Hang Resistance in IE9


IE9 is the most reliable browser available today in terms of hang resistance. In this post on IE reliability, we show how we engineered IE9’s hang resistance so that an unresponsive tab no longer affects the rest of the browser, similar to how IE is resistant to a crashing tab with Tab Isolation. We highlight new user experiences that help you recover from hangs caused by Web pages, long- running scripts or slow add-ons so that you can browse continuously without interruption. We also perform a competitive analysis of IE9’s hang resistance to other browsers. You can use our Test Drive demo or watch this video to check out IE9’s hang resistance in action.

Understanding IE’s Hang Resistance

Next to crashes, hangs are the most common source of user disruption in applications. When an application hangs (or “freezes”), the entire user interface becomes unresponsive. Browser hangs are equally unpleasant. In earlier IE versions, when a tab hangs, you could not interact with the other open tabs or browser windows. When a site hangs in a tab in IE9, the rest of the browser remains unaffected. Let’s dig into some IE internals to understand how this works.

Like other browsers, IE’s browser frame window and its tabs are separate components. They communicate with each other via a shared message queue. When you perform actions like clicking a link on a tab or dragging the IE window, IE adds these actions as messages into the queue. Each message has a desired recipient. The frame window and tabs process the series of messages in order.

Browser hangs will occur when the message queue becomes stuck. For example, a tab becomes unresponsive due to a plug-in, so the tab cannot process any messages designated for itself in the shared queue. This holds up the other messages in the queue, creating a traffic jam effect. As you fruitlessly click around the browser, your actions generate messages that get queued up but won’t be processed since the hung tab still hasn’t processed the current message!

IE9 doesn’t allow this traffic jam to occur. The frame window periodically checks on the tabs to see if they’re hung. When the frame finds out that a tab is hung, it detaches the messages for that tab into a temporary location. This frees up the queue and the other tabs and the frame window can continue to process their messages while one of the tabs is still hung.

Diagram of message queue for IE frame window and its tabs

When the hung tab is responsive again, it answers the frame window’s earlier check. Once the frame window receives the response from the now-responsive tab, it moves the messages for the previously hung tab back into the queue so that the tab can process them again.

In rare cases where IE9’s hang resistance cannot prevent the rest of the browser from hanging, the Windows Hang Experience kicks in. This is the same experience as IE8 and other browsers.

Take a look at the following diagram to understand how IE9 provides a hang-resistant user experience:

Diagram showing tab content independence

A tab houses its own instances of toolbars, BHOs, and plug-ins, as well as the scripts and HTML code required to display the Web page. If any of these components hang the tab, the message queue jams, making the whole browser briefly unresponsive. It takes the frame window under 3 seconds to detect that the tab is hung and detach its messages from the queue. Once that occurs, then only the contents of that tab (colored orange) are unresponsive. The other components are not affected and continue to function properly.

This experience is an immediate improvement over other browsers and earlier versions of IE. The browser is more reliable and responsive. Sometimes you may not even be aware that one of your tabs is hung.

Since the browser frame remains responsive, IE9 also provides you with the ability to recover from hangs caused by the Web page, plug-ins, long running scripts or add-ons. In the following sections we describe the user experience for these scenarios.

Recovering from Web page and Plug-in Hangs

We designed the hang recovery experience to stay out of your way while you’re browsing. When the Web page or one of its plug-ins causes a hang, the contents of the tab naturally become unresponsive. There are no other immediate indicators that a hang has occurred. You can continue browsing in other tabs. You can close the tab, refresh the tab or even navigate to another Web page if you don’t feel like waiting for the Web page to respond.

When the tab remains unresponsive for 8 seconds after the frame window detects it to be hung, it is likely permanently hung. In that case you may want to recover the contents of the hung tab. IE displays the hang recovery notification for that tab when you interact with the tab.

Screen shot showing "site is not responding" notification with "Recover webpage" button

The notification makes it clear that the Web page in the tab is not responding. Additionally, IE dims the title text for that tab to indicate that it is hung. This helps remind you that the tab is hung if you decide to continue browsing in your other tabs.

Pressing the “Recover Web page” button on the notification initiates tab recovery for the hung tab process and sends an error report for the hang in the process. The other tab processes remain unaffected. As we’ve previously discussed, these error reports are crucial to identifying the top hangs to address and fix.

