IE is certainly making good progress lately. However, when every other browser is releasing updates every 6 weeks, Microsoft could substantially benefit from not only a faster release cycle, but also automatic background updates.
Yes because every company will eventually be a startup with only a handful of employees and little investment in technology so that they can jump on the newest technology without costing the company billions. /snark
No, what's going to happen is SaaS companies are going to stop supporting old browsers and their webapps won't work in the browser the IT department is clinging to, the IT department heads will get fired and replaced by people who do a better job prioritizing. "Let's keep using IE8 because it's the newest browser to correctly run that app we developed in 2002" is going to cost people jobs.
You might be surprised. Supporting extra stuff can have serious IT costs. When the head of IT says it's going require x dollars, y FTEs, or z weeks in schedule slippage to bring on that support, the CEO turns to the VP and says, "Jim, is this really worth it?"
But things can shift slowly. I worked at a big media company when the iPhone was launched. The next day, the CIO emailed the entire company of 20k and said that iPhones would not absolutely not be supported, and to please stop asking no matter what VP/SVP/EVP you were. It was years before iOS was permitted on the network.
This is exactly my point. This type of thinking is simply going to be weeded out. More efficient IT heads will replace the slower and costlier ones. Companies will lose potential employees who don't want to carry around a BlackBerry Curve or use Lotus Notes. The market has a way of correctly slow moving companies.
I think we agree more than disagree, but I still take issue with your statement of "more efficient IT heads" and that they need to be "weeded out." It's not a type of thinking -- there was a clear business reason. Re: my example of the large media company I worked at:
In 2007, it was a very small minority of employees who could afford the $599 device and wanted extra support for it outside of BlackBerry Enterprise Server. By 2010, there were a lot of iPhones and the case could be made that it would meet the needs of a significant portion of the company while pushing device costs to the employee as a BYOD, thus saving money while still being able to enforce some security policies thanks to iOS 4. The head of IT wasn't being a stodgy old fart (ok maybe a little) -- he/she made a rational policy decision that I think was correct at the time. Today I think they're on Google Apps.
The problem I'm seeing now is that dozens of employees are sharing logins for the same SaaS tools. And all the passwords are the same so that it's "easy to remember." A single password compromise could leak a lot of data and cause a lot of damage.
People like you and me know encrypt drives and walk away from a stolen laptop with only the cost of hardware, but IT managers today seem like they have a bigger headache to manage.
My point is that the old IT position of updating slowly is being attacked on multiple fronts. From users who want to use modern technology, from SaaS companies who are increasingly dropping support for old-but-not-that-old browsers, and from browser vendors who are all moving towards rapid release cycles. Even IE.
I'm saying that the IT head, when it eventually gets to IE11 in 6 years and turns off automatic updates -- a feature that is on by default --, is going to find himself rather isolated and in my opinion a new crop of IT is going to form that's more nimble as a result.
Just this week gmail decided that being able to open my calendar in a single click was a horrible idea. It only took me a minute or so to figure out wtf happened. But multiplied across a large number of people of varying computer literacy and near weekly changes you are talking about a lot of lost time. Much better to have an annual or biannual training day where changes across a range of software products can be introduced.
I suppose someone will be along to stamp NOT AGILE on my forehead shortly.
Maybe because Windows XP is 12 years old now, and people who use Windows XP are mostly from third world countries who don't really care wether they have IE6 or IE11 installed.
There are, sadly, a ton of Enterprise clients still using Windows XP as well as a ton of home users in first world countries.
Windows XP still works "good enough" for most computing tasks. Having said that, since I'm used to more modern operating systems I find I'm lost on XP now but for people that never left it, it generally works "good enough".
No, it still sounds equally daft. You are aware that the latest version of Firefox runs on XP, right? So what "services on the web" do you think are affected by running XP?
That doesn't help at all. I think you mean applications on the web that aren't compatible with Internet Explorer 8 which is the latest version of Internet Explorer that Windows XP can install. But that is negated by the fact that the latest versions of Chrome and Firefox work fine on Windows XP so that is not an issue.
For a computer-buyer? It's 6 years old, not 12. If you bought a computer in January 2007, it ran Windows XP. That computer could've been a 2 Ghz dual-core 64-bit machine that should still run fine today, but the operating-system bundled with it is end-of-life.
We all know computer sales are in a slump because quite frankly a computer from 2007 is fine at most of today's computing tasks. So are you really shocked that users of such machines are bothered by the fact that their OS is now unsupported with no free-upgrade path?
people who use Windows XP are mostly from third world countries
We're still being migrated from XP to Win7, if they keep doing the upgrades as people's computers come off lease (so they avoid extra downtime) I think the projected finish date was mid/early next year. At least 3 of my coworkers (~20 person team inside IT) are still on XP, this is a largish US corporation.
