Guest post by According to data compiled by StatCounter, Firefox overtook Internet Explorer (IE) as the dominant browser in Europe at the end of 2010. During December, 38.11% of internet page views involved Mozilla’s Firefox web browser, whilst Internet Explorer fell back to second place with 37.52%.
According to StatCounter, Firefox’s share decreased from over 40% early in 2010 to 38.11% late in the year, but Firefox still overtook IE due to a steeper reduction in IE’s share, probably due to a corresponding increase in Google Chrome.
It will be interesting to monitor ongoing trends in browser market share when the newer, more competitive versions of IE (IE9) and Firefox (Firefox 4.0) become available in the near future. If Firefox 4.0 is compelling enough, Firefox may be able to hold on to its recently-gained #1 ranking, especially considering that Firefox 4.0 will bring advantages to Windows XP users where IE9 will not.
In this statistics, Firefox is marginable ahead of MS IE. This was in December when many people were more at home that at work, hence it is likely that more people use the browser of their own choice rather than their company’s choice. Do not be surprised if MS IE get slightly ahead again in January.
It does not really matter who is leading, and this kind of statistics is much too imprecise to talk about fractions of percentages. What is important is that MS IE is well below 50% and there is both Firefox and Chrome that have significant market shares.
The result is that web developers can no longer code for just one browser, they will have to code following standards to fit to many browsers. Finally! Now we just need the same trend to happen in other countries as well.
Edited 2011-01-06 23:07 UTC
We need the same trends to happen in operating systems too
Edited 2011-01-06 23:44 UTC
http://gs.statcounter.com/#os-ww-monthly-200912-201012-bar
Maybe it could happen. If someone were to release a new OS and market it as aggressively as Google has marketed their Chrome browser. Even so, not everything that’s hyped and advertised is a success; people still have to like the product and be motivated to use it.
So invent new operating system;
Flood the airwaves and internet with advertising and hype;
Give it away for free;
Work deals with OEMs to get it pre-installed and on store shelves;
Hope that people actually like it. That’s the tricky and unpredictable part.
That has been done. BeOS was killed by Microsoft because step four is not permitted by them.
BeOS actually shipped on some computers along Windows, however Microsoft forced the OEM to use their boot loader, so BeOS could not be chosen without booting the computer from a floppy.
Naturally, that didn’t last long.
Microsoft still employs these tactics today against OEMs. Dell had a problem back in the day because they were considering putting Linux on Desktop machines as a choice – Microsoft threatened to cancel their Windows discount – meaning instead of $35 for Windows, they would have to pay $109. Big difference!
–The loon
I’m sure Jean-Louis Gass~A(c) had nothing to do with it.
Funny how M$ hasn’t been able to force OEMs to only bundle Windows Server.
There are plenty of small shops that do not buy from M$ directly and yet don’t offer Linux desktops. Funny that.
As for Dell they had problems with Ubuntu breaking machines with updates. As much fun as it is to blame M$ the situation is a lot more complicated. Linux would easily have at least 5% of new computer sales if early on Linus had allowed binary drivers and set some basic standards for distros. Linus puts open source above market share and doesn’t care about making a good desktop OS so maybe you should start complaining about him.
Servers are a completely different kettle of fish…
Not only are a significant portion of servers bought with no software whatsoever, but linux is also a very significant player in the server market and thus MS don’t have as much leverage here as they do over desktops.
The difference is that there is demand for Linux on the server.
WS 2008 is $700 which is the price of a cheap 1U server.
Win7 is about $50 after trialware commissions and likely costs OEMs less to support. Linux also has issues that need to be resolved and it needs a better office suite.
Edited 2011-01-07 15:33 UTC
It didn’t help that Be held out for too much money when negotiating with Apple. As successful and nicely Unixy as alternative (NeXT) has turned out to be, I think if Be had agreed to Apples terms or not demanded SO MUCH more it would have been better for Be, though probably not better for Apple (that would depend on whether or not they got Jobs back at some point).
It’s not even as easy as “Dell pre-isntalls broke because of Ubuntu updates and Linux did not allow binary drivers”. Dell should have chosen a better distribution that delivered stable updates and superior hardware support. Ubuntu was not the best new user distro available at the time, just the best marketed.