Screen shot showing "site is not responding/Windows is checking for a solution" alert with "Cancel" button

Once the hang recovery notification appears, the following actions on the hung tab will also send an error report for the hang:

  • Closing the tab
  • Refreshing the tab
  • Navigating to a different Web site on the tab
  • Closing the IE window

If the tab manages to recover from the hang by itself, the notification disappears and the title text returns to normal.

Recovering from Script Hangs

Long-running scripts are another source of hangs in browsers. In earlier IE versions, IE halts script execution and displays the long-running script dialog once the script has executed a given number of instructions without reaching completion. IE needed to use this mechanism to detect the long-running script since there was no generic hang detection available back then.

With IE9 hang resistance, we no longer need the instruction-based detection mechanism and can now deliver a significantly improved user experience. When a long running script hangs a tab, it no longer affects the rest of the browser. Once the frame window identifies that the tab is hung, you can press the Stop button on the address bar (or the Esc key) to stop running the script. The Web page becomes responsive immediately and you won’t lose any of the data you may have entered on the page. However, the Web page may not function properly since IE has halted script execution.

Once the tab remains unresponsive for 8 seconds, clicking on the tab displays the long-running script notification:

Screen shot of "site is not responding due to a long-running script" notification with "Stop script" button

Pressing the “Stop script” button stops the script execution immediately. The rest of the user experience is identical to that of Web page hangs. For example, if you navigate to a different Web page on the currently script-hung tab, IE will stop running the script and proceed with the navigation.

Recovering from Add-on Hangs

The Add-on Performance Advisor integrates with hang resistance to help you recover from hangs caused by browser add-ons in IE9. As we previously discussed, some add-ons can take a long time to initialize when you launch IE or open a new tab. Since the tab cannot continue instantiation until its add-ons have finished loading, a really slow add-on will cause the tab to hang. As with the other types of hangs, the rest of the browser will not be affected by the slow add-on.

When the frame window detects that a tab is hung, it also checks to see if an add-on is being instantiated in the tab. If that’s the case, IE displays a variant of the slow add-ons notification once you click on the unresponsive tab. Clicking “Choose add-ons” on the notification invokes a variant of the Choose Add-ons Dialog which shows you the add-on that’s currently loading.

Screen shot showing "Speed up browsing by disabling the add-ons you don't want" dialog

You can keep waiting for the add-on to load or disable the currently loading add-on so that you continue browsing. You can also disable the other add-ons that are currently enabled. Once you disable your add-on(s) and press “Recover Tab” in the dialog, IE commits the changes to your add-ons and recovers the hung tab. With this new functionality your browser will no longer be permanently affected by add-ons that have significant performance issues.

Hang Resistance in Other Browsers

A modern browser should be resistant to hangs in its tabs, similar to how an operating system is resistant to hangs in its applications. Let’s take a look at how IE’s hang resistance and recovery functionality compares to the latest versions of other popular browsers available today.

Hang type IE 9 Chrome 10 Firefox 4 Opera 11 Safari 5
Web page hangs          
Plug-in hangs          
Script hangs          
Add-on hangs          
Legend:
  Full hang resistance — Hangs are isolated to the tab; rest of browser is not affected
 
  Partial hang resistance — Hangs are isolated to more than one tab; rest of the browser is not affected
 
  No hang resistance — Entire browser is affected by hang

As you can see, only IE9 is fully resistant to the various types of hangs we tested in the above table. Additionally, IE9 gives you the ability to recover from all these types of hangs, such as by stopping a long-running script or by disabling a slow-performing add-on.

Google Chrome offers the next best resistance against hangs. Since Chrome isolates its plug-ins into separate processes, a plug-in hang will affect all the tabs that have that plug-in instantiated on their Web pages. Chrome also offers a hang recovery experience although it is slightly different from IE9. If the hang lasts for one minute, Chrome displays a modal dialog asking you whether to kill the page or keep waiting.

The remaining three browsers do not have built-in hang resistance functionality, although Opera is designed to be resistant to script hangs. It conducts its script execution in a separate thread so that a long-running script will not bring down the browser itself. Instead the script incurs a noticeable impact on the browsing performance across all tabs and windows. Since Opera’s add-ons are script-based, they inherit the same type of resistance against hangs.