Because one of those browsers is a core component of the operating system, used in a ton of places and embeddable by other programs which means a major update can potentially break a lot of things. And the others are just standalone programs that only have to worry about their own performance.
I'm just saying it doesn't have to be that way. They could keep the core os component running the old version and have the newer standalone version of the browser running next to it.
There may be some minor technical hurdles with all the shared libraries. Sure. But there's no reason it couldn't be done.
Microsoft could still build a standalone version of IE if they really wanted. The reason they cut off support is laziness, greed (pushes people to buy new windows licences), and there's likely some legal issues due to the old anti-trusts cases where MS argued that IE couldn't be separated from the OS.
What about the code base we all have to maintain supporting those shit browsers? Plus I doubt it's that fundamental a change to the code base. Just point IE to a few newer DLLs and write a different deployment package. If MS are able to update .NET, DirectX and all their other libraries and maintain backwards compatibility then I'm sure they could with IE as well. There's no real technical reason hampering them, just a lack of incentive and plenty of excuses.
And the ROI point is moot because it's already been pointed out that all of the other big browsers still support XP.
Every 3 years its the same bullshit, MS release a new version of Windows and suddenly all their new software cannot run on their older versions of Windows - as if this new OS magically changes the way how Trident works. As someone who's hacked his way around some of Windows internals, I don't buy MSs marketing, it's just a ploy to sell more copies of their latest OS.
Seems to me that most people still on XP are stuck on XP due to ingrained resistance to updates. Who says they will be allowed to update IE even if they could?
If IE is making progress in usability, I don't really see it. I typically use Chrome for three reasons:
1. It's fast
2. It has better plugin support (i.e. Adblock)
3. It doesn't have random freezes
Just today I thought I'd give IE 11 another try on my Windows 8.1 system. It wasn't long before I started experiencing weird random freezes that I literally never get in Chrome. Anyone reading my comment history will show that I use a lot of Microsoft technologies (Visual Studio, Windows Phone, etc) but I am frequently left wondering what the heck Microsoft is doing with IE because it's not working.
IE's compositing and rendering are far ahead of Chrome, honestly. Direct2D and DirectWrite are really really good at using the GPU to render content, and the D3D-based compositing is really good. Microsoft is very good at graphics (and it helps that they control the whole stack of course).
If there's one thing I was really impressed with when trying out Windows 8 and whatever version of IE it came with, it's the hardware acceleration when scrolling. I've never seen such smooth scrolling before, ever.
The fact here is that IE is only available on Windows, whereas Chrome and Safari support all platforms. The common denominator for them is OpenGL. BTW, is OpenGL inherently inferior to DirectX?
Yea the fact that they are still making update option is going to keep fragmentation an issue for IE, at least, from what I understand, with IE11+ they are making it an opt-out not opt-in
I believe these claims are based on sunspider, which is not really considered a great benchmarking suite. Something like Mozilla's Kraken or Google's Octane may give better real-world results.
FWIW, I only recently understood what made a benchmark suite 'bad'. Basically VMs often speed up by identifying patterns which can be optimized. But there's a balance because this analysis itself has a cost and if the speed up is not much (or does not happen often), then it can slow down other apps. Apparently SunSpider has a number of places where it's easy to add special optimizations, but the result may actually degrade more common code execution.
All browsers optimize for SunSpider, but relying on those numbers alone can be deceiving.
Does it support WebRTC yet? IE always seems to be roughly ~2 years late in supporting some of the most important and future-forward HTML5 features.
Maybe they'll push WebRTC in IE12 (not guaranteed), but then it will be some key encryption standard that everyone else has decided on and they're supporting, but IE12 will be lacking. They need to keep IE more up to date with the implementation of the standards.
This seems to show the same thing (not a perfect test, but gives a rough idea about how far behind the IE team is in supporting new HTML5 standards):
What's worse, these aren't the kind of things I can use feature detection to figure out. They're not missing features, they're features that worked in the last version that are broken in this one (IE10), that will remain broken in this one, and may be fixed in the future.
So thanks for fixing one of those, IE11. Maybe you could get around to event.detail some time.
~~~
The biggest problem I have with IE is disjointed updates. They don't care enough to fix the one thing they need to fix. And don't tell me they're doing automatic updates now, they need to go back and fix all versions. Including on Windows XP.
In 2009 the IE team released IE8 on Windows XP. Some time before before that I set my parents up with Chrome 1.
Do you know what they're using almost 5 years later? Chrome 30. And they have no idea. And we're all the better for it.
Do you know what they would be using if it were not for Chrome?
It's not IE11. Or 10, or 9. Nevermind automatic updating, plain old updating isn't an option for my parents.
IE11 has crazy regressions in keyframe animations and transitions. Stuff that worked fine in IE10 and every other browser on the planet now requires voodoo like using settimeout to delay for a magical number of ms before setting certain style properties.