Servers are also another beast entirely. Microsoft never went to court claiming that “servers sold with no OS will only be used to run unlicensed Windows copies” where Mr Gates did say exactly that in relation to consumer desktops with no OS or alternative OS pre-installed. Why compete in a fair market when one can manipulate the legal system to block competition.
How is Apple selling $1600 laptops without Windows installed? Is that also part of a conspiracy?
The Linux desktop isn’t even that popular with geeks. Forget OEMs, figure out why most tech site readers are running Windows.
”
How is Apple selling $1600 laptops without Windows installed? Is that also part of a conspiracy?
”
I really can’t take you seriously when you ask such questions. We’re obviously talking about pre-built retailers selling machines with windows; a category Apple clearly does not fall within. Trying to drag Apple into your defense blatantly dodging the point. But, if you need it broken down:
First, I simply pointed out a historical fact that Microsoft original claimed that any desktops sold without an OS would only be used to infringe Microsoft’s copyright; ie. run unlicensed copies of Windows. This was the defense offered against anti-trust investigation over OEM contracts requiring Windows be installed on each unit shipped out the door regardless of the consumer’s choice to buy Windows or not. It demonstrates a history of manipulating the legal system to remove consumer choice and fair competition (“fair” in legal terms). Later evidence like the Halloween documents demonstrate that Microsoft has continued to exploit backroom strategies against fair competition. And, what does Microsoft do now that they can stipulate Windows only as a contractual obligation? They now instead threaten to remove discounts from OEM who dare sell pre-built machines without Windows on them. 35$ per unit for Windows becomes 109$ per unit for those who don’t comply. “gosh, you have a nice shop here.. it’d be terrible if something happened to it.. we can make sure something doesn’t happen to it.. just do what we tell you and your shop will be safe.”
(Coke and Pepsi do the same; deep discounts evaporate if you sell both products.)
Second, Apple is a hardware manufacturer who produces it’s own embedded OS to run on that hardware. It is clearly not a pre-built retailer who sells Windows based systems. Again, why would you try and drag this straw man into the discussion?
(you know, questioning how Microsoft does business does not invalidate your personal choice to use Windows.. just encase that was a point of confusion for you.)
The point is that Apple is making plenty of money selling non-Windows notebooks and desktops. Whatever Microsoft did in the 90s hasn’t stopped Apple from gaining nearly 14% share in the US. That percentage would be a lot higher if they allowed OEMs to sell it OSX on whiteboxes. Dell in fact asked Apple if they would license OSX.
Regardless of what Microsoft did in the past Linux has still had plenty of opportunities to advance on the desktop. It is FREE after all and anyone can download and install it. When you have a hard time giving your software away for free you can’t blame your commerical competition.
Blaming Microsoft at this point is just a cop-out.
Steve Jobs thinks that this is not true. He knows about the long-term situation, more than most of people. That’s why they are not licensing OSX to other companies.
Microsoft has been using its power to avoid that the other advance. That is why Microsoft was found guilty and convicted, at least three times.
Microsoft pays people to create problems to others, as you can see in the following link and in the antitrust case, in trials in USA and also in Europe.
http://antitrust.slated.org/www.iowaconsumercase.org/011607/2000/PX…
Amen, brother! Now I just wish all of the Apple fanbois could learn this lesson — you know, questioning how Apple does business does not invalidate your personal choice to use Apple products.
Edited 2011-01-08 18:19 UTC
Microsoft didn’t make it easy, that’s for sure. And neither did Apple. But people who wanted run BeOS were certainly able to; I’ve been using it since the year 2000 when PE was released.
Ditto Linux. If people really want to use Linux, or even buy Linux-loaded computers, they could. Easily.
But yet BeOS is dead and Linux is holding a steady ~1% of desktop users.
Makes me think the real hurdle is this: people still have to like the product and be motivated to use it.
Nitpick – Linux is not only for desktops. Linux probably has the major share of servers and handheld/mobile devices, and it is dominant in supercomputers and embedded devices (such as TVs, PVRs, GPS units, routers and switches). This alone makes Linux by far the most numerous OS in use today.
Furthermore, depending on how you count it, Linux has perhaps 5% of the desktop share, and growing.
http://www.omgubuntu.co.uk/2010/12/linux-hits-5-marketshare/
Apart from that, carry on.
That 5% comes from w3schools, a technology-oriented website. I’m surprised their stats don’t show a much higher Linux user base.
