HTML5, Modernized: Fourth IE9 Platform Preview Available for Developers


IE9 started from the premise that the modern web will deliver HTML5 experiences that feel more like native applications than sites. Building on hardware-accelerated SVG, canvas, video, audio, and text, developers will use the power of the whole PC to achieve great performance. On the modern web, developers will use the same markup across different HTML5 browsers.

With IE9, we have worked much more closely with the developer community. Developers have had an earlier (and more frequently updated) look at the platform. With that early engagement, developer feedback has had a bigger impact than before. People have downloaded IE9 Platform Previews over 2.5 million times. The samples on the IE Test Drive site have had over 20 million visitors. We appreciate the positive feedback and all the specific issues logged in Connect. They’ve helped us make demonstrable improvements that the community has noticed.

The fourth Platform Preview of Internet Explorer 9, available now, shows the opportunity of fully hardware-accelerated HTML5. You can run new test drive samples that show modern SVG and native JavaScript integration in action. In March, we promised to deliver platform preview releases approximately every eight weeks. With this installment, you will find more performance and more support for same markup. You’ll also find many fixes to issues reported in previous Platform Previews. Here is a video of some of the test drive samples:

Note this video uses the HTML5 video tag (with the H.264 codec) if your browser supports it, and falls back to other methods otherwise. It’s a good example of same markup in action.

Fully Hardware-Accelerated HTML5

The performance benefits of hardware acceleration are clear from running different sample sites side by side in IE9 and other browsers. Browsers that implement partial hardware acceleration – for example, text-only, or video-playback only, or image-only acceleration – offer inconsistent and possibly unpredictable platform experiences to developers and end-users.

IE9 offers consistent, fully hardware-accelerated text, graphics, and media, both audio and video. Try Hamster Dance Revolution, IE Beatz, or MSNBC Video in different browsers to experience the difference. Psychedelic Browsing demonstrates what HTML5 canvas can do when it’s fully accelerated with the GPU.

IE9 PPB4 and Firefox 4  showing the Psychedelic Browsing demo.  IE9 gets 1789 revolutions per minute, FF4 gets 829 rpm

IE9 PPB4 and Chrome 6 showing the IE Beatz demo, IE9 gets 36 frames per second and 160 beats per minute, Chrome 6 gets 7 FPS and 115 BPM

IE9 PPB4 and Chrome 6 showing Hamster Dance Demo, IE9 gets 41 fps Chrome 6 gets 4 fps

Modern SVG

With Platform Preview 4, we’re excited to show highly-interactive and integrated, or modern, SVG. Typically, developers think of SVG as the graphics format for static engineering diagrams and images. With HTML5 and hardware acceleration, SVG is an excellent choice for a new class of interactive, animated scenarios.

You can see great SVG performance, animated via JavaScript, with the SVG Dice example. The sample shows striking performance differences between browsers animating the same SVG markup, as well as the benefits of being able to style SVG with CSS. Unfortunately, you can see differences in how different browsers display the same SVG; as an industry we have more work to do so the same markup delivers the same results.

Native JavaScript Integration

We’re committed to the right foundation for HTML5 applications, including performance and ensuring the same markup and same script work across browsers. One aspect of doing these things well is integrating the JavaScript engine natively inside the browser, rather than bolting it onto the side to support multiple JavaScript engines as some other browsers do today. How a JavaScript engine is integrated into the browser is as important as the engine itself for real-world HTML5.

The fourth Platform Preview moves the new JavaScript engine, codenamed Chakra, inside IE9 and brings them together into one single, integrated system.

Through this deep integration, the performance of real world websites significantly improves, and IE9 becomes the first browser to have a shared DOM between the browser and the script engine based on ECMAScript5. The benefits start with real-world performance and consistency.

The easiest way to understand the importance of this fundamental change is by looking at how earlier versions of Internet Explorer integrated JavaScript. For the last 15 years Internet Explorer has supported multiple programming languages including JScript, VBScript, and even specialized languages such as Perl. While this gave developers choice it also came at the cost of performance and features. The browser and these script engines communicated through COM which could cause performance problems. Each script engine had its own language specific view of the DOM which created discrepancies. Additionally, the browser was forced to use a least common denominator approach which made adding new features challenging.

Diagram of IE8 vs IE9 JavaScript Integration Model.  In IE8 the JavaScript engine was outside of IE, in IE9 the JavaScript engine is inside IE and shares the IE DOM

In the fourth Platform Preview, we’ve moved the JavaScript engine inside IE9. With this change, communication between the browser and script engine is now direct, which significantly improves performance for real world websites. We now have a single DOM, shared across all browser subsystems including JavaScript. This change ensures a consistent and interoperable view of the document. And this single DOM is now based on ES5, which prepares the entire system for the future.

When programming the IE9 DOM from JavaScript, objects now feel like native ES5 objects because, underneath the covers, they actually are ES5 objects. This approach brings the benefits of ECMAScript5 to the DOM. With the fourth Platform Preview, IE9 becomes the first browser to have a fully discoverable DOM through ES5 reflection features. IE9 is the first browser to apply ES5 bindings to DOM objects, enabling a full Inheritance view of the DOM, and taking advantage of the WebIDL specification as the foundation for this support. Together, these changes provide developers a natural ES5 based programming model. Try some of these enhanced DOM capabilities out for yourself to see how well your browser’s DOM and JavaScript engine are integrated. IE9 will continue to support additional programming languages through the legacy model, but we strongly encourage developers and enterprises to take full advantage of the benefits of JavaScript moving forward.

Platform Preview 4 also continues improvements to the JavaScript engine itself. One measure of JavaScript performance is the Webkit Sunspider microbenchmark. Here is a chart of the latest results:

WebKit Sunspider Results from 8/2/10, IE9 PPB4 get's 4th

The differences between browsers on this microbenchmark are converging within thousandths of seconds on tests that repeat operations many, many times to find any differences at all.

Real-world HTML5 performance often reflects the entire browser’s performance, not just the JavaScript engine in isolation. This video, for example, shows a side by side comparison of different browsers running an HTML5 canvas. The performance differences between browsers are striking and do not reflect relative JavaScript performance. We encourage other browser vendors to consider following our lead in designing for end-to-end performance and natively integrating JavaScript engines rather than treating script as a separable subsystem to be optimized in isolation.

Same Markup, and Tests

To assess the quality and completeness of a browser’s standards support, we look to the official standards bodies. Their open, consensus-based process is the best way to bring the community of browser vendors and web developer and design professionals together in building a test suite.

With Platform Preview 4, we’re contributing 519 new tests to the standards bodies. Based on community feedback, we’ve also updated five of the previously submitted tests. This brings the total number of tests we’ve contributed during IE9 development to 2,138. We welcome your feedback on the specific test cases. Please continue to provide feedback on the test cases to the appropriate W3C working group. In case of ES5 test cases please provide test case specific feedback via Microsoft Connect. We also invite you to submit your own test cases to the standards bodies as well. You can find the test cases Microsoft has developed at the IE Test Center.

These test cases represent a strong start on a complete and comprehensive test suite of the web standards developers expect to work consistently across browsers. While the suite is not complete yet, it is interesting to note how interoperable some of the same markup is across different browsers:

Cross Browser Test Results Summary from IETest

Some people use a particular test or website as shorthand for standards compliance. Different sites test different subsets of different standards to different depths. Acid3 is one that some people in the community have cited. It tests about 100 fragments of a dozen different technologies. Here’s a screenshot of how today’s IE9 Platform Preview runs today’s Acid3 test, going from 83 in the previous platform preview to 95:

IE9 PPB4 showing Acid3 score of 95/100

As IE9 has implemented more of the standards that developers use and value, IE9’s Acid3 score has continued to rise. The remaining points involve two particular technologies (SVG Fonts and SMIL animation of SVG) that are in transition.

Support for SVG Fonts in the web development and font communities has been declining for some time. There’s already been discussion without objection of dropping SVG fonts from the Acid3 test. The community has put forth a proposal in the SVG Working Group to give SVG Fonts optional status.

Instead, developers can use the Web Open Font Format (WOFF, supported in IE9 Platform Preview 3 as well as other browsers) for both HTML and SVG content. It works well in conjunction with the CSS3 Fonts module and has broad support from leading font vendors (e.g. here, “a majority of font makers have already settled on WOFF or services like Typekit as their format of choice”). WOFF fonts are a better long-term solution for many reasons discussed previously.

Similarly, support for SMIL animation of SVG in the web development community is far from strong. The leader of the SVG standardization effort wrote that not supporting SMIL in its current state is probably best “since the SVG WG intends to coordinate with the CSS WG to make some changes to animation and to extend filters.” There’s already work started to reconcile CSS3 animations and SVG. Developers interested in animating SVG can use JavaScript, as the samples in the test drive site do today, with consistent results.

Getting sites ready for Beta

With the fourth Platform Preview, we strongly recommend developers, designers, and partners to start getting your sites ready for the IE9 Beta.

  • Test your site in IE9 Standards Mode.  This mode provides the best performance and interoperability and will offer additional benefits in the IE9 Beta. We suggest using the HTML5 doctype. More details here and here.
  • We recommend sending IE9 the same standards-based markup your site sends other browsers. More details here and here. From the feedback so far, and our experience with sites, the best way to get your site working in IE9 Standards Mode is to start from the same markup other browsers receive rather than IE6, IE7, or IE8 markup.
  • Use feature detection, not browser detection to handle any cross browser differences in behavior or feature support.  This keeps your site working even as browsers change.
  • Please continue to report issues on Connect if your site doesn’t look or work right, and you’re giving it the same code as you’re giving to other modern browsers. With IE9 Platform Preview 4, we’ve fixed over 100 community-reported issues. We will fix even more between now and the IE9 beta and want your feedback.
  • Consider the experience for IE9 Beta users if you find that sending the same markup creates more issues than you can resolve in your production site. It is possible that running your site in Compatibility View is better for your users.
  • Take advantage of HTML5, CSS3, SVG, DOM, ES5, and more… all described here in the developer guide.  We’re excited to run the amazing experiences you bring to the web using these new capabilities, taking advantage of hardware through IE9.

Platform Preview 4 is an important milestone on the way to beta. It is the last preview before the IE9 Beta. The IE9 platform is nearly complete. We ask that developers and partners start testing in preparation for the beta and prepare their sites to take advantage of IE9’s new capabilities. We continue to welcome your feedback via Connect.

