (https://www.netmarketshare.com/faq.aspx#Country ) So some country having a big share of internet users according to CIA and small and volatile number of users counted by Netmarketshare can be the reason.
China has about 22% of internet users according to CIA. Is there a way to estimate how many are counted by Netmarketshare?
I'm actually very curious to find out how that will turn out in the long run because another source of the same kind definitively show quite a different trend for MacOS (and IMHO more continuous & realistic) : http://gs.statcounter.com/#os-ww-monthly-201512-201612
I have a feeling this is skewed by massive enterprise purchasing of workstations - they will naturally gravitate towards a stable windows version, and tend not to get upgraded very often.
I'd like to see a chart of desktop computer market share that people purchase for themselves.
Depending on the industry, a lot of your tools might not just be Windows only but also tied to an old (and possibly unsupported in the case of XP) version. This is endemic to any industry involving capital equipment like machining centers or biotech lab equipment that have serviceable lifetimes measured in the decades. A decade ago this wasn't really a problem because this equipment was isolated but now with industrial IoT and modern IT networks they're becoming more and more vulnerable.
One mom and pop machine shop I ordered from got tired of using USB drives to haul CAD files and machining instructions between systems so they started using Google Apps with Drive as the sharing mechanism. In order to do that, they had to hook up several 10 year old Windows XP machines to the internet and within weeks they got hit by the blaster worm or some other pre-SP1 virus. That cost them weeks of machine time on their older CNCs and hundreds of man hours doing incident reports for their bigger clients like the defense contractors. The situation in the clinical biotech companies I worked at was never that bad because of HIPPA compliance and more focus on networked services like lab databases but still atrocious. So much of our industrial infrastructure is built on a house of cards.
> The situation in the clinical biotech companies I worked at was never that bad because of HIPPA compliance and more focus on networked services like lab databases but still atrocious.
I can attest that academic biotech labs are still pretty atrocious, for the reasons you mention.
Windows XP (or super old OS X installs, pre 10.5) are pretty common on non-networked machines that function as appliances.
It's not uncommon to hear about a machine controlling a $X00,000 device being compromised by malware that road in on a USB drive, despite the air gapping.
Sometimes I feel that it would be better for vendors to use a super minimal Linux distribution for each of these workstations that basically function as an appliance. Not only could the vendor reduce the attack surface for any potential malware exposure, but the one-off nature of the distro and small overall number of workstations would provide a minimal incentive for malware authors to target.
I believe the data is global so the numbers are probably skewed by China and other sources as well. FWIW, I sell a few add-ins for Excel, targeted at B2B users. I get roughly 60k uniques/month, and my Windows visitor stats are:
I wonder how many macs that would remove. Most people that use a mac that I know got their mac from work. That they use it for personal stuff doesn't mean it's not a business laptop.
On the contrary, most people I know outside of my immediate work surroundings (which is skewed by working around a lot of startups) bought their own Macbook (or had their parents buy it, as it were). They were very popular among the arts and commerce crowd at university.
Macbooks seem to have cornered the 18-25 female demographic for computers.
I see macs more among developers, but that's because I don't tend to hang around privileged rich kids that much these days, since I work for a living.
When I was at university though, the vast majority of computer science students were running either Linux or Windows. I think on the balance of it, it was mostly Windows, but the Linux users seemed to get better grades and had a better idea of what they were doing.
Macbooks were the computer of choice for the Arts and Commerce faculty students (most of whom, incidentally, did not have jobs).
... citation required? With windows terminals on every office desk, every point of sale station, every library computer room, etc. I don't think it is inconceivable that "enterprise" purchasing outnumbers personal purchases.
Interesting side note: ive only swen Windows 7 discs so far on markets in SEA. Even cambodia seems win7 dominated. But they also sell cheap laptops with mint on them.
Not to speak for the original poster but as a dev I'd like to know what my addressable market is if I'm developing a desktop-focused app, and excluding biz machines and people who will never pay would help there
If you're developing an app you need to figure out addressable market separately for it anyway. None of the apps I've developed until now really have the same % as any of those graphs - your audience will vary with the use-case.
The only thing the original proposal would do is make Apple look better, but it wouldn't really be USEFUL for anything.
With 88.67% of desktop OS being different flavors of microsoft OS, it seems a bit overwhelming that its still that high , especially after a massive acceptance of ubuntu. Its shocking that Apple has failed to make inroads in any other country outside the USA. One wonders why it never crossed their minds. This gives Microsoft a great opportunity to win back this big iPhone using crowd by making their Apps platform independent which they are doing already. I see Microsoft coming back big time like the windows 98 , 2000 and VB 6.0 kind of good times awaits them now :)
In the european market most people simply dont see the value in a mac. Even in switzerland where a $3000 laptop is not such a big of a deal you see this money invested in alienware or high quality business machines over macbook a lot.
I think they simply failed to keep up with the quality they once had.
1) If Microsoft is already/still at ~90%, is there really room for a big time come back? Getting the rest of the market is small potatoes compared to what they already have.
2) How much of the market does desktop represent, anyway? Most people I know do not own a desktop. They may have a laptop, but they may also only have a tablet and/or phone.
Now, I believe you are actually suggesting that Microsoft can leverage their desktop market to get back into tablets and phones, but I don't know . . . so far, their efforts have been disappointing, and frankly, people are used to their Android devices, so Microsoft will actually be the weird phone even if it is the normal desktop.
They've been steady at 90+% market share for a couple decades now. The modest increase in MacOS market share this past decade has never been more than a couple of percents worldwide. A drop in the bucket, really.