Statscounter shows Linux at 0.75%. Supposedly their statistics are collected at 3 million different websites.
I would agree that there are quite a number of stats collecting agencies which are quite determined to show an apparent very low rate of Linux usage on the desktop. I would even point out that stats collecting agencies wishing to show a very low rate of Linux desktop usage are far more common, and considerably better funded, than those (such as w3schools) who do not have any such agenda.
Actually, there seems to be quite a number of Internet sites who appear at first blush to be Linux-related, but when you read their articles they always somehow manage (with remarkable consistency) to paint a negative picture of Linux particularly on the desktop.
None of this has anything to do with the observation that Linux is by far the most common OS used today.
Edited 2011-01-07 04:57 UTC
It’s proving tough to beat in the mobile market, that’s for sure.
And in the supercomputer market (91.9% Linux)
http://www.top500.org/stats/list/36/osfam
and in the server market (by units, not revenue):
http://en.wikipedia.org/wiki/Usage_share_of_operating_systems#Serve…
On netbooks and tablets, Linux originally had 90% but Microsoft tried to kill it by allowing retailers to only sell XP
http://en.wikipedia.org/wiki/Usage_share_of_operating_systems#Netbo…
This attempt by Microsoft was sufficient to drop the Linux share on netbooks and tablets to about 32%.
However, with Android 3.0 ARM tablets soon to become available, I would predict that this share will increase once again to somewhere over 50%.
With desktops becoming less and less important overall, this is beginning to get very interesting. Perhaps one day it will no longer be impossible for Linux to claim a significant share of the desktop as well, just as it already has in the rest of the domains where an OS is used.
PS: It is interesting to note that the netbook and tablet market is estimated at 20% of the total market, and that Linux has 32% of the netbook and tablet market. That is 6.5% share right there. It is interesting that some statistics counters still try to insist that Linux has ~ 1% share in light of simple observations like this, is it not?
Edited 2011-01-07 05:14 UTC
For netbooks, it wasn’t only Microsoft’s efforts. The first netbooks shipping with a further crippled fork of about the worst possible distribution choice ever didn’t help even a little.
Exactly so. The choices of Linux distribution for the earliest netbooks were consistently about the very worst choices that could have been made.
It beggars belief that one on other OEM couldn’t have put a decent Linux distribution on a netbook. However, if one Linux distribution works on a given machine, then any other distribution can also work. It wasn’t long before an entire cottage industry sprang up were individuals or small groups of people were putting far better distributions on netbooks than the OEMs were offering.
One has to ask … how is it that a single person or just a few individuals working part-time without documentation could come up with a far better Linux distribution for a given netbook machine than the original maker of the machine could?
Linux had 90% market share, that is until they had competition. Once XP netbooks became available, and were priced competitively…
Once Microsoft insisted on XP and twisted the arm of netbook OEMs to provide XP and remove Linux, people could no longer buy Linux netbooks in ordinary stores.
Seriously, one weekend I went to a store and the only netbooks they sold were Linux, and they were selling well, and yet the very next weekend at the same store there was nothing but XP netbooks for sale.
What did they do, burn all of their Linux stock or something?
And those same agencies are determined to show IE in decline? Who is funding this conspiracy?
Whichever party might see a potential profit in maintaining a myth that Linux is not much in use.
There must be a conspiracy … it isn’t that Linux distros don’t work properly and there is plenty of regression issues on each new release of a popular distro.
I have seen myself on my own machine (which uses a complete intel chipset, every in intel on my machine). Audio not working in Fedora 13 but video working (detected my dual monitors and set correct resolution on install). Fedora 14 vidio not working, but audio is working.
OpenBSD in conparison … everything works perfectly fine since 4.5.
Even Windows XP 64bit Service Pack 2 (which is known for poor compatibility) can correctly detect my video resolution with the default drivers. Why can’t fedora, ubuntu, openSUSE get this right??
Most machines are made for Windows, but not for Linux.
If you want Linux to run perfectly, just buy a mchine that is built to support Linux. Get something that some technical person has certified will run Linux, even better, get a machine that has Linux pre-loaded.
After all, when you want to run Windows, you do so on a machine that comes pre-loaded with Windows. Do the same for Linux and you will find Linux works far better than Windows.