IE11 has been, in my experience, the largest source of awful, browser-specific hackery since the Galaxy S2 or Blackberry mobile browsers.
You might think that, if you've worked with IE10, it's going to be smooth sailing. It's not. IE11 is a regression in comparison.
The browser quirks aren't things like "Oh, you need a vendor prefix here." They are things like, "no matter what, keyframe animations will never work at all in the presence of the Facebook API, and transition animations will work sometimes (but not always!) because reasons, and sometimes (but not always!) animated elements will reset back to their original position after the animation completes."
Because of these browser-specific quirks, you'll be spending a fair amount of time in the IE11 developer tools, which are very nearly unchanged since interns wrote them in IE7. Worse, they frequently crash and take the browser down with it. Or, they get out of sync with the browser, and "Inspect Element" stops working until you restart the browser. You'll be restarting IE and staring at the "Searching for solutions..." dialog quite a bit.
When the tools DO work, you'll find they've added a bunch of new stuff like the "UI Responsiveness Wizard", but that's a poor consolation prize for losing the great debugging features of other browsers, like "Break on attribute modification" and "stabililty."
And, after spending days poring through JS, trying to find any sort of IE11-specific bug or behavior on these complex sites that work in very nearly every other browser, including IE10 and an array of mobile browsers going back to Android 2, you'll find that adding random timeouts will start fixing some of these weird behaviors.
I've been mired in IE11 support for the past several months. It's been a nightmare.
Oh my lord. Thanks for breaking the fragile new trust I had in Microsoft's improved embracing of the web.. They actually made it so that you now have to test and optimize for every IE version out there.
The IE vision of embracing the web is being incredibly dogmatic about "you shouldn't have to do anything special for IE" while simultaneously introducing a slew of IE-specific awful.
This is what drives me crazy with MS. They really improved IE this time around and actually fixed my biggest beef, which is the lack of decent web developer tools.
So yeah, they fixed that, then once I started using the tools, I suddenly realized they completely stripped out the Document Mode/Browser Mode so now you have to figure out some other way of testing old versions.
Oh, but wait! Browser Mode is now back in, not in an update mind you, but only available in Windows 8.1. Seriously, WTF!?
I think the big thing is they're very similar to Firebug and Chrome's developer tools now. Before it took an act of congress to change CSS, HTML and debug JS. Now, it's very similar to other browser's.
Even looking at the side-by-side shots of DOM Explorer, I have a hard time telling which one is new and which one is old.
The parts that are most important to my work have largely gone unchanged, and the debugging tools I use most from Chrome and Firebug are nowhere to be found.
I'm a Microsoft engineer that worked on animations in IE10. Unfortunately, we're not aware of the IE11 regressions you experienced. Could you please share a test page that demonstrates the issue?
> The biggest problem I have with IE is disjointed updates. They don't care enough to fix the one thing they need to fix. And don't tell me they're doing automatic updates now, they need to go back and fix all versions. Including on Windows XP.
Their claim is that it's impossible to provide modern IE for older versions of Windows, because IE is so deeply tied into the OS it requires features and libraries that are intimately tied to newer versions and don't/can't exist on (say) XP.
(Not saying I endorse that as a reasonable position, just explaining why they do things the way they do.)
Their claim is that it's impossible to provide modern IE for older versions of Windows, because IE is so deeply tied into the OS it requires features and libraries that are intimately tied to newer versions and don't/can't exist on (say) XP.
One of the core reasons is that new IEs uses security support from the OS, like low privileges, that didn't exist in old Windows OSES.
So why not un-tie it from the OS? If Chrome, FF, & Safari can do all that they do without deep os integration, why not IE? someone said "One of the core reasons is that new IEs uses security support from the OS, like low privileges, that didn't exist in old Windows OSES" well I don't believe IE11 is any more secure than chrome or firefox, and they don't need that so why does IE?
I know the business model used to be "release a new version of IE, make compatible for only the new version of windows, force everyone to buy a new windows license." and that worked back when there was no competition. but trying to do that in todays market means most savvy users will just use a browser that updates without needing an OS update (or any user action, for that matter).
Really, why doesn't IE just start doing an iterative auto-release cycle like ff or chrome?
Because the availability of IE is part of the Windows platform -- a Windows program gets to assume that IE is installed and the various APIs provided by IE are available.
If Microsoft wants older Windows software to work, it has to ensure that when an app written against IE4 invokes the WebBrowser control in IE11 it still behaves as expected.
IE doesn't need Windows so much as Windows needs IE.
This is also the case in Windows (now) but only for Metro style applications. Those compiled for Windows 8 use the IE10 control and for 8.1, the IE11 control.
That's not a reason - it's an excuse. Every other major browser works fine across Windows versions and MS claims they can't? It's a marketing decision. Microsoft is an OS company so tying their latest and greatest browser to an OS upgrade is compelling.
