From ZDNews: “Microsoft’s Internet Explorer 6.0 was the most popular browser on the Web last month, with a total global usage share of 44.7 percent, according to Net analysis company OneStat.com. Microsoft IE 5.5 took the second spot with 26 percent of the global usage share, followed by IE 5.0 with 25 percent, AOL Time Warner’s Netscape Navigator 4.x with 1.6 percent, Netscape Navigator 3.x with 0.5 percent and Opera 6.x with 0.4 percent. The numbers show that ‘Microsoft dominates the browser market with a total global usage share of 97% on the Web,’ said Niels Brinkman, one of the founders of OneStat.com.” Let’s wait and see if the release of Mozilla next month will at least overthrown Netscape 4.x which currently holds 1.6%. Konqueror and OmniWeb can only deliver better scores if their perspective operating systems they run on gain more popularity (Linux ~1%, MacOS <3%).
Survey Shows IE 5/5.50/6 Dominating 97% of the Browser Market
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Where the phuk are they getting those stats? I though that MacOS, atleast, would have more market share than 3%. Is the world realy THAT sad of a place?
>Where the phuk are they getting those stats?
It is my “job” to track down such stats from SEVERAL companies. WebSideStory Inc, is the most well known and respected one, and they say that MacOS/X had less than 3% at the beginning of the year.
As for Google (according to their site hits), their stats show that MacOS had around 4% in March 2002.
http://www.google.com/press/zeitgeist.html
Other *professionally* statistic companies also agree with WSS on MacOS, about the “less than 3%”.
In any case, they all agree on Linux: Less than 1% of the desktop market.
Is pretty questionable do statistics about that
opera and konqueror can fake their identity
about all non windows users have a *bit* more knowledge about web and usually they kill cookies.
How do they make statistics?
I know that the thecounter.com’s global stats have IE 5.x still at 55% and IE6 at 32%. I wouldn’t claim that their global stats are more representative of the web as a whole, but I have trouble believing that 97% of the web uses IE5.x,6.x. Maybe 90%, but there are a lot of educational institutions out there that use Netscape.
This is a good read too:
http://investor.cnet.com/investor/news/newsitem/0-9900-1028-8228199… (it pretty much says that Linux has about 0.24%)
and this one is the Google stats for January 2002
http://www.google.com/press/zeitgeist/jan02-pie.gif
BTW, we have already talked about this here:
http://www.osnews.com/comment.php?news_id=427
LuZero, about the fake of identity. Not many people are doing this (faking it
, mostly geeks and people who understand what the heck that browsing string is. And that would not really change the overall stats too much. Their percentages are too small anyway.
Mozilla is 1.0 is coming soon…
ciao
yc
“about all non windows users have a *bit* more knowledge about web and usually they kill cookies.”
These stats don’t come from cookies though. The browsers send this information as part of the HTTP request string and the Web server logs record it. No cookies required.
But as far as web browsers “faking it”, this isn’t very common anymore. At one time, most browsers masquraded as “Mozzila” so that they could access the Netscape enhanced content of Web sites and fool cgi scripts that determined what content to deliver based on the type of browser it thought was visiting the site.
I think Opera might still be the lone exception. IIRC, it masquarades as Netscape by default (this behavior can be turned off).
But if you look at a web server log, you will find that even IE sends “Mozzila” in its ID string. But IE also sends its real identification.
One should note that Opera doesn’t spoof as Opera in its default config.
As for Google (according to their site hits), their stats show that MacOS had around 4% in March 2002.
http://www.google.com/press/zeitgeist.html
Hmm. Zeitgeist, isn’t that a Bruce Sterling novel
“I think Opera might still be the lone exception. IIRC, it masquarades as Netscape by default (this behavior can be turned off).
But if you look at a web server log, you will find that even IE sends “Mozzila” in its ID string. But IE also sends its real identification”.
(I am not 100% sure but) I think I agree.
Many Apache Admin configure it to Mozilla – so Microsoft had to include both ID’s (Mozilla+IE5-6) to improve page Layout and speed.
At the same time, konqueror by default identifies as Mozilla; I do not customize a fresh KDE installation for long time but if I recall it well the default is Mozilla.
On Opera I think the default was IE5 (not Netscape).