Here are two companies in the US that will pre-load Linux onto machines for you:
System 76
http://www.system76.com/
here are their desktop machines:
http://www.system76.com/index.php?cPath=27
ZaReason
http://zareason.com/shop/home.php
here are their desktop machines:
http://zareason.com/shop/Desktops/
These machines are pre-loaded with Ubuntu Linux, just as Windows machines are pre-loaded with Windows. This allows you to compare apples with apples. These machines will all run Linux perfectly, without a hitch.
Enjoy.
Edited 2011-01-07 12:43 UTC
It was working … it is not now … they have regression issues.
The chipset of my motherboard is fully intel, graphics, video, chipset etc. etc.
Intel supports Unix/Linux for my chipset by writing drivers with either a MIT or GPL license for Xorg … and distro maintainers still break it, between versions.
I have total lockup on Video … but Another Open Source Operating system (OpenBSD) is absolutely fine.
OpenBSD have a relatively small development team compared to other open source projects. (I contribute to the OpenBSD project by buying CDs).
It is not the distro maintainers who break Intel drivers for Linux, it is Intel.
Intel themselves write Intel drivers for Linux, not distro maintainers or anyone else.
Some Intel chips and systems do not support Linux:
http://arstechnica.com/open-source/news/2009/12/poulsbo-mess-casts-…
http://www.happyassassin.net/2009/01/30/intel-gma-500-poulsbo-graph…
The point remains … if you get some party who knows what they are doing to provide you with a Linux installation (just as you would get someone to provide you with a Windows installation), you will have absolutely no trouble with it whatsoever.
I happen to build my own systems, and I research all components before purchase to make sure they have open source Linux drivers. Once I have researched the components and put together such systems, I have absolutely zero trouble with them. Truly. Not once in literally hundreds of Linux installations on my current machine has there ever been a problem.
Edited 2011-01-07 14:08 UTC
I also built the machine myself, and bought Intel because they had opensourced the code for the drivers for every component on the chipset.
So please explain why OpenBSD manages to work fine (with a smaller install base and less developers) and Linux doesn’t??
Edited 2011-01-07 14:33 UTC
Not to get inbetween the two of you but “Linux” or “Ubuntu Linux”? Perhaps there are distributions with better hardware support than Canonical’s products provide? I’ve heard good things about Mint including the polish and hardware support that Ubuntu does not.
I tend towards Debian myself where stability and regression seem to be managed far better. Not that I offer this in comparison to the BSDs.
the problem is I shouldn’t have to “suck it and see” everytime something gets updated. The point I was pressing was that other OSes just don’t have this problem.
However lemur2 likes to ignore the core issues. such as telling me I should change my hardware to be “Linux Compatible” … nobody knows what that will be in several months time.
Linux Distros are the only operating systems that do this. It is because they don’t have good dev model IMO. The Kernel ABI is unstable bu design, which completely defeats the point of having a Interface and what they should be used for. Every programming book and developer I know, thinks that an interface should stay the same and the code that sits behind it should be changed.
No, he told about buying hardware from manufacturers that… say that the hardware works with Linux. This solves some problems :-).
No. If you buy a Macintosh without knowing if Windows will work there… you may face big problems working with Windows.
There are other examples, of course.
There are some things that are not easy to be talked about. I’ll try to put the results of past conversations:
A binary-only driver is very bad news, and should be shunned. It doesn’t respect users’ freedom, users are not free to run the program as they wish, study the source code and change it so that the program do what they wish, and to redistribute copies with or without changes. Without these freedoms, the users can not control the software or control their computing. As Stallman says: without these freedoms, the software controls the users.
Also, as Rick Moen said: binary-only drivers are typically buggy for lack of peer review, poorly maintained, not portable to newer or different CPU architectures, prone to breakage with routine kernel or other system upgrades, etc.
In the article of http://www.kroah.com/log/linux/stable%5Fapi%5Fnonsense.html it’s explained that:
Linux does not have a binary kernel interface, nor does it have a fixed kernel interface. Please realize that the in kernel interfaces are not the kernel to userspace interfaces. The kernel to userspace interface is the one that application programs use, the syscall interface. That interface is _very_ stable over time, and will not break.
The author of the article says that has old programs that were built on a pre 0.9something kernel that still works just fine on the latest 2.6 kernel release. This interface is the one that users and application programmers can count on being stable.
