Slashdot Mirror

According to Wikipedia's page on usage share and Chuck Upsdell's browser trends page, mobile browsing accounts for under 10% usage. Mac and Linux users also take up under 10% usage. Most of the non-IE usage is people using Firefox or Chrome on Windows computers.

fair enough, how about a different survey.

http://upsdell.com/BrowserNews/stat.htm

it also warns against taking the numbers as verbatim but it is using 5 sources. this is a problem with browser market share. how can you tell a good survey? how can you correct for browsers reporting as something else?

the numbers from this site aren't much different than the other survey. what do you work at? how often do you see ie6? hell i've seen ie4 and ie5 in the past month on old servers. within the last 18 months i've seen win3.1 on a laptop. those are exceptions but ie6 is still very very common.

Most that work with browser issues on a daily basis considered IE6 dead and IE7 a spasm.

IE usage is closer to 80%, but it is still dropping. Give it a few more years, and it'll be down to 70%.

How often did you hear this argument last year, when the trend lines for Firefox were looking very, very good?

It was the trend lines that interested me here, But alternative sources do not paint a brighter picture; Usage share of web browsers, Browser News Stats

At 800x600 I have to use the horizontal scroll bar to read that page. It is very annoying. Almost 50% of the world uses 800x600. It might be that your clients would have higher resolution, but for commercial design it best to try to satisfy at least 90% of all visitors.

Almost 50%? Perhaps in 2002 but browser news peg 800x600 around 14%. Since his target market is english speaking people with an interest in games I am sure looking at his site statistics will be well within the 90%

Mozilla was more popular than Opera soon after Mozilla 1.0 was out (in 2002) and Phoenix (as Firefox was known back then) was just starting. Firefox is not more popular due just to hype, as Mozilla browsers were more popular before the Firefox hype ever started. As far as I can tell, Firefox use is still climbing, as is Safari use, and Opera use is holding steady. This same pattern has persisted for years. I don't see any evidence that many Firefox users are switching to other browsers.

Use what other webdevs use statistics. You should always try and make sure your users can access your website, without users what sense is there to even bother making a website? Whatever the majority of your important users are currently using for software is what you should be coding your site for.

You could also analyse your visitors along with the statistics found here.

http://www.upsdell.com/BrowserNews/stat_trends.htm

So what your saying is that the snotty conceited wannabe guy is still insanely more popular than you, despite your attempts to tell the whole school that you're actually smarter, and therefore better?

I think the no-friends thing might be a denial of the reality

Additionally I believe that your entire post is a freudian slip belying some serious unresolved social issues

(Disclaimer: This post created in Firefox.)

For you information:
http://www.upsdell.com/BrowserNews/stat.htm

KHTML based : 2.9%
Opera : .4%

so before considering about IE1 and 2 you should consider these great KHTML browsers first.

Yeah, but how do you know that WebSideStory's numbers aren't taken from sites that attract mostly Windows-only users? For all we know, some of the sites that they monitor could exclude all non-IE browsers.

That seems highly unlikely, given that many of WSS's customers are large global corporations that have big, professionally-created websites- Disney, Best Buy, Fox News, Bank of America, Freddie Mac- a wide range of clients in a variety of sectors. Surely some of their sites exclude non-IE browsers, but they are unlikely to make a significant difference in WSS's numbers.

But my point was not that WebSideStory's numbers are accurate, although by the fact that it is their business to know this kind of stuff and they seem to have been around a while I tend to believe them, but merely that the visitors to your site are not likely to be anywhere near a representative sample of the internet at large. There's lots of info out there that seems to corroborate WSS's numbers. Your little window on the world is nowhere near the big picture that WSS is seeing.

Google used to list this stuff in their zeitgeist, but they seem to have stopped that. Too bad...

There are a few stats services that can mistake Opera for IE, but not many. I doubt this move will help Opera's stats much. Even taking into account under-reporting of Opera, Chuck Upsdell estimates Opera usgae at about 1-2%.

So I think to really get some meaningful stats about installed bases for Firefox, Opera, Linux, and the like, we must survey lots of sites (1000's) from all manner of target demographics. Any other method of statistical analysis would have some bias.

Sorry for the munged links. corrected:

Source 1: http://www.w3schools.com/browsers/browsers_stats.a sp
Source 2: http://www.upsdell.com/BrowserNews/stat.htm

IE doesn't have over 90% of the market. Currently the number is somewhere around 60-80% and falling. Still too much, but the situation is improving.