Thanks,
Dean Hachamovitch
General Manager, Internet Explorer

Comments (283)

  1. adamtki says:

    @David Taylor – While enabling script debugging can have a small impact on performance, it would be unusual to see that sort of performance difference.  So if you have more information about this, we would love to get this feedback.  Please go to connect.microsoft.com/ie and send us your feedback.

  2. Anonymous says:

    Who cares what IE9 is capable of when only 25% of Internet Explorer users will actually be using it 5 years from now?  

    Unless IE9 stands for "Microsoft is officially not making web browsers anymore", I can't find a reason to be excited.  As it is, I can only look forward to the joy of creating an awesome HTML5 website loaded with all sorts of interactive media and then having to create a retarded version of the same website special for the morons still on IE6, 7, and 8.  Please, do the web developing community… nay, the WORLD a favor.  Immediately stop development on IE9, dump the code, and walk away forever.  Microsoft has fallen too far behind to do anything but harm the evolution of the internet by trying to stay in the game.  Simply put, you cannot and never will be able to catch up to the browsers that work with the dead weight of your disastrous history and your users' unwillingness to actually upgrade.

  3. Anonymous says:

    I can keep copying and pasting this from all the IP's in the world, guys…

    Who cares what IE9 is capable of when only 25% of Internet Explorer users will actually be using it 5 years from now?  

    Unless IE9 stands for "Microsoft is officially not making web browsers anymore", I can't find a reason to be excited.  As it is, I can only look forward to the joy of creating an awesome HTML5 website loaded with all sorts of interactive media and then having to create a retarded version of the same website special for the morons still on IE6, 7, and 8.  Please, do the web developing community… nay, the WORLD a favor.  Immediately stop development on IE9, dump the code, and walk away forever.  Microsoft has fallen too far behind to do anything but harm the evolution of the internet by trying to stay in the game.  Simply put, you cannot and never will be able to catch up to the browsers that work with the dead weight of your disastrous history and your users' unwillingness to actually upgrade.

  4. Danny says:

    This is such a great leap from IE8's low acid 3 of 20 odd, well done :D

  5. Mr. Pink says:

    Awesome! So my mothers farmville is going to run a lot smoother.  Makes me want to get it already.  How about some demos of real world apps performing better with it.  That's what we all want to see.

  6. Interframe says:

    Awesome! Cant wait to see the IE 9 beta, I hope the UI of IE 9 is a huge step up from IE 8, just like with performance.

  7. Demon says:

    Wow, welcome back to browser wars.

  8. anon says:

    Hamster Dance Revolution, IE Beatz, or MSNBC Video links are all broken. perhaps a temporary thing..?

  9. Edgard Castro says:

    I think one of the biggest advantages of Chrome is not even the web rending speed. Today this is almost nitpicking because we are talking about miliseconds. The thing with Chrome is that the browser is way snappier than IE or Firefox. You click a button, it does the action almost instantly. IE8 and Firefox 3.6 has this "mushy" feeling when clicking buttons, creating tabs and so on. And this is annoying as hell. Hope IE9 come closer or better than Chrome at that aspect.

  10. Delix says:

    Why is it taking so long to put out a public version of the browser

  11. blah says:

    ROFL, once again Opera is only mentioned in the corner of the chart.

  12. Vishwac [MSFT] says:

    @anon: Yes, this should be a temporary issue with the demo files replicating across several servers.

  13. Fiona says:

    The new IE engine seems to be very good, but i hope it will get a new gui and better addon support too.

  14. DanielHendrycks says:

    Nice, could you include Opera in your videos, considering it does well? (Also test the latest Firefox build, IE9 is unstable, so the Firefox in the demo should be bleeding edge, too)

  15. InfiniteLuke says:

    Btw: Firefox 3.6 isn't the newest Firefox. Theres Firefox 3.6.8 (huge difference to 3.6) and the furure one is Firefox 4 (currently in Beta2… therefor comparable to IE9)

  16. mathias says:

    Nice work!

    Are there any plans for supporting Web Workers (http://www.whatwg.org/…/current-work)? It would be great to have this feature, too.

  17. Asbjørn says:

    For some reason I'm getting ~4000ms (and that's not a typo) in Sunspider. This is a huge difference to previous platform previews, which always had performance that was comparable to Chrome. I'm running a Core2 Quad Q9550 with 8GB DDR3 RAM on Windows 7 x64 (Danish version).

  18. twisted coder says:

    That is kick ass !!!

  19. PQ says:

    Hi Dean, great work! Will IE9 support WebSockets?

  20. John Dyer says:

    You still can't move a [video] node in the DOM without it breaking. (This bug was filed back in PP3)

  21. badger says:

    Fixed most of my SVG/CSS bugs! Awesome work.

  22. Stilgar says:

    Better addon support is crucial. Microsoft has this MEF technology for the .NET Framework. It seems like a perfect fit for a new IE addon system.

  23. sushant_u_s says:

    This browser is greate I am waiting for this…………

  24. Austin says:

    Despite your best efforts at the video fallback, all I get is a "Install Silverlight" image and link; no way to watch the video. Not having a Theora or a Webm fallback is understandable considering your your stance on the codecs. Only having Silverlight and not Flash is just obnoxious. This is a pathetic attempt at same markup.

    At least provide a link to the video in the fallback. Why are people so scared to put a naked link to the video; you do want people to watch the video, don't you?

  25. FelipeKM says:

    Really awesome, I want this!

  26. Stilgar says:

    Austin the videos on Channel 9 are available for download. Not having Flash is the right thing to do. Silverlight does the same and is MS tech. Why the hell should they have Flash? If you were willing to install Flash you can install Silverlight too.

  27. EricLaw [MSFT] says:

    @Asbjørn: Do you have the Script Debugging feature enabled in the F12 developer tools? If so, your script will be running in the slower interpreted mode.

  28. Anti-troll says:

    >Firefox 3.6 isn't the newest Firefox

    No one said it was. And since they're showing both Firefox 3.6.4 and Firefox 4 getting smoked by IE9a4, it's silly to suggest otherwise.

  29. FremyCompany says:

    Only two words: Thank you!!

  30. John says:

    This is great. Thanks IE Team.  But how about WebSockets, File API, and Web Workers???

  31. ssl says:

    The line below (hiding a given layer) used to work before prev 3 and 4 – and still works with all other major browsers like those benchmarked above…

    document.getElementById(top.layerID).style.cssText="z-index:-1;visibility:hidden";

    however, easily fixed (prev 3 and 4) by using the old…

    document.getElementById(top.layerID).style.zindex="-1";

    document.getElementById(top.layerID).style.visibility="hidden";

    …but have I missed something here about cssText? (probably I have :-)

  32. Meni says:

    IETeam: WOW, Great work (i'll bet everyone's eyes are pretty red)

    PMs, and Marketing: Booooo, as i said before: please stop using the "same markup or script" use "standards compliant". Also stop dissing other browsers so much. When you get to their level of maturity and compliance we'll talk. in any case you guys are not on the same level as the tech guys.

    Still MIA: webgl

    One more point: Your insistence on hardware acceleration might hurt you in situations like netbooks, and mobile phones.

    overall, this is not the same old MS

    meni,

    open source fanboiii

  33. Richard says:

    This is a huge milestone! Congratulations to the whole IE Team!

  34. Blah says:

    Yeah, it's sure embarassing when IE trounces the "mature" browsers. Pfft.

    And apparently you missed the videos of PPB#1 smoking the other browsers on ION netbooks.

  35. Josemi says:

    Nice work! For when WebM support?

    Best regards,

    Josemi

    http://blog.josemi.net

  36. Frank Olivier [MS] says:

    @John Dyer The video DOM bug will be fixed in a future update; the fix did not make it into Preview 4.

  37. meni says:

    One more thing: these are great times. standards, open, patent free (well, video, let's forget that).

    2 extra things:

    a. Why the f%&&^%ing heck didn't this happen 5 years ago?????????????? who is to blame? which city do i have to look at?

    b. Just to stick it to the old guard at MS: silverlight – hahahahaha. windows phone 7 – has no HTML5 hahahahaha

    [just imagine this future: all sites written in silverlight.]

    meni,

    open source fanboiii

  38. Matt says:

    EPIC! Nice work IE team!

  39. John Hrvatin [MSFT} says:

    @anon

    The Test Drive site and download page took longer to propagate to all servers than it usually does.  Everything should be stable by now.

    Thanks!

  40. Asbjørn says:

    @EricLaw: No, it is disabled (the button says "Start debugging"). But it seems to be running in interpreted mode, as enabling debugging makes no change to performance. Strange indeed… Should I file a bug?

  41. Michael N. says:

    Congrats, there is now a ton of SVG content working without any changes:

    Atlas for the European Election 2009 in Germany

    vis.uell.net/…/atlas.xhtml

    Voronoi Price Kaleidoscope (consumer price index, basket of goods)

    http://www.destatis.de/…/PriceKaleidoscope.svg

    Personal Inflation Calculator

    http://j.mp/9FlI8o (some limitations in functionality)

  42. Tom says:

    IE Team, I love you.

    That's all.

  43. Danny says:

    However, we still need Text-Shadow and HTML5 Forms! Please?!

  44. Richard says:

    Sorry for the off topic, but why not display just a paragraph of every post in the blog's home (with a "read more" link)? That page is now very long and it is a little bit difficult to get to older posts. Thanks.

  45. abhi says:

    cant wait for ie9, it will be great ….low on disk useage and at par with quality

  46. sun says:

    Great work,

    i am proud of the IE team,

    HTML5 forms would be a nice innovation

  47. xD says:

    Parece k ba a llover, el cielo se esta nublando!

  48. EricLaw [MSFT] says:

    @ssl: We're not able to reproduce your cssText issue here. Could you please either file a bug on Connect or send me email containing a complete repro page?  thanks!

    @Asbjørn: I'm checking with the scripting team. I don't know of any way for you to get stuck in Interpreted mode. Inside the full version of IE, do you have the "Disable script debugging" boxes ticked inside Tools > Internet Options > Advanced?  thx!

  49. Garth says:

    Fantastic work.

    As others have also mention. Complete HTML5 form support would be awesome.

  50. Rene Pilon says:

    Based on this – you've just sold a few more Windows 7 licenses – Kudos!!

  51. Naval says:

    This is going to be a great competition to the chromes and fFOXes of the world!

    Keep it up MS.

  52. Steve says:

    This will run on Linux and MacOS right?  If not, IE will remain the only browser that exists that can only run on Windows.  IE9 looks great no doubt, but your harming the general computing ecosystem by only releasing it on Windows.