That article reflects the view of a large portion of Linux kernel developers: the freedom to change in-kernel implementation details and APIs at any time allows them to develop much faster and better.
Without the promise of keeping in-kernel interfaces identical from release to release, there is no way for a binary kernel module like VMWare’s to work reliably on multiple kernels.
As an example, if some structures change on a new kernel release (for better performance or more features or whatever other reason), a binary VMWare module may cause catastrophic damage using the old structure layout. Compiling the module again from source will capture the new structure layout, and thus stand a better chance of working — though still not 100%, in case fields have been removed or renamed or given different purposes.
If a function changes its argument list, or is renamed or otherwise made no longer available, not even recompiling from the same source code will work. The module will have to adapt to the new kernel. Since everybody (should) have source and (can find somebody who) is able to modify it to fit. “Push work to the end-nodes” is a common idea in both networking and free software: since the resources [at the fringes]/[of the developers outside the Linux kernel] are larger than the limited resources [of the backbone]/[of the Linux developers], the trade-off to make the former do more of the work is accepted.
On the other hand, Microsoft has made the decision that they must preserve binary driver compatibility as much as possible — they have no choice, as they are playing in a proprietary world. In a way, this makes it much easier for outside developers who no longer face a moving target, and for end-users who never have to change anything. On the downside, this forces Microsoft to maintain backwards-compatibility, which is (at best) time-consuming for Microsoft’s developers and (at worst) is inefficient, causes bugs, and prevents forward progress.
ABI compatibility is a mixed bag. On one hand, it allows you to distribute binary modules and drivers which will work with newer versions of the kernel (with the already told long-term problems of proprietary software). On the other hand, it forces kernel programmers to add a lot of glue code to retain backwards compatibility. Because Linux is open-source, and because kernel developers even whether they’re even allowed, the ability to distribute binary modules isn’t considered that important. On the upside, Linux kernel developers don’t have to worry about ABI compatibility when altering datastructures to improve the kernel. In the long run, this results in cleaner kernel code.
I already told you … Intel write the Linux drivers for Intel chips. Intel own that code, and even though it is open source, Intel retain control of it (because it is their code). Intel say what does and does not go into their code for Linux drivers for Intel chipsets.
OpenBSD developers are the ones who write the OpenBSD drivers. It is clearly in the interests of OpenBSD developers to make sure their own code works, and if it doesn’t work any longer because of a change elsewhere within OpenBSD they can change it to correct such a regression.
Have you ever heard of the term “Wintel”? It is a big thing in desktops. If Linux made significant inroads into desktops, it wouldn’t be quite so imperative to have Intel chipsets in desktop machines, would it?
You do the math.
PS: AMD don’t write the open source drivers for ATI/AMD chipsets. AMD simply publish the programming specifications for their chipsets, and they leave it up to the open source community to write their own drivers.
Edited 2011-01-08 11:46 UTC
You mean the reality that Linux is not much in use, right?
But why believe reality when you can make up unfalsifiable conspiracy theories to “prove” your point?
If people read all the comments before posting, we avoid to see the same comments being posted again and again
http://www.osnews.com/thread?456282
http://www.osnews.com/thread?456294
Excuse me?
The reality is that Linux is the most used OS of any.
Not on the desktop. That is what we are talking about, right?
And I’m not even sure that you are right about it being the most used OS overall. It may have a plurality, but not the majority.
We keep hearing how GNU/Linux is a niche system, but the analyst that takes as parameter the desktop installations to measure its success, or he is completely blind, or is a clumsy manipulator: desktops and laptops are a tiny subset of all the devices that can keep a running operating system inside. Not counting appliances, routers, switches, servers, remote controls, clusters, GPS, supercomputers, PDAs, and finally, the largest group of all, the mobile phones, is ignoring more than ninety percent of the market.
— Paul C. Brown.
I thought we were too, especially since the story is discussing Desktop Browsers.
Mobile OSes and browsers are calculated separately at the stat websites. Which only makes sense since they are two completely different markets that don’t compete against each other.