Going Forward

IE9 is the only browser today that is fully resistant against tab hangs caused by Web pages, plug-ins, scripts or add-ons. These hangs no longer bring down the rest of the browser. IE9 also provides you with the ability to recover from these types of hangs. Coupled with improvements in IE’s crash reliability and add-on ecosystem reliability, IE9 is the most reliable version of IE we’ve ever shipped.

The Web and the add-on ecosystem will continue to evolve in the months to come. We will continue to find and address the top crashes and hangs that you report to us through Windows Error Reporting. We’ll deliver these fixes to you through Cumulative Updates to ensure that we maintain IE9’s reliability for everyone.

Thanks again for using the IE9 pre-releases, submitting your feedback and sending us your error reports to help us drive reliability throughout the product cycle. We look forward to delivering additional reliability improvements to IE in the future!

—Herman Ng, Program Manager, Internet Explorer Fundamentals

Comments (54)

  1. José says:

    Listen guys, IE 9 is a step forward, but could you please stop with the FUD? Thank you.

  2. CvP says:

    Occasionally I get "…. is not responding" upon scrolling/closing a tab which seemingly doesn't affect the tabs at all (after clicking the button). The page doesn't even reload. What's the deal?

  3. Anonymous says:

    @José

    As someone who has used these other browsers, this is not FUD. Microsoft hasn't just put a lot of effort into the usability of their browser, they have given other vendors usable feedback to improve theirs. Process isolation is something microsoft pioneered, this is something they deserve credit for.

  4. David Bloom says:

    Opera is actually single threaded (this is why performance in other tabs is reduced if a slow script is running in one tab, even on a multi-core system). Within that single thread, most of Opera's components (scripting, layout, parsing, etc) basically use rudimentary cooperative multitasking.

    Reference: dev.opera.com/…/web-workers-rise-up "Opera is built as a single-threaded browser with support for a wide variety of platforms, so our current implementation of Web Workers interleaves code execution in the single UI thread. Other browsers, however, may have multi-threaded architectures which enable simultaneous execution of code."

  5. @José says:

    Please enlighten us with specifics, or are you just spreading anti-MS FUD?

  6. Aleksandras says:

    100% Agree with José.

    IE9 is a step forward indeed. However, considering performance, it is anyway stubborn and massive application.

    And another thing – the marketing of IE is getting better too, because now it's more opened, but disinformational tactics make many experienced users sick. Believe us.

  7. Alexander says:

    Ironically, my Firefox 4 hanged for a couple of seconds when loading this page 🙂

  8. NJ says:

    Why can't people just accept that it's better than FireFox and others…?

    Everytime Microsoft releases something, people start saying how bad it is; without giving them the benefit of doubt!

    I've worked with FF in the past but I switched to IE as it performed faster, less crashes,…

  9. FremyCompany says:

    It was one of my huge request for IE9. I must agree I wasn't aware you did so much progress on this. But it's still not perfect. For example, a feature to improve: in my bug report, the Test 2 (alert) still fails in IE9 : connect.microsoft.com/…/deny-of-service-ie9-autorecovery-engine-is-better-than-ie8-one-but-still-doesnt-cover-major-dos-sources

  10. Eric Caron says:

    Nice progress, but http://crashie8.com/ still makes my other tabs unusuable when I click the "Recover webpage" button.

  11. Lance says:

    @NJ

    It is a problem of credibility when PR/Marketing will creep often into postings, laughable posts/sections trying to mention grand improvements phrased so not to acknowledging previous failings and highly selective testing/scope to white wash being good in some areas as being dominant entirely.

  12. Parrotlover77 says:

    I am a big fan of IE9 and use it as my primary browser now.  That said, this article is a bit misleading.  I have run into situations where IE very quickly let me know a tab crashed, the probable cause of why, and allowed me to quickly attempt to reload the page again.  This is really nice.  But this is not always the case.  Since starting to use IE9 RTM the day after its released, I have had dozens of occurances where a webpage fails to load and the entire Internet Explorer process hangs (UI is unresponsive).  It's possible that if I was patient, IE/Windows would present me with a crash dialog and offer recovery some number of minutes later, but I'm not patient — and I don't think many users are.  The good news is that even if I kill the process, the session recovery works pretty nicely in IE.  And sometimes when the offending IE process is killed, not all my browser windows/tabs actually disappear.  So, progress is being made.  Still, the sandbox is far from complete.  And, begrudingly, I must admit that Chrome's crash-resistence and crash-recovery, at this time, is still superior.