  53. John says:

    "Take advantage of HTML5, CSS3, SVG, DOM, ES5, and more… all described here in the developer guide.  We’re excited to run the amazing experiences you bring to the web using these new capabilities, taking advantage of hardware through IE9."

    How do you propose we do that since XP is still being sold?

  54. TiredOfLies says:

    XP is no longer "being sold". Anyone silly enough to pay for XP is getting ripped off.

  55. Firefox Fangirl says:

    I only have a few questions which I think haven't already been asked. 1) Will you be releasing a 64-bit beta in September or just a 32-bit one? 2) Will you be shipping a 64-bit version of this browser at launch? IE 8 64-bit can't use HTML5 and has no Adobe Flash Player. Furthermore, as much as I'd like to see the death of flash in favor of HTML5 the web is thick with it (and it's used for more than just video too) so a 64-bit version would be useful. Having established this: 3) Has anyone here heard anything recently about Adobe's progress on 64-bit Flash Player for Windows? I swear, they move so slow…

  56. @Firefox Fangirl says:

    Adobe has discontinued all AMD64 forks of flash. If you still have any browser on any system using 64 bit flash you should uninstall it since it is out of date and will not receive any new security patches.

  57. Anonymous says:

    @Firefox Fangirl

    Adobe has discontinued all AMD64 forks of flash. If you still have any browser on any system using 64 bit flash you should uninstall it since it is out of date and will not receive any new security patches.

  58. Firefox Fangirl says:

    I see. Thank you for the news, Anon.

  59. JW says:

    Great news!

  60. Steve Warnock says:

    @Asbjørn  I got 315.33ms on PP4 and 3660.8ms on IE8.  You sure you did not accidently get IE8 in there instead of the PP?

  61. lynn says:

    Very nice. Html 5 Forms would be icing on the cake.

  62. Daniel Hendrycks says:

    Hi, here are some things I noticed:

    Sputnik- IE9 had 100 errors (no typo), little room for improvement; great job

    HTML5Test- IE9 has to improve dramatically, I cannot emphasize this enough

    Peacekeeper- IE9 is slow

    Acid3 and SunSpider- IE9 has little room for improvement; great job

    General Site Compatability (with little to no coding errors)- Slightly better than mediocre, many well coded sites are broken

    Dromaeo- IE did not finish

    Charka- RAM hog

  63. Wow says:

    The "value" of your analysis is plain to all who read it.

    You have no idea how much memory Chakra uses, and you provide no support at all for your compatibility claims. "HTML5Test" is an entirely fabricated benchmark which uses browser-sniffing hacks and has obvious bugs (h264/MPEG4, anyone?); it grants points for formats which are entirely nonsensical (bloated PCM audio, seriously???) and/or deprecated before they were even completed (WebSQL).

  64. physssssssssssssssssssssssssssssssssychadelic says:

    the physch test is pyschadelischious

    congrats to Microosft for stepping up to the challenge and removing the IE team handcuffs

  65. Hyper says:

    I tried to install the new preview on 2008 R2 and it say does not supported.  Why? All the previous PP works fine.  Any workaround/fix?

  66. Sid says:

    Should have flipped the D2D switch on in Firefox!

  67. Kishor Gurtu says:

    Great work. Wish it would also support the Location API as location-based apps are taking off.

  68. Poobalam says:

    Exciting work guys.  Been using Firefox and Chrome lately for developments and had to say IE team have steep curve to claim.  What makes the Chrome great is its UI responsiveness, its snappy start-up time, able to rearrange tabs, fast page-loading.  It seems IE9 show great promise in rendering aspect of it and wants to see developers choose IE9 in the future as there default testing browser.  Don't want to see old style thousands of tool-bar being able to installed and screw IE's performance, that's so 90.  Please eliminate the nightmare of tool bar in IE.

  69. someone says:

    Good job IE team. I stopped checking out anything and everything about Internet Explorer ever since you decided to drop XP support. It's way too insulting it ignore 60% of the world but good development for W7/Vista users.

  70. Patrick Dengler [MSFG] says:

    @badger

    let me know which bugs in CSS/SVG that we may have missed (patd@microsoft.com)

  71. stonee says:

    Html5 videos are encoded in h264 format, h.264 offers better compression and quality ,so we're playing videos natively with IE9 since it supports h264-encoded videos. And how to converter html5 videos to other video formats, we need video converter like ifunia, who declared they are dedicated in creating affordable and easy multimedia software to simplify your digital life, to do it?

  72. Razor says:

    so good, that 60% of market browser won't be able to see :D

  73. Paul says:

    Remember IE team, you have to do well on the HTML5Test.com too. It's like the Acid test for HTML 5. Web Workers, Web Sockets, HTML5 forms, Geolocation, WebGL, session and user history.

  74. mike says:

    It's just too bad that all your other browsers are such a horrible disaster. Hopefully you can stop making our lives hell with this release.

  75. hamster high score says:

    two hours of my life dedicated to hamster dance. high score of 17,100 on nvidia 26000. never dropped below 60fps. love what you're doing.

  76. noname says:

    Method appendChild in IE9 is very-very slow (more than ie8). Why? ( found a little test on russian site – vflash.ru/test_javascript.html )

    i think in the first ms needs optimized dom+js. in the second – web workers, web sockets, web gl etc…

  77. someone says:

    I use chrome only,I don't see anything about IE !!!!

  78. Really impressed with the new version of IE9, it just keeps getting better and better. Are we likely to see CSS3 Transforms & Transitions before a beta?

  79. ZappedOne says:

    Awesome job IE team! I just want to second the opinion of others and hope that the UI performance/design receives the same level of attention as everything else has up to this point. Keep it up! Thanks.

  80. Stu says:

    Wow, first version where comments are generally positive, congrats! (you must have pretty thick skins by now anyway!)

  81. Danny says:

    Please add tabs on top in the GUI like in Chrome and other new browsers. The titlebar is such a waste of screen space.

  82. Ben says:

    IE Team: Congratulations. The progress being made in IE9 is extremely impressive. I am very impressed. Having given up on IE many years ago in frustration, I eagerly look forward to giving IE9 a fair chance as my default browser. Keep up the excellent work. There are many people who really appreciate it.

  83. Charlie says:

    Great work guys – a massive improvement on past versions of IE. However, I feel I should point out that one or two CSS3 features, especially gradient support will be sorely missed if not implemented – I'm sure a lot of developers would love to see an end of the days of chopping buttons and backgrounds out of images to get gradients.

    A community project is in progress to bring some CSS3 display functionality to IE6-8 (http://css3pie.com/), but native support in IE9 would make everyone's lives a lot easier. A useful feature chart can be found at http://www.impressivewebs.com/css3-click-chart. Hope you can squeeze some of this in!

  84. Paul Rouget says:

    Do you enable Direct2D rendering in Firefox 4?

  85. Rostislav Hristov says:

    window.history.pushState please!

  86. dontchangegui says:

    This is great but please DON'T change GUI. I heard you are going to clone chrome's! No! Speed improvements are good but only the gui really matters. Chrome/opera gui is a plain joke (tabs on top???), completely unworkable and frustrating. Firefox doesn't have tab grouping (and plugins for it don't work well) and acceleration button. Please leave ie8 gui mostly unchanged (although, search in history/favorities would be nice).

    People choose browsers because of their usability, and usability depends on gui mostly, speed doesn't matter much. If I wanted chrome's gui I would be using chrome.

    The superior gui is the only reason people are using ie8.

  87. Mark says:

    This is looking really good, the IE team have far surpassed my expectations!

    I second the calls for CSS gradients, isn't that just the kind of thing Direct2D is good at accelerating?

    It's also good to see the realisation that performance is not just about how quickly you can run isolated javascript benchmarks, but how quickly you can get the results back to the DOM and render them. This is something Chrome is particularly good at, and why it feels faster than FF.

  88. Pierrot says:

    Great work!

    A couple of points that you've probably heard before but- what the heck:

    * it would be great if the work your doing now was paired with an aggressive drive to get people to upgrade.  Could this update be made automatic by default unless the user opts out?  In general I'm against this kind of thing, but the world is waiting for this browser to become The IE.  Until that happens, your work is pretty useless and we might as well force IE9 and IE8 to run as IE7.  That way we have to write crap code once, not crap code (for IE7) + kinda crap (for IE8) + nice (for IE9).

    You guys and gals are doing a great job- I hope you get the support to get your work onto as many PCs as possible as quickly as possible.  Please don't make us wait until 2015 to really use any of the cool stuff you've put together!

    * Could we see more frequent releases which would incorporate bug fixes, performance tweaks and minor functionality?  For example, if I knew that IE9.1 would support CSS gradients then I'd be ok with leaving a solid color background for a few months… These updates could be transparent to the user maybe?  This would maybe reduce the disappointment felt by those whose favourite features didn't make it into IE9- because IE10 will be a looong way away.

  89. anonymuos says:

    I just have to say: The progress made with IE9 is immaterial as your browser won't support 60% of world's computers and 71% of enterprise customers who still run Windows XP. I say IE team start living in a more realistic world. You want us to dump IE6, give us IE9 then. You think XP users deserve a browser now that fares so poorly as shown in your cross browser test results summary table? Please treat all Windows customers equally and give the IE9 renderer+Chakra for XP but keep hardware acceleration exclusive to Windows 7. Does our feedback have no effect on your Windows 7 fanaticism?

  90. eXPerience says:

    What's the point really targeting a minority of users? 27% (Windows 7's 14% and Vista's 13%). It's an improvement which will pay off several years later. Useless. MS still does not get the web and puts its vested interest of pushing Windows ahead of the web.

  91. martens.ms says:

    What about comparing the Internet Explorer Platform Preview with Chromium Nightly under Linux?

    What about that? Scared? Chromium is more as twice times faster under Linux than on Windows systems. Think 'bout that.

  92. Bala Natarajan says:

    New Kick ass UI please..Final Nail in the Coffin!!

  93. giuseppe says:

    Object.defineProperty is broken:

    i get

    "SCRIPT445: Object doesn't support this action"

    when used to update an existing descriptor:

    var o = new Object();

    Object.defineProperty(o, "test", {get:function(){}});

    Object.defineProperty(o, "test", {set:function(){}}); //ERROR

    i did not get the error in the preview 3 and both the setter and the getter worked

    ecma262 5th edition (december 2009):

    15.2.3.6

    "The defineProperty function is used to add an own property and/or update the attributes of an existing own property of an object."

    i would have submitted this to Connect, but for some reason my private postal address, phone number and other irrelevant data is demanded, which i will not give away. and as the entered data is validated against geo data, i find it too tedious to come up with valid fake data.

    g

  94. giuseppe says:

    note: the additional empty lines between the lines of code in my previous comment were added automagically by the blog engine. they were not present in the textarea where i entered them.

    also the first click onto 'post' seems to always reload the page without posting or any feedback on what happened.