In contrast to other operating system vendors, Microsoft both refused to license its operating system without a browser and imposed restrictions — at first contractual and later technical — on OEMs' and end users' ability to remove its browser from its operating system. As its internal contemporaneous documents and licensing practices reveal, Microsoft decided to bind Internet Explorer to Windows in order to prevent Navigator from weakening the applications barrier to entry, rather than for any pro-competitive purpose...
Microsoft's executives believed that the incentives that its contractual restrictions placed on OEMs would not be sufficient in themselves to reverse the direction of Navigator's usage share. Consequently, in late 1995 or early 1996, Microsoft set out to bind Internet Explorer more tightly to Windows 95 as a technical matter. The intent was to make it more difficult for anyone, including systems administrators and users, to remove Internet Explorer from Windows 95 and to simultaneously complicate the experience of using Navigator with Windows 95. As Brad Chase wrote to his superiors near the end of 1995, "We will bind the shell to the Internet Explorer, so that running any other browser is a jolting experience." ...
Starting with Windows 95 OSR 2, Microsoft placed many of the routines that are used by Internet Explorer, including browsing-specific routines, into the same files that support the 32-bit Windows APIs. Microsoft's primary motivation for this action was to ensure that the deletion of any file containing browsing-specific routines would also delete vital operating system routines and thus cripple Windows 95.
So while it was originally a marketing decision -- the two were tied together specifically to put competitors at a disadvantage -- the implementation of that decision was not just marketing. It was implemented technically too, to the point where removing IE from a copy of Windows would break the OS. So regardless of the motivation behind the decision, the fact remains that the two things are tangled up with each other quite deeply.
Microsoft could have gone back at a later point and un-tangled them, but that would have been a huge project, and there was no real motivation for them to do so -- the settlement agreement that ended the antitrust case didn't require them to, and the market wasn't rejecting Windows with IE tied into it. So it stayed tangled up.
The big irony is that by tying the OS and browser together to get an advantage over their competitors, they ended up disadvantaging themselves. Those competitors can now turn out new versions of their browsers at a fast clip, because they're not tied to any release schedule other than their own. IE is tied to Windows' release schedule, however, so it improves at a glacial pace compared to other browsers, making it less competitive over time.
The other major Browsers on Windows do not solve, work-with, nor integrate with the metric-tons worth of enterprise and business issues, needs, and past + current solutions. Nor do they take backward compatibility issues seriously.
This is probably 50% of the game which makes Microsoft money with the Enterprise.
>Every other major Browser on Windows does not solve, work-with, nor integrates with the metric-tons worth of enterprise and business issues, needs, and past + current solutions. Nor do they take backward compatibility issues seriously.
the life would so much happier without all that enterprise software crap that works only on IE. "Works only with IE" is like a hallmark of usability/quality/performance one can expect from the hallmarked application. And is not just correlation, it is usually a strong causation as the real reason behind it is that the application was most probably produced by mediocre at best team using some enterprise framework treated like a black box.
>This is probably 50% of the game which makes Microsoft money with the Enterprise.
absolutely. I'm always fascinated by how user-candy MS ecosystem looks to an Enterprise on the first date (vs. say Linux) and how bad things become almost immediately once they move in together.
I'd prefer to see MSIE market share continue to decline to less than 20%. The would warrant not having to make any specific browser hacks to get it to work.
That's part of what's frustrating about the larger move away from version sniffing (e.g. jQuery no longer supporting it) and IE playing games with its User Agent string. I'm in favor of it in theory, but in practice there are a handful of things where you simply need to know the browser vendor or version in order to provide a usable page.
Take those JavaScript numbers with a grain of salt. Microsoft optimizes for and touts their JavaScript numbers using the outdated SunSpider benchmark only.
Let's not miss the point here. IE11 will be massively easier to support than IE8 or 9. Old versions are disappearing relatively more quickly. There are fewer random impossible-to-debug quirks. Flexbox is becoming a viable option.
Microsoft is moving in line with the rest of the browsing world and we get to complain about things like favicons and event.detail bugs. In a few years, you might not even have to write IE-specific styles at all.
So sure, get upset about idiosyncrasies. But at the end of the day, doing good development is getting much easier.
Just to be clear, the rest of the browser world offers modern browsers across a wide array of operating systems. Microsoft can't even be bothered to offer a modern browser across more than two of their own Operating Systems.
Chrome spans at least 7(And that's a conservative estimate) Operating Systems. Firefox the same. And both update in a sane manner.
It's nice that Microsoft offers a good solution to the 5% of users on the cutting edge, but they are still ludicrously behind the curve.
>Just to be clear, the rest of the browser world offers modern browsers across a wide array of operating systems. Microsoft can't even be bothered to offer a modern browser across more than two of their own Operating Systems.
Keep in mind the rest of the two browsers aren't tightly integrated with any OS, thus can enjoy the flexibility what IE can't.