Mobile Browsers:
http://gs.statcounter.com/#mobile_browser-ww-monthly-200912-201012
Mobile OSes:
http://gs.statcounter.com/#mobile_os-ww-monthly-200912-201012
Anyone else find it interesting that none of the Linux-based mobile OSes use any variation of “Linux” in their names? It’s difficult to even find the word “Linux” at any of the product websites. Here’s the home page for the Motorola Droid, for example:
http://www.motorola.com/Consumers/US-EN/Consumer-Product-and-Servic…
Does “Linux” have such a negative connotation that the word is virtually shunned, even by the companies that earn their money from Linux-based products?
There are some things that can’t be denied. For example, the effects of people choosing.
There’s more in http://articles.cnn.com/2010-03-01/tech/europe.microsoft.web.browse…
Wikipedia in October said 2%:
http://stats.wikimedia.org/archive/squid_reports/2010-10/SquidRepor…
Be Inc also made a big mistake early on to waste money on the PowerPC and fighting with Apple over trying to get access to the hardware. The time and money would have been better spent working on the x86 version, improving hardware support and maybe developing a middleware line to fill in the gap where there was a lack of mainstream software for end users. To place all the blame on Microsoft simply ignores the missteps that Be Inc made which undermined their long term profitability in much the same way that Corel lost focus when it started going in all directions: Netwinder, Linux OS etc.
BeOS lacked drivers !! and applications – an OS in itself is not good for anything at all. Dude, I must know, I bought all my hardware for BeOS back then, and stayed with the crappy Matrox for way too long!
Everybody was looking at BeOS back in the day, and they even gave it away for free the world around on CDs bundled with PC mags.
BeOS didn’t take off, because it could not be used by too many people quite literally, be it for hardware reasons or usage scenarios. The teapot was cool, but not good for anything.
It didn’t even have a proper browser for the longest time, even by the browser standards of the day. Even Scott Hacker (author of the BeOS bible) awoke to that fact only late – he once uttered amazement over at Benews about what the web had on offer after using an Apple for browsing for a change, insted of the trusted Net+…
Be Inc. just didn’t put enough developers behind BEOS when they most needed them, and failed to gain 3rd party developer support.
I think that you will find that Linux (i.e. the kernel) has already overtaken windows.
Interesting how it appears that early European Chrome users were switching from Firefox, but as Chrome began a serious marketing campaign many Internet Explorer users made the switch.
Firefox eventually overtook IE, but not because more people were using Firefox; Firefox usage remained flat, IE dropped, and Chrome increased.
http://gs.statcounter.com/#browser-ww-monthly-200912-201012
With Firefox remaining steady, IE dropping, and Chrome increasing how long until an even 30-30-30 split occurs?
I agree with both of you.
As for operating systems, Minecraft is a great example on how Java works across operating systems.
There are lots of programs that say Java is crap but they are wrong. It has no more problems than any other language.
Java the language is OK but when it comes to making native looking applications it still sucks. Games like minecraft don’t require a native interface.
While I agree that things need to be portable, it’d be an uphill battle to get Java popular on Linux.
There’s a general sentiment among Linux users that Java GUI apps, for whatever reason (I’ve heard Swing’s underlying implementation is partly to blame) are pain enough that it’s better to write a competitor than to use them.
I don’t know whether Java developers just generally suck or whether something about the Java+Swing combo sucks, but I know Java apps are the heaviest, least responsive things on my system short of Firefox and Chrome. (I include Chrome because it loves to gobble up 50% of my 4GiB of RAM under my normal usage patterns)
I also know Sun had a history of under-servicing Linux which Oracle has to live down. (eg. When informed that Java 5 was crashing because it static-linked against a non-multithreaded Xinerama (multi-monitor API) and displaying empty gray windows on non-reparenting window managers like Awesome, they just let it sit. I’m told Java 7 fixes both problems, but it hasn’t been finalized yet.)
Of course, there’s also the sentiment that, given how little market penetration Java has on Linux outside server farms (The only Java app I really use with any frequency is an XML schema translator by the name of TraNG), it’s better to just use a higher-level language like Python or Ruby or even Perl. (A view bolstered by all the hidden Python installs on Windows… for example, the copy of Python 2.5 in HP’s standard load-out for WinVista and Win7 “value added” utilities)
“I don’t know whether Java developers just generally suck or whether”
Yes, that is exactly it.
Android^aEURTMs Users Eclipse IPhone^aEURTMs for First Time, comScore Says
http://blogs.wsj.com/digits/2011/01/05/androids-users-eclipse-iphon…
“For the first time, more Americans say they are using cellphones powered by Google Inc.^aEURTMs Android system than Apple Inc.^aEURTMs iPhone, according to a new survey from technology-tracking firm comScore Inc.