    Still, IE9 is leaps and bounds ahead of previous IEs, better than Firefox and Opera, and is finally, once agian, a very "modern" and enjoyable browser.  I just get a little tired of the propoganda.  I wish this blog was more "techy" and less "marketing."

  13. Kevin says:

    Crashie8.com doesn't even dent IE9 in my tests.

  14. Graeme says:

    Did he say he was using Chrome 6 Beta in that video?? Chrome is up to version 10 in Stable and 12 in Dev!

  15. Anon says:

    @Graeme

    They're probably up to version 15 in planning.

    What's up with the Chrome team and their version numbering. They done gone crazy.

    Glad to see the IE team "grew a pair" and went to "MSIE 10" in the useragent, instead of chickening out like Opera did.

  16. @Graeme says:

    Chrome's versioning is meaningless.Version 6 went stable September 2010.

    That's 4 versions in about six months. At that rate ( I mean …cadence <ahem>) can you tell the difference between the updates? Do you think they made significant improvements during that period, in the area discussed?

  17. hang says:

    Firefox detects scripts eating too many resources and allows the user to stop them. I doubt I have ever seen an add-on hanging Firefox but I guess that in most cases script protection still works for them.

    IE way to detect and solve hangs can be better (process isolation should do some good here) but you should show some real data, as far as I know Firefox correctly detects scripts taking too long and the number of cases where it fails are close to zero.

    Plug-ins run in a different process in Firefox and I've seen Flash crash several times without affecting anything else than flash content.

    I've seen IE 9 (maybe it was the beta) hung doing some tests, the whole UI was unresponsive and the only solution was to kill through task manager. The same can be said of F4 betas which I crashed repeatedly doing some WebGL tests.

    I'm a Firefox user and I haven't seen the final version of F4 hung or crash, the same can be said of my more limited experience with IE9.

    I think the post is quite misleading. The whole point is that IE9 architecture prevents hangs while others don't. Truth is that architecture is not everything about avoiding hangs and that everyone is in a grey area and not the black and white picture presented here.

    Also, I would expect some real use cases and data from feedback to support this claims, without it IE9 and other browsers real hang resistance is completely hypothetical. I would be disappointed if all this effort makes irrelevant differences in a field where a simple session recovery mechanism avoids almost all the damage.

  18. Schalk Neethling says:

    I do not agree at all. The latest example is that IE9 is not launching at all for me on Windows 7 Professional 32bit. So, I thought maybe I will remove it using Programs and Features and then add it back. Still no luck. I have also heard from colleagues that since upgrading the browser it is crashing for no apparent reason, such as searching google from google.com. The developer tools are also very  unreliable when changing browser and document modes. Please, please either fix these issues or stop producing a browser.

    You have tried to do a lot of good since IE7 but IE9 is proving to be the new IE6

  19. Herman [MSFT] says:

    @hang:

    While Firefox has long-running script detection, the whole browser is not responsive until the detection kicks in and halts script execution. IE9's hang resistance allows the rest of the browser to be responsive when a long-running script hangs a tab.

  20. @Herman says:

    …but this page doesn't say that.  It's a real shame that the credibility of an seemingly decent product is undermined by obvious and consistent cherrypicking of results.  From some perspectives IE9 may be the fastest browser, and from some perspective IE9 may have the best hang resistance – but by claiming this is the only perspective these articles come across as untrustworthy.

  21. Nathan says:

    IE9 team, you guys could really do with a dose of humility. Your claims are often misleading and your stabs at other browsers are often off the mark. How about you endear yourselves to the tech community by doing your best work and not rubbing it in the faces of your competitors? You really haven't yet earned back that right yet.

  22. schalk@ossreleasefeed.com says:

    Why on the 19th of April are you testing anything against Firefox 4 beta 3 when the final is out? And what about Opera?

  23. Bitter Anti-IE9 People says:

    Stop crying now.