Why can't IE be released in a form that is not so tied to the OS, and leave the old IE bits where they lie? In XP, for example, they could leave the rotting old IE8 guts in, and release a standalone IE11 that opens instead of IE8.
Of course, I would much rather they just exit the browser market entirely. They keep embarrassing themselves over and over again. While new releases are (finally) somewhat modern, there are still incredibly frustrating new bugs with every release that take years to get fixed, holding the web back again and again.
The Safari picture is a bit more... complicated. The latest version of Safari only supports Mavericks(I believe?), but 93% of all online Macs are eligible for the free upgrade to Mavericks.
I'm afraid it won't. If anything it's just one more browser to have to worry about.
All the installations of IE7 and IE8 in enterprise/healthcare aren't going to budge just because a shiny new IE is out. I don't know what it's going to take for that segment to truly go to zero.
This is not true. I have windows 7, I check updates, nothing.
Also, IE out there is IE 8.
And WebGL can't do what we do w/ DOM, so nothing of use in IE11, other than they still don't update IE8 users, so I still have to support old IE.
Since Microsoft only compares using the outdated Sunspider benchmark, I decided to find out how IE compares using modern, proper benchmarks. Here are the results:
As you can see, IE11 lags pretty far behind both Google Chrome and Mozilla Firefox in modern JavaScript benchmarks. So, it would appear that IE is specifically geared towards the outdated Sunspider benchmark. That optimization coupled with the poor performance on Octane and Kraken is likely why Microsoft only ever mentions Sunspider.
Testing Notes: The above test was performed on a fully-patched (as of 2013-11-07) Windows 7 Ultimate x64 machine with antivirus and all other background apps disabled. Each browser was started fresh for each test and had no extensions/add-ons enabled. Hardware-wise, the machine is an Intel Core2 Quad Q6600 running at 2.4GHz with 8GB of PC2-6400 DDR2 RAM clocked at 400MHz (x2 = 800MHz) and an AMD Radeon HD 6700 running at 850MHz with 1GB of RAM @ 1200MHz running Catalyst 13.9.
You're using an unstable Dev release (aka alpha quality) of Chrome and pre-alpha build of Firefox whereas I'm using stable. That's why you got the same performance as I did on Octane with Firefox on your much-faster PC.
Additionally, I'm running Windows 7 and you're running Windows 8.1 so we're comparing different browsers. IE11 on Windows 7 is not the same browser as IE11 on Windows 8/8.1 (hooray for OS-tying).
IE back in the game... boggles the mind a bit. For many years I haven't allowed IE to run on any machines I've managed. There have been a few WTFs in my time in the industry, this is one of them. Like when Apple started selling risc/unix workstations and then intel/unix workstations, no?
Well, I downloaded IE11 to replace IE10 on my VM, and all the issues I had with my current WIP are still present. I'm using `calc()` heavily for a quite complex responsive layout, and paddings are missing on some places and CSS transitions don't work properly while everything works just fine in Firefox and Chrome. I haven't been able to find what's causing these issues, but given IE's history, I suspect that a fix could simply be `zoom:1` or `min-height:100%` or whataver hack du jour you have to remember this time around.
So not at all impressed by this release so far. Will have to wait another year for IE12 and hope it can really compete with FF and Chrome, that means, opening IE and having no surprises.
IE 11 adds a pain for developers who now need to support IE 7,8,9,10,11 and quirks modes (see http://habrahabr.ru/post/201172/, use google translate if you don't understand Russian). The basic premise is that all those IE versions use different engines, with different bugs and supported features/workarounds.
109 comments
[ 3.1 ms ] story [ 151 ms ] threadBut things can shift slowly. I worked at a big media company when the iPhone was launched. The next day, the CIO emailed the entire company of 20k and said that iPhones would not absolutely not be supported, and to please stop asking no matter what VP/SVP/EVP you were. It was years before iOS was permitted on the network.
In 2007, it was a very small minority of employees who could afford the $599 device and wanted extra support for it outside of BlackBerry Enterprise Server. By 2010, there were a lot of iPhones and the case could be made that it would meet the needs of a significant portion of the company while pushing device costs to the employee as a BYOD, thus saving money while still being able to enforce some security policies thanks to iOS 4. The head of IT wasn't being a stodgy old fart (ok maybe a little) -- he/she made a rational policy decision that I think was correct at the time. Today I think they're on Google Apps.
The problem I'm seeing now is that dozens of employees are sharing logins for the same SaaS tools. And all the passwords are the same so that it's "easy to remember." A single password compromise could leak a lot of data and cause a lot of damage.
People like you and me know encrypt drives and walk away from a stolen laptop with only the cost of hardware, but IT managers today seem like they have a bigger headache to manage.