… and all fake our IDs to something like “StarWarsBrowser” and let’s see how much we hit next month on google
No one called me for the survey… I must not have been home I guess 🙁
I use IE6 and also use Proxomitron to fake the HTTP headers and Javascript so that it says I’m using something other than IE. This way, it will throw off those websites that try to send you ‘annoying ad x’ if you’re using IE
As far as Mac & Linux are concerned, you can argue and bicker whether it’s 3%, 5%, 7%, or whatever .. but then, they’re both like 100 miles behind Windows any way you slice it, so what difference does it make ?
… and all fake our IDs to something like “StarWarsBrowser” and let’s see how much we hit next month on google
how bout chaing it to say “NetPositive”
Analysis tools like WebTrends identify my Motorola StarTac cell phone as IE 5.0, for example. Most non-Netscape or IE browsers identify themselves as one of the two so they won’t encounter “Internet Explorer is required to view this page.” when visiting websites assembled by morons. Opera has a configuration setting to have it idenitfy itself as Netscape, IE, or Opera for the same reasons. Additionally, what websites did they measure? My website logs show a little less than 60% of visitors were using a browser which identified itself as “Mozilla”.
http://www.bayerwerke.com
For crying out loud, it seems as if everyone here is talking about stuff without ever having seen a httpd-access.log or its equivalent from a web server other than apache.
The “Mozilla” 4 or 5 refers to the browser compatibility (or “family” if you will) string. That is followed by the browser ident string and perhaps by OS/version. IE4-6 are “mozilla4” compatible, whereas netscape4+, Mozilla (the browser), Konqueror present themselves as “mozilla5”.
All graphical browsers will probably have “mozilla” as first string.
Geez.
And, what percentage of people even realize there is another browser?
Statistics like this hardly matter when the reality is 95% of people don’t realize or care what browser or OS they are using. Once computer illteracy rates go down a bit, so many people might not use whatever MS shoves down their throats, becasue they might actually realize they have an option.
I wanted to note something about the google stats –
what is the 5% ‘other’? If that were any compbination of something besides Windows, that makes a statistically significant impact.
PLus,. I like that almost 50% of folks are using Windows 98…the differences between that and 2000 and XP are immeasurable to the average user anyway. Silly, how with WIndows many people never upgrade since it costs money and tyhey engineer it to cost new hardware too, while Linux, you get all upgrades for free on the web… like KDE for instance.
typo, i meant changing. or however you spell it. man, i need get some sleep.
>>they’re both like 100 miles behind Windows any way you slice it, so what difference does it make ?
<<
Where we like it, no reason to stick out like a sore thumb and become a prime target for hackers (or better put, computer vandals), and as long as our technology is better, then that is all I care about!
“Where we like it, no reason to stick out like a sore thumb and become a prime target for hackers (or better put, computer vandals), and as long as our technology is better, then that is all I care about!”
Well, that’s cool … maybe you can convince some of your comrads to come around to your way of thinking, instead of trying to ‘convert’ people who don’t agree that your ‘technology’ is better to begin with.
OK, I see lots of comments but nobody’s saying which browser they use. I use Opera 6, been using it since 5. Great browser. Mouse gestures are great. As always, at some point IE will have it, and all will wow on the “innovation” from M$. I also changed my Opera’s ID to ID itself as Opera, not as anything else. I did it because I’m proud of Opera, and because I’m proud of not using IE. I’m a rebel. I want to prove that there are good things out there. I also left a message to Opera people telling them to ID themselves to Opera by default: if they don’t, web designers won’t take them into consideration, which creates a vicious cycle, etc.
So… TALKBACK!
>>Well, that’s cool … maybe you can convince some of your comrads to come around to your way of thinking, instead of trying to ‘convert’ people who don’t agree that your ‘technology’ is better to begin with.<<
Actually I don’t try to convince anybody to do anything, they can make that decision on their own… I was a PC user and I made that decision on my own when leaving it for the Mac platform. If somebody comes up to me and asks, “hey what computer do you think I should get?”, I will point to either Mac or PC depending on what they say of what they’re going to use it for, I’ve used both platforms and each serves a purpose. But your arrogant remark of a “100 miles ahead” was open for debate, not sure it if was leaning torwards ‘technology’ or ‘marketshare’, but I assumed both! Being that Windows is a forced marketshare doesn’t automatically mean it’s a better platform!
’nuff said
Sorry…
I use OmniWeb 4.1 mostly, but also use Mozilla and Chimera as my alternative browsers. I really like Chimera and I think it will probably become the web browser of choice on the Mac in the near future, if all goes well with the development that is 🙂
“For crying out loud, it seems as if everyone here is talking about stuff without ever having seen a httpd-access.log or its equivalent from a web server other than apache. ”
I know perfectly well what it means. I have written custom log file stats software.