About 26% of U.S. smartphone users said an Android device was their primary phone in the three months ending in November, compared with 25% using Apple Inc.^aEURTMs iPhones, comScore said. BlackBerry maker Research In Motion Ltd. still held the first-place spot, but its share of users declined to 33.5% from 37.6% in August.”
PS: sorry about the garbled apostrophes in the title, I can’t fix it now.
Edited 2011-01-07 03:55 UTC
Look at the browser situation in Armenia:
http://gs.statcounter.com/#browser-AM-monthly-200807-201101
Hey, don’t worry, IE still gets second (with just over 20% share) in Germany, the country where Firefox is king:
http://gs.statcounter.com/#browser-DE-monthly-200807-201101
How about South Korea? A country where the word, open source, means something so bad that they rely on Microsoft Windows to death.
http://gs.statcounter.com/#browser-KR-monthly-200807-201101
But in North Korea Firefox leads.
http://gs.statcounter.com/#browser-KP-monthly-200807-201101
Seriously, it looks like those statistics is based when about 3 users have their computer turned on.
I was surprised by the stats for the Philippines:
http://gs.statcounter.com/#browser-PH-monthly-200807-201101
The cause of Firefox overtaking IE seems to be due to Chrome’s popularity; I wonder if Chrome/Firefox will eventually cut into one another’s usage share. I guess for most people both browsers offer the same appeal of being a choice other than IE.
It will be interesting to note if the fast/lean style of Chrome will be more appealing than the slightly-slower/more-extensible style of Firefox.
It will be even more interesting to monitor what happens with Crome and Firefox when the slightly-faster/yet-still-more-extensible Firefox 4.0 is released.
Well, I still think Chrome is slightly faster:
http://arewefastyet.com/awfy2.php?machine=9
What is important about Firefox 4 is not just the webpage performance. But the UI slowness and many other similair things people would like Mozilla to fix.
And I have to say, Firefox 4 does solve many of them.
I’m running Firefox 4.0 beta 8 and Chromium 8 on the same system.
Neither one is any faster or slower in UI responsiveness than the other.
Firefox 4.0 beta 8 is in general slightly the faster of the two, although there are some cases where it isn’t.
One thing to try: run the Kraken benchmark in both of the latest dev-releases of Chrome and Firefox (of course at a separate time). Mid-benchmark, start up another tab, browse around in both.
On my system, Firefox: sluggish, Chrome: responsive!
This is also what I mean by ‘slightly-faster’, Chrome tends have better responsiveness.
Another thing to try: run the Kraken benchmark in both of the latest dev-releases of Chrome and Firefox (of course at a separate time). After doing just that one thing check the benchmark results.
On my system, Firefox: fast, Chrome: okay-ish.
Kraken is a heavy test of javascript performance. If you happen to be a browser user who can look at and interact with only one page at a time, because you have only one set of eyes and ears, then Firefox is the one with better responsiveness for your purposes.
As I said, Firefox 4.0 beta 8 is generally the faster browser, although certainly not in all circumstances.
I think the speed is basically the same on Kraken/Sunspider, and Chrome seems to have put some heavy work into the V8bench results. They both have excellent performance.
Regardless, even with one set of eyes and ears, you may have background tabs running some scripts. I would think you still want your foreground tab running in a responsive way?
What is your definition of faster? I believe you stated before the proper way to determine this would be to take the average speedup across the top 3 benchmarks:
http://arewefastyet.com/awfy2.php?machine=9
It looks like Chrome is now faster by your own metric.
Chrome’s new Crankshaft (aka V8) javascript engine (the green tace) is very fast at running Googles V8bench benchmark. On this test, Crankshaft (green) scores 3882, whereas TM+JM (purple) scores 2821. Since the higher score is better for this test, V8 is 138% better than the TM+JM score on this benchmark.
For the Sunspider benchmark, the green and purple traces are neck and neck. Nothing in it. If anything, JM+TM is 1% better (lower) than V8, but this difference is probably within the margin for error for this test.
For the Kraken benchmark, where again lower scores are better, Crankshaft scores 16048 whereas JM+TM scores 6975, so on this test Mozilla’s JM+TM javascript engine is 230% better than Chrome’s score on this benchmark.