I'm saying that the IT head, when it eventually gets to IE11 in 6 years and turns off automatic updates -- a feature that is on by default --, is going to find himself rather isolated and in my opinion a new crop of IT is going to form that's more nimble as a result.
Just this week gmail decided that being able to open my calendar in a single click was a horrible idea. It only took me a minute or so to figure out wtf happened. But multiplied across a large number of people of varying computer literacy and near weekly changes you are talking about a lot of lost time. Much better to have an annual or biannual training day where changes across a range of software products can be introduced.
I suppose someone will be along to stamp NOT AGILE on my forehead shortly.
http://www.netmarketshare.com/operating-system-market-share....
~1/3 of windows users are still forced to run IE8 at best. If Chrome, Firefox and Opera can run on XP why can't IE11?
Windows XP still works "good enough" for most computing tasks. Having said that, since I'm used to more modern operating systems I find I'm lost on XP now but for people that never left it, it generally works "good enough".
Well... when those computing tasks increasingly involve using web services, then no, Windows XP doesn't still work "good enough".
We all know computer sales are in a slump because quite frankly a computer from 2007 is fine at most of today's computing tasks. So are you really shocked that users of such machines are bothered by the fact that their OS is now unsupported with no free-upgrade path?
We're still being migrated from XP to Win7, if they keep doing the upgrades as people's computers come off lease (so they avoid extra downtime) I think the projected finish date was mid/early next year. At least 3 of my coworkers (~20 person team inside IT) are still on XP, this is a largish US corporation.
And the ROI point is moot because it's already been pointed out that all of the other big browsers still support XP.
Every 3 years its the same bullshit, MS release a new version of Windows and suddenly all their new software cannot run on their older versions of Windows - as if this new OS magically changes the way how Trident works. As someone who's hacked his way around some of Windows internals, I don't buy MSs marketing, it's just a ploy to sell more copies of their latest OS.
If IE is making progress in usability, I don't really see it. I typically use Chrome for three reasons:
1. It's fast
2. It has better plugin support (i.e. Adblock)
3. It doesn't have random freezes
Just today I thought I'd give IE 11 another try on my Windows 8.1 system. It wasn't long before I started experiencing weird random freezes that I literally never get in Chrome. Anyone reading my comment history will show that I use a lot of Microsoft technologies (Visual Studio, Windows Phone, etc) but I am frequently left wondering what the heck Microsoft is doing with IE because it's not working.
Does anyone know how true the claims of 30% faster JavaScript performance are? That sounds too good to be true.
FWIW, I only recently understood what made a benchmark suite 'bad'. Basically VMs often speed up by identifying patterns which can be optimized. But there's a balance because this analysis itself has a cost and if the speed up is not much (or does not happen often), then it can slow down other apps. Apparently SunSpider has a number of places where it's easy to add special optimizations, but the result may actually degrade more common code execution.
All browsers optimize for SunSpider, but relying on those numbers alone can be deceiving.
Maybe they'll push WebRTC in IE12 (not guaranteed), but then it will be some key encryption standard that everyone else has decided on and they're supporting, but IE12 will be lacking. They need to keep IE more up to date with the implementation of the standards.
This seems to show the same thing (not a perfect test, but gives a rough idea about how far behind the IE team is in supporting new HTML5 standards):
http://html5test.com/results/desktop.html
You can get the prototype WebRTCn as a plugin for IE through Microsoft's Interoperability Bridge here: http://html5labs.interoperabilitybridges.com/prototypes/cu-r...
A lot of browsers certainly chrome implement experimental features with vendor prefixes just to get high up in html5 tables such as those.
No mandatory codec (VP8) No SDP
We'll see where this all goes.
http://html5labs.interoperabilitybridges.com/cu-rtc-web/cu-r...
IE10 is awful and introduced weird regressions into canvas and event.detail.
Canvas clipping regions are completely broken in IE10: http://jsfiddle.net/simonsarris/8nNnb/
Microsoft's reply was essentially "lol, next release" : https://connect.microsoft.com/IE/feedback/details/782736/can...
They're fine leaving it broken for IE10 for all time.
Event.detail never resets when you move the mouse: http://jsfiddle.net/uDhwX/
In fact event.detail doesn't reset even after you've reloaded the page
The IE team closed that one as won't fix back in June: http://connect.microsoft.com/IE/feedback/details/789773/ie-i...
Both of these are just fine in IE9.
What's worse, these aren't the kind of things I can use feature detection to figure out. They're not missing features, they're features that worked in the last version that are broken in this one (IE10), that will remain broken in this one, and may be fixed in the future.
So thanks for fixing one of those, IE11. Maybe you could get around to event.detail some time.
~~~
The biggest problem I have with IE is disjointed updates. They don't care enough to fix the one thing they need to fix. And don't tell me they're doing automatic updates now, they need to go back and fix all versions. Including on Windows XP.
In 2009 the IE team released IE8 on Windows XP. Some time before before that I set my parents up with Chrome 1.