    As a user who uses Firefox 4 as my main browser, and IE9 as my secondary browser, I can vouch for this article.

    And that http://crashie8.com/ Website didn't crash IE9 either. I can still access the other tabs.

    Furthermore, that website does not execute the same Javascript if you change the Browser Agent String to Firefox or Chrome or Opera — Meaning, the programmer of http://crashie8.com/ is CHEATING. http://crashie8.com/ Will crash Firefox, Chrome, etc… when your Browser Agent String is changed to Internet Explorer.

  24. raw says:

    You should take some precautions against continuous modal dialogs (alert,confirm,prompt) like the other browser vendors have done, either make them un-modal inside the tab or add a checkbox to the bottom to prevent more dialogs from that webpage. Annoying successive dialogs can force a user to close the entire browser.

  25. Kurt Dane says:

    The video doesn't work. Uploading a WebM version would remedy the situation. It's interesting that Microsoft continues to struggle in getting something as simple as HTML5 video working on this blog. Frankly, it's a bit amateur hour.

  26. IE user feedback says:

    IE create a problem when debug web application in it. Firefox and chrome both not close itself as IE so i need to refresh their if i want to test something. but in IE their is hard to do that you need to reproduct step everytime for testing issue again how let's take a look

    suppose i have a issue in page B

    then in Firefox and chrome i just need to refresh to see a look

    in IE he close itself whenever they have much tab who i browse not debug because it's close itself you need to run application again and go A go to B then reproduce the issue by spent some second.

    so their is nothing you can make better for debug the application in IE even it's support javascript debugging.

    conclusion: maybe their is option for close automatically after debug or not

  27. dlh2009 says:

    I think that Microsoft has done a great job at improving IE. I still think that there is some work that needs to be done. I have been using IE as my primary browser since Windows 98. I have beta tested every single version of IE, and I still think that IE9 is a great improvement.

    I also believe that IE9 is much better than the other browsers. I have used Firefox 4, Chrome 12 Dev, and Opera 11.10 and have not found those browsers to to fit my needs. Personally, I think it is just preference because of the fact that I have been using IE for so long and have choosen not to navigate to someone else.

    I do have concerns for the people who are having issues installing IE9 on their computers. I have had a few people who have tried to install it and it either crashes during install or it doesn't seem like it installed correctly.

    I give kudos to the IE team, and I hope that they keep up the great work!

  28. Thomas H says:

    I'm very dissapointed with IE9.

    It always changes over to the Compat- mode – even for all brand new sites. It would be great if IE was adapting to what is coded out in the world rather than trying to make developers adapt to IE.

    Also there are many many problems with IE9 -64 bit and fonts. It might be that the graphics drivers are not compatible with IE9.  This page for example is impossible to load because of the font Helvetica – http://store.apple.com/us

  29. Sven Tycker says:

    On my computer IE 9 crashes as soon as it i s started for all users on the computer except one.

  30. Biggebass says:

    Looks good but….there's a bug somewhere. My IE9 32bit (on Win7) cannot open new tabs. The tab is created but completely empty. If I switch to 64-bit version it's OK. I have to use the 32-bit for my Bank Account ID because it doesn't work with 64-bit version.

  31. @Biggebass says:

    Your symptoms sound like ActiveX control or BHO issues.

  32. Torbjörn Andersson says:

    @Sven Tycker

    I haven't used IE9 much, but I had problems with IE9 crashing on start-up too. Well, not the browser itself but it might as well have because it got stuck in an infinite loop of the first tab crashing and being resurrected. That problem went away when I turned off hardware acceleration, but maybe you've already tried that?

  33. nstlgc says:

    All this talk about "hang resistance", and crashie8.com still freezes the entire browser. At first it only froze its own tab, but when I clicked "recover webpage", IE9 hung completely.

    So no, not as good as you guys think.

  34. RP says:

    I tested and had the same experience as nstlgc.

  35. Vitor Canova says:

    I came back to IE in IE8 after Firefox starts to be very slow in mid 2009. I like very mutch IE9 but I think that is very dificult to create plugins. If IE support Plugins made with Silverlight of WPF and plugable with MEF will be very good.

    You said "Chrome isolates its plug-ins into separate processes". IE9 use one instance for every tab. But can it increase memory usage?