  95. Ruben says:

    Ooh, goody. No CSS3 text-shadow, no background gradients, no multiple columns, etc. It would be nice if you'd be a little more candid about your CSS3 support (which is pretty limited).

  96. fr says:

    Looking good, now that your Sunspider and graphics performance is good, I would be interested to hear why your Peacekeeper score is still quite far behind Chrome, Opera etc.

  97. Ali says:

    Please

    - html5 form (input element type & attributes & validation)

    http://www.w3.org/…/forms.html

    - Drag and drop

    - Geolocation

    - Web Workers

    HTML 5 test (300 Full Test):

    Chrome 6.0.472.22  ->  217 / 300 (+ 10 bonus points)

    Chrome 6.0.458.1  ->  221 / 300 (+ 10 bonus points)

    Chrome 6.0.447.0  ->  219 / 300 (+ 10 bonus points)

    Firefox 4.0.nightly  -> 199 / 300

    Chrome 5.0.375.86.Final  -> 197 / 300 (+ 7 bonus points)

    Safari 5.0.1  ->  165 / 300 (208 / 300)

    Opera 10.60  ->  159 / 300 (+ 7 bonus points)

    Firefox 3.6.6  ->  139 / 300 (+ 4 bonus points)

    IE 9 P.4  ->  96 / 300 (+ 3 bonus points)

    IE 9 P.3  ->  84 / 300 (+ 1 bonus points)

    IE 8.0  ->  37 / 300

    http://html5test.com

  98. giuseppe says:

    update to my first comment about defineProperty:

    it appears that the 'configurable' field of the descriptor is now 'false' by default, as opposed to 'true' in the previous previews. this seems to follow the ecma spec more closely:

    8.10

    "Any fields that are not explicitly listed are considered to be absent."

    where 'absent' would be interpreted as 'null' (or 'undefined'?) and thus evaluated to 'false'

    so i believe this is not a bug. (glad that connect didnt want my participation ;)

    g

  99. JP says:

    Great stuff IE team! Gets better with each preview.

    Enjoying reading this blog a lot, keep up the good work.

    I'm at work so haven't installed this yet, but I'm running Server 2008 R2 – is the alleged problem listed above likely to be rectified or not?

    If any of the IE team get a spare couple of minutes (!) please put in some WebGL support, as that really would be the icing on the cake.

    (I realise it would be a hell of a lot more work than a couple of minutes)

  100. harry richter says:

    @ fr

    Peacekeeper seems to be very hardware-dependent!

    On my Dell Latitude D630 (Win7, X86) I get:

    Chrome 6.0.472.22:  2229

    IE9 PP4:            2228

    Opera 10.60:        2198

    I would not call that "lagging behind"!

    Cheers

    Harry

  101. Jayp says:

    Great stuff IE team! Gets better with each preview.

    Enjoying reading this blog a lot, keep up the good work.

    I'm at work so haven't installed this yet, but I'm running Server 2008 R2 – is the alleged problem listed above likely to be rectified or not?

    If any of the IE team get a spare couple of minutes (!) please put in some WebGL support, as that really would be the icing on the cake.

    (I realise it would be a hell of a lot more work than a couple of minutes)

  102. giuseppe says:

    @EricLaw, Asbjørn

    here IE9pp4 is also 'stuck in interpreted mode':

    sunspider 0.9

    on german vista home premium 32bit (all updates and sp)

    IE9pp4:

    script debugging disabled:  Total: 4519.8ms +/- 0.3%

    script debugging enabled:   Total: 4521.4ms +/- 0.5%

    IE 8.0.6001.18928:

    script debugging disabled:  Total: 4678.6ms +/- 0.5%

    script debugging enabled:   Total: 5164.4ms +/- 0.4%

    both mentioned debugging checkboxes unchecked in IE8 preferences

    g

  103. EricLaw [MSFT] says:

    giuseppe: Do you find faster performance if you *CHECK* both checkboxes labeled "Disable Script Debugging" in the IE Internet Control Panel advanced tab? The blog team is looking into the "Scroll to top" problem upon posting a comment.

    FWIW: The OS marketshare claims made by various other comments are outdated, and I believe it's worth mentioning again that Win7 is the fastest growing OS in history.

  104. MathML_(Time_has_Come) says:

    How about implementing MathML?

    There is an awesome font for math that is free and open to use

    The STIX fonts.

    Webkit (Safari, Chrome) are working hard at implementing MathML.

    Gecko (Firefox) is already MathML enabled.

  105. giuseppe says:

    @EricLaw, Asbjørn

    possible cause (solved my problem):

    a third party 'addon' (or how they are called) that i thought was only available to IE8 was appearantly loaded in IEpp4 (possibly earlier pp too)

    same configuration as above:

    IE9pp4:

    no debugging, no addons:     397.0ms +/- 1.1%

    all other results are about the same as before.

    conclusion: IE9pp4 loads IE 'addons' and IE 'addons' can engage interpreted mode for javascript in IE9pp4.

    g

  106. fr says:

    @harry richter none of the results I have seen on my machines or posted by other people have been that close, there is usually a significant difference as shown here img411.imageshack.us/…/benchies.png (not my results but posted by someone else today).

    It is not the only test where IE9 still seems to have some catching up to do either, for example if you run the DOM core tests at Dromaeo there is also quite a difference.  I would have done the full test suite but it errors in the current preview.

    http://dromaeo.com/?id=112176 IE9 PP4

    http://dromaeo.com/?id=112177 Chrome 5.0.375

    http://dromaeo.com/?id=112178 Opera 10.61

  107. Thangaraj says:

    Really great work! I'm definitely rooting for IE9!!

    Just hoping this sort of growth keeps happening even after IE9 goes stable. As someone else mentioned, if we knew 9.1 will be released with bugfixes & feature updates that IE9 missed, there will be much less anxiety involved ;)

    Sincerely hope you don't make us wait 2 years for IE10 with no releases in between!

  108. giuseppe says:

    @EricLaw

    sorry i did not see your response before submitting my previous comment.

    debugging was disabled by IE8 settings, i just wrote some nonsese not remembering whether checking enables or disables.

    thank you about the status of fixing the 'scrolltop' problem.

    g

  109. Brian LePore says:

    I'm currently seeing 5 FPS for the Hamster Dance (I cannot believe I just typed that) on my Vista box.

    I noticed that filters worked are still working in CSS files. I thought they were going to be dropped when in standards compliance mode? I know how to use conditional comments to give older IE certain features, but when the changes are 1-2 lines I find a CSS hack for IE6-7 is simple enough … so now I think I'm going to need to find CSS hacks for IE8 and below to give filters and let IE9 behave like everyone else.

    I think the above issue may be why my company's main application are tending to behave very quirky. Right now it looks like we're going to need to do much more work to support IE9 than we needed to for IE8. I am actually surprised because we do use feature detection instead of browser detection for 99.999% of the things we use. Having some serious issues getting the debugger to run for some reason.

    I know it is useless to ask for an exact date for IE9's release, but can we get an idea of when IE9 won't be out by? Like say "it will not be released in the 4th quarter of 2010".

  110. Osvaldo Doederlein says:

    On my system, Psychedelic score is IE9p4 = 1870, FF4b3 (beta3 candidate builds are just out; with D2D&DW turned on) = 1754. So IE9 is just 9% better. Latest Chrome 6 is atrocious though, with just 8 rpm.

  111. Asbjørn says:

    @EricLaw, giuseppe: For me, checking Disable Script Debugging (other) IE8 Advanced options did it. I'm now at 352 ms. By the way, IE9 is looking great. I'm hoping that the UI is just as improved as the engine.

  112. Senthil says:

    Do you indented to support websocket? This is the best thing happen after Ajax, for real rich interactive experience and mostly other things are some how emulated using scripts.

  113. india says:

    use EPIC the best browser in the world

  114. MisterX says:

    Upgrade ES5 Conformance Suite (es5conform.codeplex.com) with new test please!

  115. Wibble says:

    Does it run on a Mac?  I don't have Vista or Vista II.

    About time you fixed your browsers.  Better late than never.

  116. EricLaw [MSFT] says:

    @Wibble: No, IE9 requires Windows Vista+.

    @giuseppe: What sort of add-on are you referring to? The Platform Preview build does not load BHOs, Toolbars, Explorer Bars, menu extensions or other add-ons. Only ActiveX controls used by the current page are loaded.

  117. badger says:

    @Patrick Dengler

    Looks like bug 571984 on Connect is still not fixed. But this bug appears in most browsers to some extent.

  118. giuseppe says:

    @EricLaw:

    the addon was http://www.ie7pro.com/

    but i cannot reproduce the negative effect by re-enabling the addon, so my conclusion was wrong.

    IE9pp4 now always matches the faster time at sunspider.

    but i can asure both debugging options were disabled (checked) and script debugging was off where stated. weird.

    sorry for false alarm

    g

  119. oiuawfeiijfe says:

    Finally, progress!!!

  120. Matt says:

    A 100% Acid 3 score should be taken for granted these days?

    Will it go out as a 'critical' update?

  121. Tyler says:

    This make me so happy. You guys are making the web and the world a much better place with the work your doing. *single tear*

  122. Chris says:

    Very well done with IE 9 thus far.

  123. MSFANBOY says:

    GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO GO

  124. wiiman says:

    epic!

    5 more points in acid3 will make ie9 the ultimate!

  125. Literacy? says:

    For those of you who read the post, Dean basically says that they don't plan to collect higher than 95, since the last 5 points come from features that are being deprecated and thus aren't going to be implemented.

    Further proof that tests like ACID are *harmful* to the web, because they result in browsers being forced to race to implement features that they should not and will need to remove later.

  126. Zero says:

    im using google chrome and i got 100/100 in the acid test.

  127. where is the download link? says:

    The link in the beginning goes to the IE test drive site, but there is no link to download the preview 4?

    thanks

  128. Jovian says:

    I am still facing issues with KB2028560.  I faced the same issue with PP3.  PP4 also crashes.  But when I uninstall KB2028560 things work fine (though a lil slower in graphics perf.  I have a Lenovo T400 with ATI Mobility Radeon HD 3400.  Device manager says that driver is up to date.  I went on the ATI website, downloaded the latest driver and tried installing again.  Still same issue. KB2028560 is causing issues consistently.  Any solutions you can suggest here?