Do you know what they're using almost 5 years later? Chrome 30. And they have no idea. And we're all the better for it.
Do you know what they would be using if it were not for Chrome?
It's not IE11. Or 10, or 9. Nevermind automatic updating, plain old updating isn't an option for my parents.
You might think that, if you've worked with IE10, it's going to be smooth sailing. It's not. IE11 is a regression in comparison.
The browser quirks aren't things like "Oh, you need a vendor prefix here." They are things like, "no matter what, keyframe animations will never work at all in the presence of the Facebook API, and transition animations will work sometimes (but not always!) because reasons, and sometimes (but not always!) animated elements will reset back to their original position after the animation completes."
Because of these browser-specific quirks, you'll be spending a fair amount of time in the IE11 developer tools, which are very nearly unchanged since interns wrote them in IE7. Worse, they frequently crash and take the browser down with it. Or, they get out of sync with the browser, and "Inspect Element" stops working until you restart the browser. You'll be restarting IE and staring at the "Searching for solutions..." dialog quite a bit.
When the tools DO work, you'll find they've added a bunch of new stuff like the "UI Responsiveness Wizard", but that's a poor consolation prize for losing the great debugging features of other browsers, like "Break on attribute modification" and "stabililty."
And, after spending days poring through JS, trying to find any sort of IE11-specific bug or behavior on these complex sites that work in very nearly every other browser, including IE10 and an array of mobile browsers going back to Android 2, you'll find that adding random timeouts will start fixing some of these weird behaviors.
I've been mired in IE11 support for the past several months. It's been a nightmare.
So yeah, they fixed that, then once I started using the tools, I suddenly realized they completely stripped out the Document Mode/Browser Mode so now you have to figure out some other way of testing old versions.
Oh, but wait! Browser Mode is now back in, not in an update mind you, but only available in Windows 8.1. Seriously, WTF!?
You can see the differences in the UI here: http://net.tutsplus.com/articles/the-new-ie11-f12-tools/
The parts that are most important to my work have largely gone unchanged, and the debugging tools I use most from Chrome and Firebug are nowhere to be found.
You can reach me at stevebe@microsoft.com.
Their claim is that it's impossible to provide modern IE for older versions of Windows, because IE is so deeply tied into the OS it requires features and libraries that are intimately tied to newer versions and don't/can't exist on (say) XP.
(Not saying I endorse that as a reasonable position, just explaining why they do things the way they do.)
One of the core reasons is that new IEs uses security support from the OS, like low privileges, that didn't exist in old Windows OSES.
I know the business model used to be "release a new version of IE, make compatible for only the new version of windows, force everyone to buy a new windows license." and that worked back when there was no competition. but trying to do that in todays market means most savvy users will just use a browser that updates without needing an OS update (or any user action, for that matter).
Really, why doesn't IE just start doing an iterative auto-release cycle like ff or chrome?
If Microsoft wants older Windows software to work, it has to ensure that when an app written against IE4 invokes the WebBrowser control in IE11 it still behaves as expected.
IE doesn't need Windows so much as Windows needs IE.
In contrast to other operating system vendors, Microsoft both refused to license its operating system without a browser and imposed restrictions — at first contractual and later technical — on OEMs' and end users' ability to remove its browser from its operating system. As its internal contemporaneous documents and licensing practices reveal, Microsoft decided to bind Internet Explorer to Windows in order to prevent Navigator from weakening the applications barrier to entry, rather than for any pro-competitive purpose...
Microsoft's executives believed that the incentives that its contractual restrictions placed on OEMs would not be sufficient in themselves to reverse the direction of Navigator's usage share. Consequently, in late 1995 or early 1996, Microsoft set out to bind Internet Explorer more tightly to Windows 95 as a technical matter. The intent was to make it more difficult for anyone, including systems administrators and users, to remove Internet Explorer from Windows 95 and to simultaneously complicate the experience of using Navigator with Windows 95. As Brad Chase wrote to his superiors near the end of 1995, "We will bind the shell to the Internet Explorer, so that running any other browser is a jolting experience." ...
Starting with Windows 95 OSR 2, Microsoft placed many of the routines that are used by Internet Explorer, including browsing-specific routines, into the same files that support the 32-bit Windows APIs. Microsoft's primary motivation for this action was to ensure that the deletion of any file containing browsing-specific routines would also delete vital operating system routines and thus cripple Windows 95.
So while it was originally a marketing decision -- the two were tied together specifically to put competitors at a disadvantage -- the implementation of that decision was not just marketing. It was implemented technically too, to the point where removing IE from a copy of Windows would break the OS. So regardless of the motivation behind the decision, the fact remains that the two things are tangled up with each other quite deeply.