Source 1: http://www.w3schools.com/browsers/browsers_stats.a sp/
Source 2: http://www.upsdell.com/BrowserNews/stat.htm/

74%, according to this.

Do not mod up, this is copied from another posting.

IE: 84% and falling
Mozilla: 7% and rising
Safari: 1-2% and rising
Opera: 1-2% and holding steady
Netscape 4: below 1% and falling

• Browser news puts Gecko-based browsers around 5% for most sources as of 4 Sep.
• The latest publicly-available stats from OneStat indicate just under 3%, but that was back in May.
• WebSideStory/StatMarket's latest numbers are still the ones used in a July C|Net article putting Moz at 4.5%

Here's some statistics from a different source (which actually presents stats from 5 sources), where Gecko (mozilla) ranges from 4% to 27% - it's clear that the stats greatly vary from site to site:

http://www.upsdell.com/BrowserNews/stat.htm

I feel compelled to warn that the W3Schools' browser statistics tend to reflect the browsers that web developers use more than the browsers that web consumers use. In order to get a more Fair and Balanced(tm) view of all your install base, you'll need to find a site with broader demographics than W3Schools or, better yet, a site that presents stats from multiple sites.

Combining Netscape 7 in there - that's 13.7% for mozilla derivatives. That is impressive if it's true.

I'm wondering where they got these statistics. I couldn't find any discussion of this on the site. They don't seem to jive with the Google Zeitgeist. In the Google Zeitgeist, all Mozilla variants seem to fall into the "other" category. At the May 2004 point, the other category seems to have a lower percentage than both IE 5 and IE 5.5 on Google. In contrast, the w3schools statistics say that Mozilla by itself is higher than all IE 5 browsers combined. Is this really only visitors to the w3schools site?

Also, looking at the second hit returned from a "browser statistics" query on Google highlights the uncertainty around this. They list 5 different sources of browser statistics that vary wildly. Here's the link.

With all the hubub over IE, it seems like there would be more discussion over this. Is the only way these statistics are being collected is from the useragent string? Are there any statistics on how often this string is being spoofed?

First thing I found was http://www.osnews.com/comment.php?news_id=6013 So far, so good: Linux has 3.2% of desktop share and passed Apple according to that. Another good read is http://www.ecommercetimes.com/perl/story/32706.htm l. No definite answer there. A quote: 'According to The Linux Counter, there are probably somewhere between 2,747,850 and 68,689,500 Linux users worldwide.' Great.

So maybe I can figure Linux %% out from some browser stats... http://www.upsdell.com/BrowserNews/stat.htm gives some info but its stat sources may produce rather biased results (imo). Since Google is Google is Google I trust it. So here's what I see: http://www.google.com/press/zeitgeist.html Can't be Linux is only 1%... lets look for something else.

Next thing I found thecounter.com - a web util which lets you add counter to your pages, they also publish stats from their hits. If you want to take 2 minutes and compare 2004 march results (http://www.thecounter.com/stats/2004/March/browse r.php) and eg 2003 january results (http://www.thecounter.com/stats/2003/January/brow ser.php) then you may see strange things there: linux users went down from some 0.42% to 0.29%.

I give up here. Now before you mark me as flamebait - I know there are some possible explanations like faking UA to prettend windoze. However I wonder what is reality: 3.2%(OSnews estimate) or 1%-0.28%(Google+some webcounter log data). That would be some 3/4 linux users faking UA.

Loads of people are still on Internet Explorer 5 today, that was released over five years ago

Actually, most sites that track this sort of thing show about 75% of all users on IE6, but only around 10% on IE5.

Some examples:
http://www.thecounter.com/stats/2004/January/brows er.php
http://www.upsdell.com/BrowserNews/stat.htm
http://www.w3schools.com/browsers/browsers_stats.a sp
http://www.cen.uiuc.edu/bstats/latest.html
http://www.webreference.com/stats/browser.html

As a consultant that works with many clients, I'm truly surprised at how many companies still use NS 4.x as there minimum browser standard (even for intranet apps). I can show them all sorts of statistics that NS 4.x is used by less then 2 % of internet users (Broswer stats/trends) and even less on there sites, it still doesn't matter. It takes me more time to test/code around the shortcomings of NS 4.x then to test/code the entire site.

But the clients pay my bills, so that's what I do.

Firebird has won the browser wars? That's quite the statement to say, since it seems IE still has a good ~85% chunk of the market. I think it still needs to carve quite a big piece out of Microsoft's share before we can start claiming any victories.