  36. tariq96 says:

    For those who are trying crashie8.com; try it with all the adds-on is disabled

  37. aoobtd says:

    I thought this forum was strictly about IE9       If I wanted to talk about how to  run FF or Crashie I would have gone to there forums………..can we get back to IE9 and leave the other stuff alone?

    I like IE9 but like some others get very frustrated with the HANG time….takes sometimes more than 2 minutes to load on startup and switching sites can take 45+ seconds……Is this common?

  38. hunter222 says:

    I imstalled ie9, and then uninstalled it, I like the favorites bar in ie8, ie9 doest have it.

  39. maybeican8 says:

    I would like to get rid of ie9 and get back ie8 but don't know how to do it.

  40. crashie8.com is CHEATING says:

    crashie8.com employs a JavaScript Crasher function that can crash Firefox, Chrome, etc… It won't crash IE9 though. JavaScript Crasher function works by loading a lot of text in the tab non-stop. Firefox 3 crashes. Chrome hangs. But IE9 is fine — you can even close that website anytime, no problems.

    The thing is, crashie8.com is hypocritical. It removes the supposed "crasher" code when Non-IE browsers are detected. It does Not prove anything with objective non-biased tests.

    Thus, crashie8.com is CHEATING. Everyone know it already. Hacker News discussed that, and eventually discovered the CHEAT.

    TO THE IE TEAM: Thanks for writing this article! You know, sometimes you have to defend yourselves and brag a bit. Don't let the Anti-MS bigots win with false and biased claims.

  41. To Crying Anti-IE People says:

    Move on with your lives. It's just a browser.

    Get off the basement, enjoy some sunshine.

    Life is about having fun with your friends and family.

    Don't let a stabilized browser like IE9 make you sad.

    There are many other things on which you can prove yourselves right, ok?

  42. Unknow says:

    What is problem issue with IE9:  A few users disabled windows update, Active center, add-ons and plugins before download that cause not work with IE9!! supposed be enable all windows update, Active center, add-ons and plugins before download!! after enable all and then IE9 will run very smooth and  no issues. BUT sometime not work on version and operator system with processor access.

  43. Eric Caron says:

    There's no funny-business on crashie8.com. Just a boring, static HTML file that's served identically to all browsers. I really don't care enough about it to engage in any shenanigans. If anybody wants to talk about it, my email is eric.caron@gmail.com.

  44. Please Don't says:

    When anything in Windows crashes we get the:

    "Windows is checking for a solution to the problem" dialog

    Is there a way to turn this off completely?  Never once in my entire I.T. career managing 1,000's of PCs has this ever, once come back with ANYTHING useful of ANY KIND!

    All it does is P!$$ me off that it is taking up my precious time and patience by stopping me from doing actual work and re-starting the application.  It is a complete 100% total EPIC FAIL in terms of usability.

    If you want to pop up a "toast" message that you are reporting the crash (if I have set permissions to do so as such) that's fine, but for the love of god don't you dare waste my time with a dialog of uselessness… and don't insult me with the hint of luck that it might actually find a solution to the problem that we both know full well it won't.

  45. ieblog says:

    @Please Don't: You asked, "Is there a way to turn this off completely?" Yes, there is.

    Using the search box in the start menu search for "problem reports". One of the search results will be "Choose how to report problems". Select that item which brings you to Control PanelAll Control Panel ItemsAction CenterProblem Reporting Settings. From there, you can choose to "Never check for solutions." Note that this is Windows-wide setting.

  46. Zkal says:

    @Eric Caron: Sure, the HTML is identical. The Javascript however isn't which is proven by the fact that when you make other browsers to pretend to be Internet Explorer, they'll crash too.

  47. Eric Caron says:

    Here's a javascript-free version: crashie8.com/index-nojs.html

  48. IE9 might have better hang resistence handling. However, I notice something wrong with frame message queue behavior as it might still cause inresponsive UI experience in IE.

    IE should have better UI experience without unresponsiveness. Take a look at my IE9 RTM performance issue in multiple tabs report here: connect.microsoft.com/…/ie9-rtm-all-in-one-performance-issues

  49. Pfurz says:

    This article makes it sound like Microsoft solved the halting problem…

  50. Please Don't says:

    @ieblog – I followed the steps suggested – There is no option for "Choose how to report problems". Only a list of the apps that have crashed before.