Microsoft could have gone back at a later point and un-tangled them, but that would have been a huge project, and there was no real motivation for them to do so -- the settlement agreement that ended the antitrust case didn't require them to, and the market wasn't rejecting Windows with IE tied into it. So it stayed tangled up.
The big irony is that by tying the OS and browser together to get an advantage over their competitors, they ended up disadvantaging themselves. Those competitors can now turn out new versions of their browsers at a fast clip, because they're not tied to any release schedule other than their own. IE is tied to Windows' release schedule, however, so it improves at a glacial pace compared to other browsers, making it less competitive over time.
This is probably 50% of the game which makes Microsoft money with the Enterprise.
You're thinking of the Consumer market.
the life would so much happier without all that enterprise software crap that works only on IE. "Works only with IE" is like a hallmark of usability/quality/performance one can expect from the hallmarked application. And is not just correlation, it is usually a strong causation as the real reason behind it is that the application was most probably produced by mediocre at best team using some enterprise framework treated like a black box.
>This is probably 50% of the game which makes Microsoft money with the Enterprise.
absolutely. I'm always fascinated by how user-candy MS ecosystem looks to an Enterprise on the first date (vs. say Linux) and how bad things become almost immediately once they move in together.
Closest that I have is this but that is because it turned to be a security issue: http://yuhongbao.blogspot.ca/2013/07/how-i-found-cve-2013-13...
http://i.imgur.com/SD7xrif.png
Microsoft is moving in line with the rest of the browsing world and we get to complain about things like favicons and event.detail bugs. In a few years, you might not even have to write IE-specific styles at all.
So sure, get upset about idiosyncrasies. But at the end of the day, doing good development is getting much easier.
Chrome spans at least 7(And that's a conservative estimate) Operating Systems. Firefox the same. And both update in a sane manner.
It's nice that Microsoft offers a good solution to the 5% of users on the cutting edge, but they are still ludicrously behind the curve.
Keep in mind the rest of the two browsers aren't tightly integrated with any OS, thus can enjoy the flexibility what IE can't.
Of course, I would much rather they just exit the browser market entirely. They keep embarrassing themselves over and over again. While new releases are (finally) somewhat modern, there are still incredibly frustrating new bugs with every release that take years to get fixed, holding the web back again and again.
Flexbox is still pretty far from a viable option, unless you don't care about multiline (in which case, go for it).
All the installations of IE7 and IE8 in enterprise/healthcare aren't going to budge just because a shiny new IE is out. I don't know what it's going to take for that segment to truly go to zero.
In my market (healthcare, mostly in the US):
IE8 is the most widely used browser, followed closely behind by Safari 7 on iPhone. After that comes Chrome 30 on Windows.Also, IE out there is IE 8. And WebGL can't do what we do w/ DOM, so nothing of use in IE11, other than they still don't update IE8 users, so I still have to support old IE.
Google Octane: https://developers.google.com/octane/
(Higher is better)
Google Chrome 30.0.1599.101: 11337
Mozilla Firefox 25.0: 10127
Internet Explorer 11: 7631
Mozilla Kraken: http://krakenbenchmark.mozilla.org/ (Lower is better)
Google Chrome 30.0.1599.101: 2479.4ms
Mozilla Firefox 25.0: 2653.1ms
Internet Explorer 11: 4792.3ms
As you can see, IE11 lags pretty far behind both Google Chrome and Mozilla Firefox in modern JavaScript benchmarks. So, it would appear that IE is specifically geared towards the outdated Sunspider benchmark. That optimization coupled with the poor performance on Octane and Kraken is likely why Microsoft only ever mentions Sunspider.
Testing Notes: The above test was performed on a fully-patched (as of 2013-11-07) Windows 7 Ultimate x64 machine with antivirus and all other background apps disabled. Each browser was started fresh for each test and had no extensions/add-ons enabled. Hardware-wise, the machine is an Intel Core2 Quad Q6600 running at 2.4GHz with 8GB of PC2-6400 DDR2 RAM clocked at 400MHz (x2 = 800MHz) and an AMD Radeon HD 6700 running at 850MHz with 1GB of RAM @ 1200MHz running Catalyst 13.9.
Windows 8.1 Pro x64 i7 3770k 16GB RAM
-Octane-
Firefox 28 - 10320
IE 11 - 14167
Chrome 32.0.1689.3 - 21223
-Kraken-
Firefox 28 - 1742.6ms
IE 11 - 2252.0ms
Chrome 32.0.1689.3 - 2241.0ms
Additionally, I'm running Windows 7 and you're running Windows 8.1 so we're comparing different browsers. IE11 on Windows 7 is not the same browser as IE11 on Windows 8/8.1 (hooray for OS-tying).
"Internet Explorer 11 is required in order to run this installation"
I'm not a fan of recursive installs.
So not at all impressed by this release so far. Will have to wait another year for IE12 and hope it can really compete with FF and Chrome, that means, opening IE and having no surprises.