Anyway…
Here are my browser stats for April so far:
87.25% of my visits came from Internet Explorer in general. 58.84% were from IE5, and 26.41% were from IE6. IE4 had 2.84% . There are even a few people out there still running IE 3 0.13% of my traffic came from IE3.
Netscape browsers were 11.25% of my traffic, and Mozilla/5 had 0.85%
So my numbers don’t show that IE is quite as popular as this survey seems to be claiming.
A potential problem with this is the way the information was gathered–what sites are they collecting information on? Onestat.com is not a survey company like Arbitron. They sell statistics-gathering services to web sites. In other words, the only information they can have is the median web usage of the companies that use their services, not those of a random sampling of internet users. There’s an excellent chance that most of those companies are collecting statistics about their internal web sites as well as external–we did at Intermedia when I worked there, and over 99% of the corporate desktops were running Internet Explorer.
You can see the level of discrepancy just by viewing OneStat.com’s own demo page. As of a few minutes ago, it’s got more Windows users (98%), but 7.3% of the viewers used Netscape browsers, not Microsoft ones, and Navigator 6 was more popular than IE 4 or Navigator 3. I doubt those statistics are truly representative of the browsing public at large, either; the point is that it’s open to question how well OneStat’s gathered statistics apply to the surfing public at large. A disproportionate number of intranet users is almost certain to skew the results toward Internet Explorer and Windows.
I selected this myself too. (Nice interface, feature-rich. Not MS.) And I did for the same reason: to let webmasters know that people are using Opera. I’ve sent e-mails to websites mentioning incompatibilities with parts of the site, with a response along the lines of, ‘our website doesn’t support that browser at this time.’ If I went back to a couple of those sites I’d probably find that they’d realized their mistake and would now return a “We’re sorry, but…” for their home page.
Mr. Simba
From your IP you are .edu
Guess your are running a site for *nix users (wich is good), and related, who have an old 486 on their room (11.25% for Netscape ?).
Try to check a log from a .com and you will see real world statistics (no ofense).
“Mr. Simba
From your IP you are .edu
Guess your are running a site for *nix users (wich is good), and related, who have an old 486 on their room (11.25% for Netscape ?).
Try to check a log from a .com and you will see real world statistics (no ofense).”
You make the errornous assumption that the site I gave you the stats from is connected to UNIX or the University. In fact, it is not. The site is a completely seperate site. It’s actually a .org site and is in no way geared towards UNIX users. In fact, it has absolutely nothing to do with computers at all.
Lets not forget that if google reports linux is higher, or non-IE browsers are higher, it may be partial because many linux users use google. Do a statistic on msn.com, and see what the majority of browsers are. Google also has that linux section, which also draws me to it:)
using Moz .99
“Lets not forget that if google reports linux is higher, or non-IE browsers are higher, it may be partial because many linux users use google.”
I don’t think google results would be biased anymore. Just about everyone today knows about google and knows that for general purpose searching, its search algorithms are the best in the world.
Google is given high rankings in just about any article or book geared towards helping people learn how to do online research.
In my animal behavior research, Google is usually the first place I hit before I go to the more specialized scientific search engines. I am also an author, and I just read an article an article on doing online research in one of my magazines on writing (the magazine is geared towards authors in general. No specific field). The first online resource they suggested for doing research is Google.
“I don’t think google results would be biased anymore. Just about everyone today knows about google and knows that for general purpose searching, its search algorithms are the best in the world.”
Just some data to support my speculation…
I checked my web site logs and Google is listed as the #1 refering site for traffic. So basically, the vast majority of people visiting our site got to it by finding it on Google.
And like I said, the site is not related to UNIX or computers in any way. So it shouldn’t be biased towards Linux users.
These numbers seem to be fairly accurate. Perhaps I would say that the IE usage reported is a bit too high, but not much.
According to my own weblog (a music site), most people are using IE. IIRC the logs show that about 85% are using IE, and about 12% are using Netscape and the rest is other browsers such as Opera. That is if I don’t count the spiders and bots, because they have become a large portion of the traffic lately
Surprisingly, the logs show more BeOS users than MacOS users(and I’m not counting myself), but I seriously doubt that it would apply to the real world, but it’s strange anyway, since I haven’t promoted the site in the BeOS community at all.
As for my own usage, I use Net+ about 90% of the time, Mozilla 8%, and Opera 2% when I’m in BeOS. On FreeBSD it’s 80% Links and 20% Mozilla. In windows, which I rarely use for browsing it’s 99% Opera 6 and 1% IE.