  129. ssl says:

    @EricLaw :

    Mailed you and url / example on that cssText issue? – I wrote about yesterday.

  130. Kai says:

    Nice work indeed!

    However, can someone from Microsoft tell me what are the plans for these features: CSS gradients, CSS transitions, CSS transforms, web sockets, web workers, IndexedDB, HTML5 forms, text shadow, file API, offline web apps, geolocation, CSS animation, WebGL and the flexible box layout? I am not sad if it wouldn't support all those, but I would like to hear if there are plans on supporting some of them (and what specifically). Thanks for making things possible!

  131. Aethec says:

    Please fix the white text on black background issue. Currently it is either too bold or too ugly.

    PS : What's with those comments? It seems my first one wasn't posted at all.

  132. SeanJenkin says:

    We are taking a look at the blog commenting issues some folks are experiencing. Hang tight while we investigate them. Thanks.

  133. Gixx says:

    Still no multi-column support… Focus on the important features and not the hype. Please.

  134. faris says:

    keep it up Microsoft i want to in your company what i can study to work with you

    1st computer engineering

    2nd computer science  

  135. Swathi Ganapathi [MSFT] says:

    @Hyper: The Platform Previews are only designed to install on Windows 7 and Vista OS's.

    Please see: ie.microsoft.com/…/Default.html

    Thanks!

  136. Hyper says:

    @Swathi

    Thanks for your answer.  Is there any reason why 2008/R2 suddenly get left out for PP4?  Of course I know almost nothing about the details, but it almost looks like it got arbitary block at installation.  PP1-3 works fine with 2008 R2, and I can't really see why PP4 doesn't.

    I'm sure this won't happen with the beta (that would be awful), but really, please don't do that.

  137. Daniel Hendrycks says:

    IE does badly on the Celtic-Kane JS benchmark, too. jsbenchmark.celtickane.com/Run.aspx

    (I listed other things before)

  138. Ian says:

    Man I am happy and sad about IE9 at the moment. For some reason it does not seem to show any of the images on my website. Additionally it collapses all my input fields and buttons due to the CSS 3.0 styles that I am using… I just hope that the beta will at least make it possible for me not to have to rewrite my site to make it compatible with IE9.

  139. Ilidio says:

    As a web developer I'm very concerned about these 3 websites:

    http://html5readiness.com/  (Google Chrome usually is in the first place here)

    http://html5test.com/ (Google Chrome usually is in the first place here)

    And check some thinks here:

    http://www.html5rocks.com/

  140. Ilídio says:

    Please

    - html5 form

    (too)

  141. Nanalich says:

    Is it still possible to use the new JScript engine with ECMAScript 5 features outside of IE?

    Like in WSH and/or HTA?

  142. Ilidio says:

    And please give more attention for software updates. You have to learn with old mistakes. For example IE6 has 8~9years, I'm just tired of IE6. This must be integrated with the browser.

    Is there a away to make this more automatic. Users usually don't understand the importance of having a browser update.

    This very important for the evolution of the Internet. I really bet this will make the diference in the future.

  143. SamIAm says:

    @Daniel Hendrycks – IE9 does much better than Firefox  on the Celtic-Kane JS benchmark that you mentioned (both 3.6.8 and 4.0 beta 2). Not sure why you think it did not do well.

    @Ilidio – why are you "very concerned about these 3 websites"? Take a look at the Cross Browser test site samples.msdn.microsoft.com/ietestcenter  where IE9 does much much better than any other browser. Anyone can publish a test that will make one browser or another look bad. Chrome, Safari and Firefox look bad in the one above. why is this any less valid than the ones you quote? This nonsense about html5test and other sites is a smokescreen for IE haters.

  144. Nanalich says:

    These so called HTML 5 Test websites used tons of "-webkit-" and/or "-moz-" things, so you believe such things like that can tell the truth?

  145. dave says:

    IE9 is getting better but, I cannot get the sunspider js script test to score anywhere near the posted score. It takes over 4 seconds for it to finish. Google 5 takes 321ms to finish.

    Also, IE9 doesn't always redraw when the page changes. I hope this gets fixed.

  146. fchivu says:

    Which is going to be the default rendering mode in WebBrowser control (WebOC) in IE 9? In IE 8 it was IE7 and the rendering mode was controlled by the FEATURE_BROWSER_EMULATION registry key as described in the article from blogs.msdn.com/…/more-ie8-extensibility-improvements.aspx . This produced a lot of confusion and our customers asked why IE 8 (installed on their machines) displays a web page in one way and our library displays it in a different way. Thanks.

  147. David says:

    I would be happy if also WebSockets would be implemented!

  148. PeterRow says:

    Preview 4 is totally hosed for me. I installed it on my work machine (Win 7 Pro 64bit), it downloaded and installed 3 hotfixes as part of the install, rebooted the machine. When I run it as soon as I move the mouse anywhere over the window it crashes with a "Windows Internet Explorer Platform Preview has stopped working" message, I uninstalled and reinstalled using the MSI by running from an admin command prompt whereby it installed but no longer had to install the hotfixes and still I get the same error. I have yet to install it at home (also Win 7 Pro 64bit).

    Previews 2 and 3 worked fine. The only exception being with 3 it refused to install as it kept failing when downloading the hotfixes. I downloaded/installed the hotfixes separately and then installed preview 3 and then it worked fine.

    A little gutted as I was keen to see the latest improvements and to see if the google map info tips in a web app I've been working now works correctly or not as in preview 3 they didn't.

  149. David Taylor says:

    Hi Eric and Dean,

    I was getting the same performance issues with Preview 4 as well (about the same speed as IE 8).  I hit F12 but could not solve the problem.

    Eventually, I loaded IE 8, and in the advanced settings ticked the "Disable Script Debugging (other)" – I already had the other one ticked.

    When I reloaded IE 9 Preview 4 this fixed the issue.  You better do something about this before beta.  In fact given the IE popup that lets you set this stuff when it hits a script error…..maybe you should always turn off script debugging when IE 9 first loads; or at least show some UI hint/hover that script debugging is enabled.  That will be a real user issue if someone clicks on something they don't understand and ends up getting javascript running 10 times slower ;-)

    Great job though!  I love how it is progressing.

    Dave

  150. John says:

    It won't install on Windows Server 2008 or R2, even though previous previews did. Any reason for this? Is this how the final product will be too???

  151. PeterRow says:

    Anybody else having the same problem as me, see my comment from earlier today at 8:40 am (GMT)?

    Would love to try Preview 4 but it crashes as soon as you try to use it.

  152. hAl says:

    @Nanalich 5 Aug 2010 7:50 PM

    Looking at the artikel it is no longer the Jscript engine but a Javascript engine that IE9 is using.

  153. Ilídio says:

    @ SamIAm

    because I want to use HTML5 forms; Web Workers; Web storage; and Geo Location.

    For example, many add-ons (extensions) are using this now. So if IE9 have these ones. Webdevelopers will start to export those add-ons .

    Everybody will win at the end

  154. EricLaw [MSFT] says:

    @John/Hyper: The PPB builds are not supported on Windows Server (the fact that earlier builds could be installed was a bug, since not all Servers have the required prerequisites). The final product will install on Server 2008 and Server 2008R2.

    @fchivu: IE9 will behave like IE8 in this respect and WebOCs will specifically need to select the later rendering mode or they will use the IE7 Standards mode by default.

    @hAl: Both of these are technically "ECMAScript engines." Surprisingly for many, "JavaScript" is a trademark owned by another company.

    @Nanalich: JScript9.dll is not presently supported for use outside of IE9, HTAs, and Web Browser controls, so no, it won't be loaded by WSH.

    @PeterRow: I'm sorry to hear this isn't working for you. If you're a technical user and could grab a crash dump, I'd love to have a look: blogs.msdn.com/…/collecting-internet-explorer-crash-dumps.aspx

    @Ian: It would be VERY useful if you could provide a URL of your site, or file a bug so we can see what you're encountering.

  155. Ilídio says:

    @Nanalich  So for example, you have these on webkit:

    -webkit-transform: rotate(-45deg)

    Maybe should be like this:

    transform: rotate(-45deg)

  156. Frank says:

    It's really nice to see what Microsoft can achieve if they actually try.

    I'm still very disappointed that we aren't going to get native vp8 support out of the box.

    And I'm not entirely sure how fair it is to compare the IE9 Platform Preview with hardware acceleration enabled with Firefox 4.0 beta with hardware acceleration disabled. At least on my PC Firefox still slightly outperforms Preview 4 in the tests shown when I enable Direct2D/DirectWrite (which is currently planned to be enabled be default in the final build).

  157. PeterRow says:

    Hey Eric; yeah I'm fairly technical (supposedly).

    I ended up using procdump.exe recommended that I got from blogs.msdn.com/…/what-is-a-dump-and-how-do-i-create-one.aspx because the link you provided ended the web installer just kind of sat there for ages not downloading anything so I used the following command options:  procdump.exe –e –h –ma –x

    The files a pretty big even when zipped so here are the Skydrive links to them:

    File 1 of 2: cid-175ac91641484a3e.office.live.com/…/PeterRow%5E_1%5E_of%5E_2%5E_iePreview4dump%5E_100806%5E_151958.7z

    File 2 of 2: cid-175ac91641484a3e.office.live.com/…/PeterRow%5E_2%5E_of%5E_2%5E_iePreview4dump%5E_100806%5E_152003.7z

    Hope this is of some help; I'm going to try it at home tonight and see if it works there, fingers crossed.

  158. hAl says:

    @Ilidio

    Geolocation is only really usefull for GPS equipped devices and as such is  possibly more likely to be implemented on the mobile IE versions first.

    It however is also one of the features that Microsoft could add later to IE9 without breaking compatiblity.

  159. Steven says:

    Say "NO" to IE toolbar.

  160. ieblog says:

    We are testing a fix to the comment box issues.  If over the next few days you are still experiencing issues with having to press the "post" button multiple times, please let us know in a comment.  Thanks!

  161. Jovian says:

    No love for Jove?  I posted a crash issue I faced with PP4 (and PP3), KB2028560.  See my comment above.  Can MSFT guys comment on this?  I dont care if there is no permanent fix for PP4.  But ensure u or ur partners (I suspect nvidia) get things sorted out for beta.  Looking forward to kick ass IE9.