Overall, Firefox 4.0 beta 8 is faster on javascript performance than Chrome.
So no, Chrome is not faster on the overall metrics. It wins only on Google’s own v8bench benchmark, and even there it doesn’t win by as great a margin as Mozilla’s engine wins on the Kraken benchmark.
Perhaps you failed to notice the significance of the (lower better) or (higher better) criteria on the test results graphs to which you linked?
PS: BTW, thanks for the link. I didn’t know about awfy2.
Edited 2011-01-09 06:31 UTC
Where are you taking your numbers? Kraken especially, the green trace is 5885, not 16048. This score is actually lowest (best) on this benchmark. Also, Crankshaft is more than twice as fast on v8bench.
The awfy2 numbers are what I’m looking at, you? Note that the awfy numbers haven’t been updated in quite a while, and don’t include Crankshaft, so you have to look at a awfy2.
So yes, by your own metrics, for now Chrome is faster than Firefox.
That is a very good question. Yes, I was looking at awfy2, (this site was new to me), because as you say the original awfy has not been updated since about Nov 4.
Mozilla’s JM+TM has been very much faster than V8 for a longish time on Kraken, while V8 has been faster than JM+TM in a similar fashion (but not by as significant a margin). When I looked at awfy2 about 30 minutes ago, this was still the case. I checked and double-checked the numbers. I made absolutely sure of it.
After reading your response above, I re-visited the site, and it now looks completely different. Utterly topsy-turvy to what it seemed before, and none of the numbers I took before are anything like what is on the site now.
Most perplexing.
I’m going to run kraken on my machine and see what happens.
OK, here are my results (I have only a fairly modest machine):
Firefox 4.0 beta 8
RESULTS (means and 95% confidence intervals)
———————————————–
Total: 24198.7ms +/- 5.9%
———————————————–
Chromium 8.0.552.224
RESULTS (means and 95% confidence intervals)
———————————————–
Total: 33604.5ms +/- 9.8%
———————————————–
OK Firefox’s speed for Kraken was measured at 139% of Chromium’s speed on my modest system.
This quick test confirms what I had thought … on my system, Firefox 4.0 beta 8 is slightly the faster of the two browsers.
afy2 is screwy.
Edited 2011-01-09 09:39 UTC
As far as I know Crankshaft is only available in Chrome 10 releases, I’m not sure if they’ve backported it to the 8-series release. Certainly the awfy2 page is running the nightlies (or close to it) of each engine.
Edit: Just checked on my machine (i7 920), and on Kraken my Chromium 10.0.626.0 gets 7273.1ms +/- 3.1% and Firefox beta 9-pre gets 7465.4ms +/- 0.6%, so here they are virtually identical results.
Edited 2011-01-09 11:20 UTC
That part is planned to be solved for Firefox 4.1, Firefox Mobile has that already solved.
@OSAlert
Does OSAlert report HTTP statistics to StatCounter?
In case not, why not?
Does OSAlert report statistics to other analyse institutes?
Even if OSAlert doesn’t, I’d be interested in seeing the OS and browser share that OSAlert sees. I suspect it’s mostly Windows, sadly.
Probably fewer Windows users than general-interest websites, and undoubtedly more Linux users. Mac, maybe about the same as average?
I’d like to see the “Other” ranking for OSAlert.
And almost certainly a lower % of OSAlert readers are using Internet Explorer as a browser.
Edited 2011-01-07 14:59 UTC
It’s hard to see Firefox having a good long term chance against Chrome.
Webkit is the best funded renderer, with multibillion vested interest from usual suspects.
Mozilla would do well to focus on a new, badly covered market like email (making thunderbird the obvious #1 choice for corporate email).
@Vivainio
“Mozilla would do well to focus on a new, badly covered market like email (making thunderbird the obvious #1 choice for corporate email).”
It took several years for Thunderbird 3 to arrive, but it works really well also for dealing with daily high volume.
However for corporate use everybody expect an email program to come with a calendar system. And here it falls apart. Mozilla has spend 7 years and they are still not ready with a version 1.0 of Sunbird/Lightning. The problem is that there have been put no resource priority on this, but it reflects badly on Thunderbird.
I would have though MS had far more of a hold over Europe. Wonder why there is less fear of MS lately.