00.25% - MSIE 4.0
11.85% - MSIE 5.0
06.88% - MSIE 5.5
73.89% - MSIE 6.0
04.33% - Gecko
03.18% - Opera
00.51% - Konqueror
00.51% - Other

Hopefully this will make companies realize that the Internet isn't comprised of just IE users.

Well, not quite, but the last time I looked, (which was about 9 days ago) about 95% of internet users used Internet Explorer for their web browser. While that doesn't comprise all internet users, it's close enough to make no odds.

majority? Are you sure?

Just because I said the UI is a steaming fucking pile doesn't mean I said you shouldn't use it.

Secondly, learn to take some criticism. I don't use the browser I use because I hate Opera (what a bizarre position to take...), I use it because I think it's better than anything else for my needs. I also happen to think that it's better than Opera for most other peoples' needs, but I'm certainly not shoving it down anyone's throat - these are just my opionions.

Finally, don't quote numbers to me. 6 million downloads (let's not confuse the number of downloads with the number of people who actually keep and use the app, thanks.) for a commercial product is not bad, but it's a tiny percentage of web users. According to Browser News, Gecko based browsers are way ahead of this.

I spent 3 years doing tech support with IBM for almost two dozen different companies, including no less than 4 banks. You would not believe how much the banks will compete to keep you as thier customer, and if it means making their browser Netscape/Mozilla compliant, most of them will do it. After all, for the pidly sum of $38,000, they can expect to keep at least 5,000 customers happy (and you know i'm guessing on the very very low end of the scale). 11.5% of the browser population is still a large part of a business' clientel. Another way to look at it, a user is just as likely to be gay as to use a non-MS browser. I'm just sayin'....

DavinciII's comment is modded insightful, but, as others have pointed out, he uses bogus statistics to make his case. This is so wrong.

He cheats his clients and he cheats his clients' potential customers.

Look at the stats from Upsdell's Browser News. He compares several sources, and it is a stretch to say MSIE commands 95% of the browser market, especially if one excludes older versions. Even an estimate of 90% for IE 5 and above would be discounting reliable evidence to the contrary. And that's not even accounting for the way text mode browsers are systematically undercounted.

If davinciII were honest, he would tell his employers something like

$50k for between 88% and 95% of users
$100k for any w3c standards-compliant browser, or just about anybody

But really, what are the real numbers? How many people are 4% of internet users? or 7%? or 10%? Right now that's about 25 million, 45 million and 62 million, respectively. Are those numbers that matter to davinciII's clients? I should think so, but davinciII's arguments are more insidious. He suggests that people not using the latest offerings from MSIE are using old technology. The implication is that they are poor and will not buy anything. They represent "junk traffic," and excluding them from the clients' websites is being touted as an added value.

Ick.

1. Sites that rely on graphics for navigation are disadvantaged from the git go. For sites that people visit daily, it's probably no big deal since the graphics can be cached. For attracting new users it's far from optimal.

2. You should understand how people surf. People bail on sites that take forever and a day to load. Upsdell's site has a concise summary of some older research. It is still relevant. Here's another report on the Zona research. I think if you tested this yourself you would reach the conclusion that most flash animations are not worth waiting for. There are other sites out there, easy to find and easy to use. If your site depends on flash for basic funtionality you *will* lose visitors to your competitors.

Did you happen to catch the Flash Usability Challenge at Webword? Now I can't truthfully say that flash is completely worthless, but it hardly looks like the future of the web, does it?

BTW, I don't have flash installed at the moment. I went to your site (numbera.com) and saw no compelling reason to take your advice and install it.

Bye-ya

You are lucky. Look at Charles Upsdell's browser stats or web browsers used to access Google for more typical stats.

The browser stats mentioned are collected by hitbox.com, and come from webbugs. They're used by sites to determine which browsers to market at (at least, thats what hitbox themselves say you should use them for)

If your site gets 80% IE hits, so you decide not to worry about NS compatibility, your site will not get visited by NS users. So your % IE hits will go up.

Doh. The only way to get fair statistics is from a site used by everybody which doesnt discriminate against browsers. The closest thing I know of to this is Google's zeitgeist.

The browser graph doesnt have figures, but relative proportions are obvious:
IE 6 gets about 8 notches
IE 5, 5.5 get about 5 notches each
NS4.x/other seem to have dropped from 2 notches each a couple of months back to 1 notch each;
IE 4 gets about 1/2.