  162. Daniel says:

    Will there be a more frequent update schedule? Ex: IE 9.5

  163. David S says:

    Guess my comment got lost from earlier.

    Anyway, I'm excited to go home and try out the new platform preview.

    However, I have a major gripe with those asking for XP support. You want the latest technology and the latest standards. Yet you ask for this on a platform that is 10 years old and shipped with IE6. This puzzles me. Windows XP did it's time, let it rest. If you want technology to move along upgrade. This is a business after all and I don't think it is unfair to ask for an upgrade to a decade old OS.

    As for cooperations, their major concern is $$$ and they won't upgrade just for shinier looks. If the user base like yourself is using newer technology they will eventually upgrade to capitali$e on this. Please just let XP rest.

  164. Hyper says:

    @Eric Thanks for your answer.  Hopefully by "final product" you don't mean I have to wait until the RTM/RTW of IE9 for Windows Server 2008/R2, don't I?

  165. Erikke says:

    Радует что наконец начали стандартам общим приходить, а то ужас был какой-то уже давно сижу на Mozilla Firefox и Google Chrome.

  166. Erikke says:

    Удачи вам, желаю всеж чтоб придерживались стандартов, конечно некто не запрещает свои фичи делать, тогда и народ потянется к IE

  167. michealW says:

    foutth ie9 platform preview can not support windows server 2008 r2 x64?

  168. RGCode says:

    so good , thnx

  169. fchivu says:

    @Eric, thanks for the answer. Do you have any reason for why the WebOC will still run in IE 7 mode by default? For IE 8 the reason was there was not enough time for products using WebOC to adapt to IE 8 rendering, but now with IE 9 why isn't rendering mode moved to IE 8 at least?

  170. Chris says:

    95 on Acid 3 is fine by me, if it means the IE team spend more time on other things, good job!

  171. Ilídio says:

    Please IE team make the some:

    CWmike writes

    "Taking a page from rival Google's playbook, Mozilla plans to introduce silent, behind-the-scenes security updating to Firefox 4. The feature, which has gotten little attention from Mozilla, is currently 'on track' for Firefox 4, slated to ship before the end of the year. Firefox 4's silent update will only be offered on Windows, Mozilla has said. Most updates will be downloaded and installed automatically without asking the user or requiring a confirmation. 'We'll only be using the major update dialog box for changes like [version] 4 to 4.5 or 5," said Alex Faaborg, a principal designer on Firefox, in the 'mozilla.dev.apps.firefox' forum. 'Unfortunately users will still see the updating progress bar on load, but this is an implementation issue as opposed to a [user interface] one; ideally the update could be applied in the background.' Unlike Google, Mozilla will let users change the default silent service to the more traditional mode, where the browser asks permission before downloading and installing any update."

    Source: news.slashdot.org/…/Like-Googles-Chrome-Mozilla-To-Silently-Update-Firefox-4

  172. krishna says:

    This is really WoW!

  173. EricLaw [MSFT] says:

    @fchivu: The reason is the same, and no, it has nothing to do with "not enough time" and everything to do with the fact that if you ship a browser update that breaks existing applications, users will not install it. Any application that wishes to change its rendering mode can do so with the feature control key, all other applications can depend on the stable behavior or the Web Browser Control without worry that a new IE version will break them. (Keep in mind that in many cases, it's completely infeasible to assume that an application update would even be made available were the application to break, because either the application is no longer maintained, or users wouldn't have any way to get the update, etc).

    @michealW: yes, as mentioned above and in the release notes, PPB builds are not supported on Windows Server 2008 and R2.

  174. Perf Testing says:

    IE9 Preview 4 Psychedelic score of 1840. FF nightly with hardware acceleration turned on score of 1656. Chrome nightly score of 32. Opera 106 score of 412. That means IE9 is 11% faster than FF and 6000% faster than Chrome and 446% faster than Opera 10.6.

  175. Richard Fink says:

    Dean,

    I find this post's explanation for tossing SVG Fonts under the bus completely unpersuasive. I don't know what to make of it. You've written nothing about the technical merits or the lack thereof of SVG Fonts at all. Just excusy dance-arounds.

    >"Support for SVG Fonts in the web development and font communities has been declining for some time."

    To what kind of "support" are you referring? Actual implementations or enthusiasm about the technology? If the latter, who's enthusiasm has waned?

    Did somebody do a poll?

    Chrome, Safari, and Opera all have running code for SVG Fonts.

    >"There’s already been discussion without objection of dropping SVG fonts from the Acid3 test. The community has put forth a proposal in the SVG Working Group to give SVG Fonts optional status."

    First, I object. So now you have an objection.

    Second, so what? Are you in favor of SVG Fonts or not? Does it give developers valuable options that they would not otherwise have, or not?

    >Instead, developers can use the Web Open Font Format (WOFF, supported in IE9 Platform Preview 3 as well as other browsers) for both HTML and SVG content."

    Huh? Excuse me for being a little in the dark here, but could you be specific about where and how WOFF – which is a wrapped TTF or OTF file and works no differently than the system installed web-safe fonts – can be manipulated via script in the way SVG fonts can be? Are you saying WOFF is *the same* as SVG Fonts?

    >"It works well in conjunction with the CSS3 Fonts module and has broad support from leading font vendors (e.g. here, “a majority of font makers have already settled on

    WOFF or services like Typekit as their format of choice”)."

    This has the standards process backwards and it isn't even true. If it's true, prove it.

    >WOFF fonts are a better long-term solution for many reasons"

    Says you. What are the differences, why are you declining to suport SVG Fonts, *that* is the point.

    Some elaboration would be appreciated.

    Regards,

    Rich

  176. Sincerity says:

    Petulant children shouldn't sign their post "regards" as it confuses the reader. Every technology has its fans, but that doesn't mean that every technology survives. SVG Fonts won't be one of the survivors.

  177. Richard Fink says:

    @sincerity

    SVG Fonts might not deserve to survive OR be supported in IE. But questioning why that is isn't petulance, it's a request for elaboration.

    Regards,

    Rich

  178. CvP says:

    HELL YEAH!

    ps: css transitions please. box-shadow is still buggy.

  179. Ankit says:

    Good Job IE team….. back to browser War….. beat the Chrome

  180. CvP says:

    I was trying out the SVG Dice.

    "Line: 156

    Error: Exception thrown and not caught"

  181. Chunk says:

    Dear IE Team,

    I absolutely CANNOT take you seriously until you explain why IE fell so far behind every other major browser.  Even IE8 is miles behind something like Firefox as far as usability and power.

    You can promise us the world in IE9.  And you could even deliver.  But it doesn't matter at this point because you have no credibility.

    You need to explain the past in order for me, a serious web user, to consider you a legitimate option.

  182. chunk says:

    Dear IE Team,

    I absolutely CANNOT take you seriously until you explain why IE fell so far behind every other major browser.  Even IE8 is miles behind something like Firefox as far as usability and power.  Any serious web user considers IE to be UNUSABLE in everyday use.

    You can promise us the world in IE9…. And you could even deliver on your promises…  

    But it doesn't matter because you have no credibility.

    You need to explain the past in order to move forward.   You have a serious perception problem and you can't sweep it under the rug.

    But if you just come clean, explain the past, and release a quality product, then you might actually be considered by serious internet users who set the trends.

    The decision is yours.

  183. hAl says:

    @Richard Fink

    You have not given us any practical example of SVG fonts  in a webpage that you can't do with WOFF fonts that warrants keeping svg fonts but still you object to svg fonts not being included?

  184. hAl says:

    Still double posting actions.

    Now I am getting a "Post" button on the first post action and then a "Publishing" button for a secondary posting actions.

  185. Ilidio says:

    @chuck just don't go there. I'm not a MS employe, but I understand process of development. Even thought, some of technology, only now, is possible. Sometimes it's better delayed then produce something wrong, because w3c specification, is unclear. So now is time to make things right. It's that why, I'm asking some of html5 and automatic updates.

  186. the_dees says:

    @Richard Fink

    Safari and Chrome share the same engine, so they only count as one engine.

    Now we have two vendors that decided against SVG Fonts (MS and Mozilla) and two vendors that decided in favor of Acid3/SVG Fonts (WebKit and Opera). And somewhere Konqueror still exists as well (which can be counted as an engine on its own because it greatly differs from WebKit).

    Both MS and Mozilla have stated various reasons why not implementing SVG Fonts makes sense.

  187. EricLaw [MSFT] says:

    @hAl: The idea is that after you click the "POST" button, you don't need to click the button again; your post is in the process of posting when the button turns to "Publishing."  Can you elaborate a bit more (e.g. step-by-step) on the problem you're encountering?

  188. antiTroll says:

    Heh… with a name like "Chunk" it's hard to really consider you a "serious web user."  When you try to speak for everyone, you immediately lose all credibility.

    A browser that smokes all of the other browsers in hardware accelerated HTML5 performance, IE9 is already very a "credible" browser and it's not even hit beta yet. Frankly, IE8 is still an extremely "credible" browser, with more users than every other browser, and better security features than all others.

  189. hAl says:

    @Ericlaw

    As I saw the button change to display display "Publishing" I took that as that I was required to push it again to get the comment published.

    You should not change the button like to confuse people. If I have to wait then just display a message beside it saying  "Wait for the comment to be posted" or something.

    p.s. The button looks like a 3 part picture that does not line-up when using IE6. Very ugly.

  190. hAl says:

    I  just had to wait 30 second for the comment to be posted without any hint that the post was recieved. Of course I am going to push the button  that reads publishing agian and again

  191. Meni says:

    Hi,

    I do not want this to become a flame war, especially not on this blog, by of one of the most open and different groups at Microsoft (I love your work IE team)

    AntiTrool, I'm not Chunk, but to call IE6/7/8 a credible browser is making fun of the web. I accuse Microsoft of sabotaging the web since around 2001. If it was up to the old guard at Microsoft, we would be all writing sites with Silverlight right now. The MBAs at Microsoft would love the idea that the only way to browse the web would be from a Windows platform (they kinda succeeded at around 2005), and they would love the idea of a new proprietary image format where you would pay $0.001 for impression. [I'm sure fanboys like you argue that with market share of 95% YOU COULD WRITE SITES WITH VBSCRIPT, or better silverlight, today. I AM SURE YOU DON'T EVEN SEE WTF'S WRONG WITH THIS. i see this as very dangerous that people like you are advising clients to go this way every day]

    I acknowledge MS's contributions like xmlhttprequest, and ActiveX (Just kidding on the latter)

    Apple and Google contributed ten times more to the web then MS.