A lot of that is due to the fact that most businesses use Windows almost exclusively.
Even in the tech world, outside of startups and hipster tech companies, there are a lot of developers using Windows (i.e. the vast majority of .NET developers).
I'm not sure what you mean by Apple failing to make inroads outside of the USA as well. They've made inroads all over the western world.
1% market share worldwide. I was just recently in Vietnam, unsurprisingly I didn't see a single Apple computer, and most of the computers were running Windows XP.
Apple has around 10% [0] of the market for laptop sales (most consumer computers are now laptops, I don't know any non-enthusiasts who have bought a PC in the last 5 years).
Now, factor in that those sales figures include less developed countries as well as developed countries, and you're probably look at at a decent margin about 10% of personal computer sales for home use as Apple products.
I don't know. Many sectors require windows and I assume most HN readers are bored at work. You'll probably see a similar situation although my guess is Mac and Linux will be slightly higher on the totempoll. Probably more like 3-5% for Linux I'd say. Maybe even higher.
I'd love to see the data like you say, would be interesting.
People consume HN through 3rd parties, such as Twitter, RSS, and Reddit. Whatever the stats are with the website, they won't give you an accurate view of the actual HN reader.
Last months saw an example of a distributed database/api service scraping, using hundreds of thousands of IPs of around the world, and random user agents (I think that all started with Mozilla/5.0, but the rest were from random historical user agents), to avoid rate limits.
The OSs referenced in those user agents went back to early windows versions, OS/2 Warp and Maemo (to put 2 examples that hurt knowing that were fake).
If this site take as base sites/URLs that are somewhat hit by that kind of traffic the stats could be very misleading.
If you are browsing HN, you likely live and work in a bubble with massive over-representation of the very rich (who buy Macs disproportionately) and the IT tinkerers (who use Linux disproportionately).
(This is the case even for the US, but even more so worldwide, where a Macbook costs a month's gross salary rather than a week's worth.)
Is that really notable? They are both small numbers, and those NT machines are likely at companies with upgrade policies versus people who choose not to upgrade their personal macs for free.
I see this rather in favor for Mac OS X as their upgrade system definitely works, 10.8 is not old at all (2012), Windows NT is from the 90s mostly pre-internet. It's putting things into perspective.
I wonder about NT with nothing but up to date ssh exposed vs win 10 with nothing exposed except the pile of obscure phone home protocols you can't turn off.
I suspect NT wins unless there are unpatched TCP/IP exploits.
I wonder what percentage of in-use ATMs make a similar bet.
According to Wikipedia: "As of November 2013, Apple's online store was available in 39 countries, with physical Apple Stores numbering 444 in sixteen countries (263 US/181 elsewhere).
The world has roughly 196 countries. I don't know how many of those you can buy Macs in, but I'd expect there are gaps.
However, a lot poor countries have access to cheap PCs, or local cheap PC assemblers, or users can build their own machines.
There are countries where Apple cannot legally open a retail store. They go through tie ups and partnerships with local companies to setup stores and sell their products.
For example, in India - every major city probably has at least 4-5 "Apple Stores" which are not fully owned by Apple and thus probably don't factor into these stats.
This. In thailand and cambodia apple seems widely available. None of those is a official store tho. I think i've seen more than those 400 stores in Thailand alone.
In Norway too, there is a large chain official partner, essentially an Apple Store operated by not-Apple, but the online store is official Apple, I think that is what OP was referring to. IN some countries there is no Apple Store online.
I would like to see the PornHub numbers for desktop and mobile combined. That would give a better sense of what's going on. My guess for the consumer market is that it looks like Windows is in solid shape, but the reality is that people are spending more time on mobile devices, even if they have a PC somewhere in the house.
If you define "desktop" as 'not a portable computer', I suspect MacOS will drop a lot, percentage wise. Maybe Linux is already #2 in that market niche. Plenty of rack mounted linux machines run firefox/chrome...
And if you define desktop as "linux running computer" Linux totally dominates the category.
But "year of the Linux desktop" has a clear and well defined meaning since 1999 or so -- it's about people running Linux (and Linux userland, not Android) on their personal computers, whether desktops or laptops.
Without some information on how this data is calculated it's pretty tough too take it seriously.
Not that I doubt Windows is the majority OS, it obviously is. But I didn't need a chart to know that. The chart may be inaccurate depending on the source of the data and how it was extrapolated.
"The data is compiled from approximately 160 million unique visits per month. The information published on www.netmarketshare.com is an aggregation of the data from this network of hosted website traffic statistics. In addition, we classify 430+ referral sources identified as search engines. Aggregate traffic referrals from these engines are summarized and reported monthly. The statistics for search engines include both organic and sponsored referrals."
Are there somewhere market share statistics available for desktop and mobile operating systems combined? The separation between desktop and mobile devices feels a bit outdated. How does Windows relate to Android or iOS, for example?
"desktop" is a totally useless and nonsensical classification, when they do not (which is obvious from their numbers) count laptops plugged into a screen & keyboard for the duration of a workday as a "desktop".
Laptops/desktop are a form factor / product category, not a usage pattern.
The same way as 20kg desktop PC on boats are not "laptops".
Sometimes you do need to know how its used (e.g. if you sell peripherals that only make sense for "road warriors" you will want to know how many use their machines at a specific location all day, regardless of form factor), but most of the time, laptop/desktop distinction based on form, like the one we always been using, makes far more sense.