I think the main reason that so many people are using IE is because it comes with the OS(at least the one most people use), and they generally don’t have any problems with it. People that doesn’t have any problems with a product rarely looks for alternatives, and those who have problems often thinks that this is the way it has to be and accepts it.
And then theres the small amount of powerusers who uses their computers many hours per day. They mostly look for alternatives to make their computing more effective, since they are tired of waiting a long time for webpages to load and render (this is why I use Net+ most of the time) or constantly having to move the mouse to the toolbar and press the arrowkeys(Go Opera!).
I personally spoof with Mozilla to be able to do online banking, but I doubt that there’s a large amount of people who spoof, I guess that it’s mainly Opera users who doesn’t know or care that Opera is a sheep in wolves clothing.
what is the webbrowser stats of OSNews?
I’ll never go back…. Opera Rules! The rest just droolz.
Why so many get “hot” when the statistic shows IE got the highest rating? It might represent the real situation since the world largest share of desktop OS is conquered by Window$. Majority Window$ user just use whatever bundled with the OS especially IE.
If we don’t like the statistic, let we ourselves start to improve other desktop OS so that the user will not browse using IE. Mac OSX is a great desktop OS but not so many througout the world can afford it. So the best thing is the cheaper alternative such as OSS OS.
* Don’t forget that pirated Window$ are available everywhere and nobody can give the exact statistic. So out of 97% who can say how much is using real IE rather that “free” IE. I think for Mac at least it is confirm the hardware is Mac although the OS is pirated.
>what is the webbrowser stats of OSNews?
I *just* uploaded this image, captured from Webalizer, our internal statistics program, for April 2002.
http://www.osnews.com/images/osnews-web_stats-April_02.png
Please note that OSNews receives lots of hits from alternative OS users, because of its nature. However, IE is way ahead even with all the alternative OS users visiting us.
Alternatively, you can check our secondary statistics engine, at Counted.net (however Counted loses about 6-7% of the overall stats because *sometimes* (randomly) their javascript and icon doesn’t work with some browsers). These stats are only for April 30:
http://66.181.171.71/2/42699/400/?date=0
Also, please note that Counted has a different algorithm parsing the browser string, as Webalizer has.
I also got numbers closer to the ones posted in an
earlier message.
April stats (entire websites):
http://www.reimeika.ca (mostly manga/anime – 426 pages)
MSIE = 85% (total hits from all browsers = 13000)
http://www.math.toronto.edu (UofT Math site – 8369 pages)
MSIE = 70% (total hits from all browsers = 110000)
The rest (15% and 30%) are non-IE browsers.
Perhaps IE doesn’t permeate geek/academic browsing
as much. Then there are small lies, big lies, and
statistics…
> Please note that OSNews receives lots of hits from alternative OS users, because of its nature. However, IE is way ahead even with all the alternative OS users visiting us.
Yeah I know, that’s the reason I was interested
The numbers of IE visits was a little more than I expected and the MacOS visits a little lower than I expected.
I dont see mozilla or netscape….. i figured it would show mozilla itself more and not just that mozilla string that all the browsers seem to have. The webalizer on my website tells the difference between mozilla and the others… i must be reading it wrong.
anyways i use mozilla most all the time
cant wait for 1.0!
“FreeBSD 4.4
LuZero, about the fake of identity. Not many people are doing this (faking it
, mostly geeks and people who understand what
the heck that browsing string is. And that would not really change the overall stats too much. Their percentages are too
small anyway.”
The browser I use spoofs as IE by default. I would think this is
essential for regular use of any other browser.
Many sites are only accessible if the site thinks you are running IE.
Nowadays the browser even has to pass clever Javascript tests to prove
it is IE before you are allowed in.
How can anyone possibly know what browsers are actually being used by
counting the results from web sites?
Is there anykind of non-changeble tag that could be built into the OS or browser that says this computer is this OS and this is this browser. Maybe something like this is built into the network stack already. Seams like if there was something like this and no one could go in and change it, it would make this easy. Actully have the OS identify itself through the network and then the browser since you may use more than one browser. It sounds like this is basicly what there is but since people can change such things, (make opera say its IE, or Net+ say is IE) this system becomes worthless. Granted there is reasons for changing the id since there are sites that block browers (note to if anyone running elion for PSU is reading, Stop blocking browsers other than IE and Netscape!).