    HTML5 should have been here 5 years ago :-((((

    Meni, an open-source and open-standards fanboi

  192. Meni says:

    BTW, This blog system really fubar. Eric, if it needs explaning then…….

    Just move to Blogger already :-)

  193. Errr.... says:

    Uh, hal, did you not get the memo? IE6 is dead.

    Meni, you're making fun of yourself. Could you do that somewhere else please?

  194. Justin says:

    Quite pleased with the progress so far.  someone above said not to change the UI – I think this would be a major mistake – if IE9 is going to turn an about face with supporting web standards then it needs to do the same with its aging pathetic UI.  The chrome is cluttered and unorganized, and no longer re-arrangeable like IE6 was.  Firefox and Chrome have far superior UI's that are well organized and easier to use.

    Having to manually move the reload button back to the left of the address bar to get it to work like all other browsers (even older IE versions) is a PITA and not easily discovered.

    Disabled scrollbars on the right of the browsing area are a sign of bad initial design. Until there is content beyond the vertical limits, the disabled scrollbars are a waste of space.

    The address bar lost major functionality in recent versions, removing the favicons, obscuring the true URL (the URL should be on the left, the often redundant or duplicated labels should be on the right)

    The "goto" highlighted URL at the bottom of the address bar is useless. Users know that if they press enter they are going to go to a site/keyword.  Removing this redundant item from the addressbar dropdown provides either more room for meaningful, helpful content or minimizes the amount of screen it obscures (either way is good)

    All dialogs that popup from IE need to be revisited as they are cramped and disorganized (try finding anything in the advanced tab, or the endless scrolling of restricted access in zones.

    Why doesn't typing "http" in the address bar show me all kinds of matches from my history? This kind of stuff should be real simple to fix!

    Add open frame in new tab to the context menu

    Add a spellchecker to IE! its 2010 dude! every other browser has had this built in for years now!

    Not sure whose silly idea it was to change the text color to white (on a light gray background) when clicking on the command bar "Page", "Safety" and "Tools" links but it was obviously the poorest choice available.

    flush-right-ing the command bar was also a major mistake.  Allow the user to put it wherever they want it please!

    creating new tabs is still very slow in IE8 (please don't tell me my addons are the issue – I'm not listening to that crud anymore) please fix this in IE9

    IE9 should enable tab tearing… to tear a tab off into a new window (or re-join orphaned windows)

    and so, so, so, much more.

    thanks

  195. hAl says:

    @Justin

    Tekst is not white but plain black here when clicking on the commandbar "Page", "Safety", and "Tools".

    Mayby some toolbar changed your UI?

  196. wetran says:

    @Blah,

    "And apparently you missed the videos of PPB#1 smoking the other browsers on ION netbooks."

    And apparently you missed the fact that the vast majority of netbooks are NOT, I repeat NOT, ION netbooks. The vast majority of netbooks, budget desktops, and even many mainstream desktops, have nothing but a lowly Intel GMA IGP, and with the Core i3 and i5 integrated graphics cores, they leave more and more PCs end up with only some poor performing Intel graphics chips.

    And from what PPB#1 has shown, the D2D accelerated IE9 actually performs worse on Intel graphics chips than those non-D2D-enabled browsers that relies solely on CPU, especially for Core i3 and i5 systems.

    Considering the still dominating market share that Intel IGPs have over the graphics card market, IE9 needs to run at least as fast on a Core i3's integrated graphics chip as the other browsers running on the i3 CPU itself in order to claim superiority of HW-accelerated browsing. the IE team can say whatever they want, but the average user will still think their Core i3/i5 systems are modern systems with modern graphics chips, and the average users don't do other CPU-intensive tasks when they browse the web, so IE9 needs to be at least as fast as those other browsers even with Intel IGPs and just browsing the web, else many people will actually think IE9 is slower with HW graphics acceleration.

  197. mitch074 says:

    That preview is getting very interesting. However, if benches and tests show great improvement, real-life use shows many regressions:

    – box-shadows are no longer taken into account to determine the document's viewport. As an aside, in some cases the shadows are not properly drawn (I've filed a bug about that). Well, at least, the blur radius has been fixed, and, yes, the spec doesn't say how the shadow should impact the viewport (Webkit says it doesn't, Presto, Gecko and IE9pre3 say it should).

    – there are MANY bugs present on bullet lists: position not respected if bullet is replaced by an image (a bug I just filed), and several other bugs currently present in Connect.

    – inline-block images send floats running in tears in some cases (and this is a regression over IE7/8, which were not very kind to floats).

    So, IE9pre4 good on the benches; but bad on pure CSS and HTML compliance.

  198. hAl says:

    @wetran

    Even low level IGP's currently provde decent amount of hardware accelleration. IE9 on my old budget celeron laptop with intel IGP makes mince meat of Chrome on my much faster desktop computer on all gfx related tests on the testdrive site.

  199. PeterRow says:

    Following on from my earlier post where I'd tried IE9 on my work machine (Win 7 Pro, 64Bit, 8GB RAM, Intel Quad Core, ATI graphics, Dell) and it just crashes as soon as it loads the test drive page I can happily confirm that on my home machine which is virtual the same machine just with better ATI graphics and better sound it works fine, hurrah!

  200. Andreas says:

    @MSFT:

    Obviously (and understandably) you do not comment on requests for specific features. But can you indicate whether the fact that this was the last developer preview before a public beta means that this preview is more or less considered feature complete? Or is there still some ongoing feature development that *might* end up in the final relase?  (e.g. in the CSS3 area)

    Any hint is appreciated, thanks!

  201. CvP says:

    @hAl

    "Of course I am going to push the button  that reads publishing agian and again"

    Pushing a button over and over again when it says it is doing something doesn't make any sense.

  202. hAl says:

    @CvP

    Waiting 30 seconds to inform a commenter that his comment is posted does not make sense. That should be done in less than a second. If only to inform the commneter that his comment is queued and it could take some time before it would be published.

  203. Popper says:

    <<Not sure whose silly idea it was to change the text color to white (on a light gray background) when clicking on the command bar "Page", "Safety" and "Tools" links but it was obviously the poorest choice available.>>

    It was your choice. You customized your Windows theme colors.

    Moving the Refresh button is "a PITA"? You must have something else stuck up there, since it takes one click on one menu.

    Sticking your fingers in your ears and saying: "I know that you guys would know best because you profiled the code, but I refuse to believe that slow tab startup is caused by add-ons" isn't likely to cause anyone to ignore the facts.

    Mitch: What are your bug numbers on CONNECT? Happy to validate and vote for them!

    hal/CVP: take your button argument elsewhere, please. The IE team didn't write this blog software. (hal, sniff your tcp and you'll see that the POST request takes that long. Their server is slow).

  204. Dean says:

    Just testing the new commenting to see if the bug was fixed.  Ps as others have noted if you are going to change the button text you need to gray out and disable the button so users know something is happening.

  205. adamtki says:

    @ Jovian, We have received another report internally on the same configuration as yours and are actively investigating this.  There's also an IE Connect feedback you can view at connect.microsoft.com/…/584697.  If you have additional information, you can post the information there.

  206. Daniel Hendrycks says:

    "@Daniel Hendrycks – IE9 does much better than Firefox  on the Celtic-Kane JS benchmark that you mentioned (both 3.6.8 and 4.0 beta 2). Not sure why you think it did not do well." Sorry, I was comparing it to Opera and Chrome, not a slower browser like Firefox. (I don't want to hear about why you think Firefox is great, it is getting better, and I am excited about that)

    "These so called HTML 5 Test websites used tons of "-webkit-" and/or "-moz-" things, so you believe such things like that can tell the truth?" You can check if it does that by viewing the source.

    "IE9 Preview 4 Psychedelic score of 1840. FF nightly with hardware acceleration turned on score of 1656. Chrome nightly score of 32. Opera 106 score of 412. That means IE9 is 11% faster than FF and 6000% faster than Chrome and 446% faster than Opera 10.6." Of course it beat Opera and Chrome by a large margin, it did not have HW acceleration available, yet.

    Could you please have a built-in update system, or something? That way we do not have to deal with the IE6 issue again!?

  207. ClueTrain says:

    Daniel: He did look at the source. The browser-specific hacks are plainly visible.

    IE6 users are using IE6 because they choose to, not because WindowsUpdate doesn't work.

  208. SamIAm says:

    @Daniel Hendrycks – you start with "orry, I was comparing it to Opera and Chrome, not a slower browser like Firefox.", and follow up with a quote about Opera and Chrom being so much slower. Can you make up your mind?

  209. Mitch074 says:

    The bugs I reported are 584673 (box-shadow is not taken into account in parent element's measurement), 584670 (list-style-position:inside is not respected) and 573183 (Rendering regression on inline-block elements).

  210. adamtki says:

    @Mitch074.  It's great to get your feedback.  We are currently investigating the rendering regression issue and it's something we hope to address since it worked properly in IE 8.  For the recent ones you filed, we will process them as soon as we can!

  211. SeanJenkin says:

    We've updated the post button on comments to indicate it's publishing the comment. The button is also disabled while that's happening to help with the confusion… We are continuing to investigate the time it is taking to post a comment. Thank you for your continued patience.

  212. Nanalich says:

    I would like to see a hostable version of JScript9.dll =[ it helps

  213. Daniel says:

    I tried the Fourth IE9 Platform Preview. Is the most impressive thing that I have seen in years. The hardware acceleration is just… WOW. Is the biggest gift to developers in years.

  214. mitch074 says:

    @AdamKing: I submit bug now not to have to work around them later. I hope my reports are complete and precise enough.

  215. Meni says:

    A Question for all who speak of the new IE putting all other browsers to shame (  i.e. MS fanboys :-)   ):

    Before the quite shocking announcement of standards support by the IE9, what were your thoughts of HTML5? That is, in 2009 did you even care about SVG, canvas, standard event handling and the rest? Or you had to wait and see what's Microsoft views on these are.

    What do you think of WebGL today?? Or you are waiting to see if Microsoft will support it?