Dividing by 20.5 to get market share:
IE6 39%
IE5.5 19.5%
IE5.0 19.0%
NS4.x 4.8%
Other 4.8% (includes Moz, galeon, etc)
IE4.x 2.4%

If you look at this statistics comparison you'll see that the guesstimates up there are not in violent disagreement with reality, with the exception that the 'source1' stats (assumed to be hitbox or a hitbox-a-like) are scoring low for NS and 'other' compared to all other sources now including google, whereas upsdell.com itself sees an excess of Mozilla hits. I reckon this all adds up to a classic case of self-selection.

• The most pessimistic estimates show that all flavors of MSIE combined have a 93.7% share. MSIE 5 and greater comprise about 90%, and that's really the relevant number because nobody codes explicitly for outdated versions of MSIE. If you're coding for current versions of MSIE only, and you violate w3c recommendations in such a way as to exclude other browsers, you're cutting off at least 10% of your potential readership, which is twice 5%--a huge difference.
  
• Note that other sources at Upsdell's claim significantly less than 93.7% for all versions of MSIE combined. Why would you choose the most extreme number and then round up, unless you were trying to justify decisions made for other reasons, or carelessly spreading fud and hype. The wise thing to do would be to pick a middle-of-the-road estimate, or to average various estimates. Quoting numbers like 95% makes no sense at all.
  
• The variety of agents and platforms currently in use should testify to the fact that change and diversity is part of the global computing environment. Open standards enable communications between any and all programs that implement them. Proprietary "standards" are an obstacle to the global communications infrastucture, and should be viewed as such. MSIE html appears as a "de-facto" standard only by suppressing equally compelling facts, including the temporal horizon, which ought to be in the forefront of any serious thinking about software developmnet.

CNET quotes the OneStat survey from earlier this Spring showing about 97% market share for all MSIE versions combined. This is an unreliable survey, and by citing it CNET shows a bias towards Microsoft. Read the OSNews coverage of the OneStat survey. There are some insightful comments in that thread.

Here are some more trustworthy numbers. 93.5% for MSIE is about as high as you can safely estimate. You can also check out Google's graphs, which suggest that any estimate of more than a 93% share for all versions of MSIE would be innacurate.

I've been using Charles Upsdell's Browser News for reliable browser stats. According to that site, Mozilla/Netscape/K-Meleon/etc. browsers currently have a 1.1% share of page hits, IE has about 92%, and Netscape 4 about 4%.

At lest I think I know how the tv commercials affect my anyway. I watch some program and in the commercial hear about new improved detergent. Later when I go down to the store to buy some stuff, and I need detergent. Say that I see a couble of brands, some of them are unknown to me, but there also is the detergent I have seen commercials about. I most likely would buy the product that I at least had heard about - even though it only was trough some unbiased commerical.

The problem with Internet ads is that somewhere along the way the "interpretation" of the Internet has changed from "a big library" (rember that?) to "a big shopping mall". (At least for the banner makers.)

However, in a real shopping mall, an "on sell" sign might lure me into some speciffic shop to buy someting. This while most users don't interpreat the Internet this way - like most banner people would like to think. To the users the web is still more of a library than a shopping mall.

Have a look at:http://www.upsdell.com/BrowserNews/stat_des.htm #d04

(Figures a bit old, but I think the main point still applies)

The point is that when I go to a shopping mall I went there to buy stuff, much unlike when I surf to my favourite pages. And they wonder why I find thouse blinking banner annoying? They are annoying because, most of the time, I didn't log on to buy anything.

So, to some extent I think that web ads sould be more like the tv commercials - "get this brand into your heard" kind of ad. But, the success of such ads are of course much harder to measure. Even though computer ads supposedly were to be the dream for copy writers et al. (Should be so easy to measure exposure etc etc)

Also, though it has been mentioned above, just clicking a banner on a site that I'd like to support would not help much in the long run. I'd of course have to click and buy.

http://www.upsdell.com/BrowserNews/stat.htm
Untill these numbers change, Microsoft can do whatever it likes... and I don't need to worry about some 0.25% of average users using Mozilla.

I gotthe stat from Browser News, but that stat is based on the TheCounter stat. Nevertheless, there's a good deal of other interesting statistics here, like support for various plugins and such.

Well, Browser News has graphs generated by TheCounter.com indicating around 10% market share for Netscape and about 80% for IE, though my personal experience suggest that the NS share is higher reality.

Naturally browser stats differ greatly depending on the content of the site. microsoft.com probably doesn't get that many hits from lynx, but linux.com probably does.

-=Sam=-