Also seams like the aproach to usage is wrong, they look at what people are using while looking at certain sites. How about getting ISP’s and such have their routers track this and report back to one central computer and compile the results that way the whole linux people go here, mac head there, beos users to benews type problem would be gone. Unfortently this would involve co-operation of ISP’s.
I don’t think results for OSnews are very valid, like was said there is way more variety coming here than the normal world has. If this place was correct apple would probly have 25%+ market share.
the survey is about browsers and not platform. mac might have a 4-5% market share. but they do use IE5.1 by default.
I know that I do a lot of browsing at work due to the high bandwith versus my 56K connection at home. Most people don’t have the option of using an alternative OS at work, and might not have a choice in the browser they use either. I wonder how much the work-browsing aspect might skew the results.
I know that most Amiga browsers fake their identity to be either Netscape or IE, this is done to offer more compatibily. BTW the Amiga browsers currently available, may not be the best up-to-date browsers available, but they surely are among the fastest browsers around. A new version of IBrowse will come bundled with AmigaOS4.
more ppl “view” osnews with wget than NetP
Opera 6 is my favourite curently too.. It does have one or two bugs, like the failure to diffrentiate between links that open in new windows and pop-up banners.. it blocks both or neither.. and a refusal to work at all in some Java chatrooms ( MSN Communities chat and messageboards wont work either ) for everything else its great.
well, I’m not that much into the point of arguing which browser is the best, IE or Mozilla. I just know Mozilla looks great on my monitor and that it’s running perfectly (RC1). IE6 is kinda heavy and its parent IE 5,5 was way better and sharper than the piece of fat IE 6 is.
However, and just because I feel like bringing some kind of geeky-nerdy fun, I would recommend to all of ya folks to use a long forgotten but still workable browser:
The HOT JAVA browser 3.0 from Sun Microsystems, last released in 1999. You can dowload it from the Java Sun web site. Many features of today browsers are missing, but there’s even a mail client inside! Have fun!
Wow! That’s a ton of hits!
It is weird to see that Opera only have 0.73% in osnews’ stats. considering how “many” claim they are using it. But, then again, many people browsing with IE would care less what browser they use and not mention it.
And yes, I run Opera too and let it identify as Opera. Perhaps we can improve these stats.
What the heck? I know exactly NOBODY that uses IE 6.0. That’s because I know exactly noone who uses WindowsXP, I guess.
I use Opera in Windows and QNX, and Konqueror in Linux. No spoofing.
This does not take into consideration misidentified or disguised Browsers in Linux., nor does it take into account how FrontPage mounted sites will deliberately and falsely identify Linux browsers that are capable of reading them as IE so as to provide stats for MS propaganda. I simply do not believe these stats. If they really were true, however, that would mean that much more of the world is much more stupid, uninformed, and credulous than I thought.
Both of my banks and some of my financial sites will not let me log in when I use the Opera 6.0 browser unless I tell Opera to masquerade as IE 5.0.
I am delighted that Opera has this capability, but I often forget to change it back so it identifies itself as Opera. So I may very well have visited sites that incremented the IE counter when they should have incremented the Opera counter.
I wonder how many other ways there are to artificially inflate Microsoft’s perceived market share.
-Ken Jackson
There are 2 types of user out there, us & them.
Us does the SW & HW & OS development, them is all our clueless relatives & other strangers we don’t know.
Us cares alot about the SW, HW, OS, browser we use but we only make up 1 or 2% of the whole PC world population. Them uses the 1st thing that works & cheap as possible but them also pays our salaries for what us makes. Also remember when only us bought computers, they were 10x or so more expensive, so them also allowed HW volumes to reach saturation & lowest prices possible.
Most all the stats reflect us & them bundled together, so we have to put up with the world looking very sorry. Personally I only care what the us stats are & I bet MS is only 50% of the scene since most of us uses atleast 1 OS & knows a few others. The general stats can never reflect us properly except site stats like OSnews. Even if I do all my work 8hrs a day on BeOS or whatever, I still generally surf more on IE for 1 reason or another (ok it can handle languages I could never get to work elsewhere).
Just my 2 bits
The numbers don’t mean anything, which is why these guys keep shouting them louder and louder.
Imagine if a company told you that BMW was the most popular make of car on the road. Imagine if further you discovered that the “method” used to discover this was that car parks in four parts of London, in the UK were being surveyed and all other data was ignored.
The web stat companies don’t measure stats for any serious sits (for example, they measure OSNews, which is a pretty trivial little corner of the web, but not Google, nor Microsoft.com) and therefore their statistics are meaningless.