    As for a Windows wide setting, – YES! this is exactly what EVERYONE wants.

    Actually I found the option, you need to search for "Choose" in the top-right search, then the option becomes available.

    Once there you have to choose the option for don't check (the one that isn't recommended)!

    Oh boy Microsoft, you have so much to learn.

  51. rotterdarned says:

    This is a Microsoft web page, so it's styled in a Microsoft fashion. Why some folks seem so appalled by this style, under the circumstances, is most peculiar to me – what do these folks expect? A Microsoft whose style instills confidence by any other primary method than FUD? An IE that comes to the world? No, such change would be asking too much in a single browser version's upwards move – IE9 remains a browser that expects the world to come to and revolve around it.

    But maybe with good reason … . Along with other major browsers that are built as core to their operating systems, Internet Explorer rounds out the Big Five in market share. It remains number one in share and will do so, as long as Windows remains the dominant system.            

    Most commenters seem generally pleased with IE9's advances over previous versions of the browser, although there is healthy debate regarding IE9's number one standing according to the parameters described in Herman Ng's article on "Hang Resistance." One gets the impression that Microsoft and the third party software developers that produce to suit the Windows system hold Internet Explorer to a higher standard than others hold the other big four browsers. Because IE powers the web activity of the world's dominant operating system, a higher standard actually would make sense. Where the disagreements arise is on what features comprise a standard that is "highest": speed, stability, resistance to hangs / crashes, etc?

    Anyone familiar even with the stable versions of Chrome knows and doesn't seem particularly concerned that they'll experience a higher frequency of browser crashes than they'd experienced with the other four historically. Then again, to their credit, Google's Chrome calls a crash a crash, whereas Internet Explorer still refuses such blunt verbiage and prefers to seek answers in a Windows system that take forever to find – and rarely finds such answers. The delicate term "Hang Resisitance" has been broken out into four causes, including the trickiest "Plug-In Hangs."

    The problem with such terminology is that one developer's plug-in is another developer's add-on and / or script and / or (out-of-date) web page. But, when IE finds a page to be out-of-date in this context, Microsoft is really saying that the page's developer has not come to IE, has decided that the world shouldn't revolve around IE and has elected a different standard of browser than IE to be "higher."

    The case can be made that Google is behaving most aggressively in optimizing their browser's speed because they can afford to do so – Chrome remains independent of any particular operating system and, while it's working toward powering what will be Google's eventual OS for the Workstation /  Desktop / Notebook powered by a multicore CPU and GPU arrangement (currently, at least) , Chrome's not there yet – it's still not core to any operating system. Google realizes this better than anybody, so they're making hay while while the sun shines. We can be fairly certain that, when Google finds Chrome to be truly ready for primetime, their proprietary, high-powered OS will follow shortly thereafer. It'll be fascinating to see if Google maintains such an aggressive pace for the development of their browser at that time.

  52. Brian Forgrave says:

    Playing Yahoo Cards with Internet 9, when you have the window: ( page of score or last tricks ) If you try to close the temp window from top of screen ) It will freeze the page for a few seconds. But if you close the java scrips from taskbar, then they will close correctly. If a fix for this problem is to be, it will be from routing the web-pages threw the taskbar. Good luck with fixing this problem Microsoft. Windows 7 is very cool, it rocks. I ever build my next gaming pc, it will be running Ultimate 64-bit or better if it comes along.

  53. Eric Caron is a Scammer says:

    I saved crashie8.com/index-nojs.html to my local disk, because I don't trust your server-side manipulations.

    Then I opened index-nojs.html in Internet Explorer 9. Guess what? IE9 didn't crash at all. I can even Right-Click,View Source that page.

    Conclusion: This proves all the more that crashie8.com server switches code depending on what browser it detects.

    @IE Team: Please provide a way to pause script execution, so that it's easier to catch misleading/deceptive websites like the one above.

  54. Zkal says:

    Having tested crashie8.com (nonjs version saved on local computer) on multiple browsers, it indeed does crash IE9 only due to a missing </form>. Could claim that the site coding should be fixed but let's be honest, a browser shouldn't crash if a site forgets a tag. So can safely admit I was wrong and the site in question isn't doing any browser based detection.