    What I'm getting at, is instead of getting so belligerent about FPS, we should all be happy, like i am, about the news, and hope that the web will quickly go forward. My real hope is that MS, Google, Mozilla, Opera, and whoever, will create a much much much faster standards body, and do even more. I'm talking better sound support, video camara support, maybe other client-side scripting languages, maybe bytecode support for scripting. I feel now is the time. [before some manager at microsoft will wake-up and yell: IETeam, WTF are you doing, are you trying to kill MS? stop this interoperability nonsese at once]

    love and peace

    -meni

  216. wetran says:

    @ClueTrain,

    "Daniel: He did look at the source. The browser-specific hacks are plainly visible."

    They are not hacks, it's standards-compliant prefix for experimental features. The W3C standards clearly allowed and recommended the use of vendor prefix for experimental features. And for those already standardized parts, there are prefix-less attributes provided, so for every "-moz-border-radius", there are "border-radius", for every "-webkit-box-shadow", there are "box-shadow", etc. etc. So the tests are quite valid as far as it's testing the HTML5/CSS3 feature availability.

  217. José says:

    @Errr….

    sorry to get this terrible news. IE6 is not dead yet. And that's the problem! There is some stats from a public website

    1. IE 8.0   21.162    60,05%

    2. IE 7.0   11.009    31,24%

    3. IE 6.0     3.060      8,68%

                                          100%

    K Government Rejects Calls To Upgrade From IE6

    rss.slashdot.org/…/UK-Government-Rejects-Calls-To-Upgrade-From-IE6

    Please do some research before submitting anything here…

    That's why I'm asking too for automatic updates and more html5 improvements (HTML5 forms; Web Workers; Web storage;).

  218. José says:

    "These so called HTML 5 Test websites used tons of "-webkit-" and/or "-moz-" things, so you believe such things like that can tell the truth?"

    Why webkit make those prefixes. So they can not break the interoperability and they can also be removed in the future.

    you just don't get it. It's all about inovation and a better web.

  219. giuseppe says:

    @Meni

    I agree!

    Vector graphics for the web, for example, were ready almost a decade ago. For non-technological reasons a small number of people could not agree on one single technology, leaving a lot of people in the situation where vector graphics for the web imply redundancy (bad) or cannot be used at all (not good).

    Ten wasted years (very bad).

    Also I believe that those small number of responsible people from above knew that browser vendor independent vector graphics will ultimately be inevitable and would be highly appreciated by a lot of people right away.

    And now we finally get 3D for the web. Very similar situation. The technology is ready. Demand is present. Ultimately inevitable. Now all that is needed is one single standard.

    I hope i will still have enough dignity to not be thankful to whoever is the last relevant browser vendor to support that standard in ten years.

    g

  220. must says:

    IE6 is far from dead, especially in Asia, where there's the largest number of internet users in the world. This is my site's stats for the last day

    Msie 8.0    483182    24.4 %

    Msie 7.0    251277    12.7 %

    Msie 6.0    1182516    59.8 %

    MSIE    1927158    97.5 %

    That's the sad state of the internet here in Asia. And lowly website programmers like me can't do much but to work for IE6 compatibility when 60% of our potential customers are stubborn IE6 users

  221. IE6IsDead says:

    Jose/Must, you're missing the point. The point is that "automatic updates" don't help– users on IE6 *want* to be on IE6. They'll disable automatic updates and decline IE updates just to stay there. How do we know? Because IE7 and IE8 have been on the automatic update list for 4+ years now, and still don't have 100% uptake. But the worldwide numbers are good and getting better.

    As for the UK government, unless you happen to be selling something to them, who cares? Just make a site that works with IE8 and if they can't use it, then they'll eventually see the need to upgrade.

  222. mitch074 says:

    @Must:: while Asia is still the area where IE 6 is most used (due, primarily, to the use of pirated Windows XP with automatic updates disabled), not ALL websites cater to the Asian market – and anyway, due to the completely different alphabets used there and their different appreciations for blinking stuff, loud colours and page overloads, if you really are interested by the Asian market you need to create two websites anyway:

    – one for East Asia

    – one for Occidentally-tainted countries (Americas, Europe, Africa, and Central Asia).

    Now, you can use 'recent' technologies for the Rest of the World ™, and focus on IE 6 for Asia (eventhough I don't recommend it, because it would push away all those that use more recent/free browsers in Asia – and there are).

    Personally, I disable CSS+JS and use fallbacks for all browsers older than IE 7 (and IE 7 gets no more than a token set of workarounds; forget about pixel-perfect rendering).

    @Jose: your stats simply show use share on the basis of IE browsers. If, we say that this website sees 25% of non-IE browsers, IE 6's share falls to 6%.

    On the websites I manage (which see a few thousands visitors per month), half (if not 3/4) of the IE6 hits can be traced back to bots that spoof its UA identifier. If we take that into account, it means that less than 4% of your website's visitors are actually using IE 6. Should you double development time to support IE 6? Or should you still develop FOR IE 6, leaving all other browsers in the lurch?

    All signs point to 'neither'.

    I thus welcome IE 9's improved use of standards (and not to be dismissed performance) quite warmly. Let's just hope for continued regression testing and fixing, even after RTW.

  223. Blah says:

    Mitch, it's not about piracy– IE doesn't do WGA validation anyway. Even if it did, you'd see much higher share of non-IE browsers there. The issue is that sites in that region cater to IE6 and there's a chicken & egg problem– no need to support a later version unless users use it, no need to install a later version unless sites require it.

  224. a.ha says:

    Regarding "same markup", I wonder if IE 9 will finally understand

     document.getElementById("thisShouldFloatRight").style.cssFloat = "right";

    as any other browser does. Up to version 8, IE always insisted on styleFloat (which any other browser ignores) in place of cssFloat.

    Furthermore, I wonder if IE 9 will apply the rule

     body, #ie8#hack { background: lime; }

    Not that #ie8#hack made a lot of sense, but it's indeed a valid selector, so IE 8 is wrong in dropping the entire rule. This will be used to hide CSS from IE 8 if fixed in IE 9. If not, it will be used to hide CSS from "IE 8 and later", causing much pain – later.

  225. the_dees says:

    @Adam Kim [MSFT]:

    Since you had a look at some Connect reports mentioned here, I'd like to ask this question:

    I have reported some issues on Connect. While most reports contain a notification that you (Microsoft) were able to confirm there are some (older) reports that seem to be ignored/overlooked. For example:

    559441 CSS – A report about incorrect computed values using inheritance

    559618 CSS – Incorrectly working visibility style on inlines

    566518 CSS – outside list-item marker incorrectly clipped when overflow != visible is applied

    566968 CSS – Incorrect rendering with block formatting contexts

    I don't know why these reports appear overlooked. They vary from small issues to rather big issues. On some issues work was done, other seems completely ignored.

    There's also an issue (574235 XML – lang attribue is incorrectly understood as language definition) where I provided additional information after your response. However, since then I have never heard from you again.

    I'd really like to know if all issues are really looked at :)

    @a.ha: You should try the Preview, both questions will be answered.

  226. a.ha says:

    @the_dees, thanks, but I can't, because of the system requirements. And I won't personally bother further, since authors will have to support styleFloat for another decade at least in any case. And whether #ie8#hack "works" or doesn't work in IE 9 is something we'll see; both is OK for authors. It's really those who talk about "same markup" that should bother.

    And thanks for pointing out 566518 (CSS – outside list-item marker incorrectly clipped when overflow != visible is applied). This is something I just recently ran into myself with IE 8. The funny thing about it is that IE 6 and 7 get it right, but IE 8 and recent versions of all other browsers get it wrong.

  227. the_dees says:

    @a.ha:

    I see. Well, cssFloat is suported FWIW.

    Multiple ID selectors are supported as well, but buggy at the moment. The behave like class selectors incorrectly behaved in earlier IE versions.

    See report #584333 (Multiple ID Selectors are only partially supported) for which I created a testcase.

    About list-markers: Unfortunately no one implements this correctly at the moment, but the specced behavior is definitely the one preferred and expected by authors.

  228. bluepicaso says:

    umm… It looks good.. But will soon lag behind others… :P as usual.

    IE are so worse of the kind. no border radius. no damn custom fonts…. just sick

  229. Netanel says:

    Can you look this site? It's doesn't work well in the 4th preview yet.

    http://html5readiness.com/

    By the way, good work! I can't wait for the first beta!

  230. hAl says:

    @netanel

    That that site doesn't work is mainly caused by the site serving different stuff to IE9 thant to other html5 browser browser making the page useless for what the domainname is suggesting that it is claiming to do.

  231. Julien says:

    That's good! but I'm asking too for automatic updates…

  232. hAl says:

    @bluepicaso

    IE9 supports borderradiusbetter than several other browsers and has supported custom fonts throu @fontface since IE4

  233. Jose says:

    @ IE6IsDead

    That's why the browser should force the update, and not the S.O. itself (Windows – update list)

  234. Simon says:

    Too bad that the "report issue" page doesn't work! I'm in IE9 and go to menu "Report Issue" > "Report and Issue" , I'm taken to the url go.microsoft.com/fwlink which redirects me to connect.microsoft.com/ie, a "can't find the page" page.

    But here's what I want to report about IE9 fourth platfrom review:

    1. anti-aliasing is really really ugly on some common fonts like tahoma and arial, especially when they are bold.

    2. It's really really SLOW compared to latest firefox, chrome, opera and even IE8! This is when it comes to javascript animations and draggable objects etc.

    Drag the knobs on this page for example, developer.expressionz.in/…/slider_using_mootols_1.2.html

    Really lagging behind the mouse in IE9. In firefox or even IE8 they are really sticking to the mouse.

    Now this is just an example, I have this problem everywhere!

  235. Miguel Web Developer says:

    Let's say … IE9 is out with out "HTML5 forms". Then MS makes update IE9.1 with "HTML5 forms".

    As web developer I will have to code something like:

    /// Example – Start

    if (  IE >= 9.1  ) {

       useHtml5Form(…)

    }else{

        useNonHtml5Form(…)

    }

    /// Example – End  

    And it will be IE9 users without "HTML5 forms". So I will start thinking maybe is to soon to implement that and I will delay the devel (on new tecnlogy) for 2 or 3 months.  And will be more versions of IE to be worry about. So I don't like the end picture….

    So why not to force the browser to update itself and it will be less versions to worrie about. Change something in that. It's a new browser maybe you can. Take the example from Firefox

    news.slashdot.org/…/Like-Googles-Chrome-Mozilla-To-Silently-Update-Firefox-4

  236. Matt says:

    There's a reason that IE has 200% of the marketshare of Firefox and Chrome– They don't randomly make updates that change the web platform functionality that cause users to find their browser unreliable and break the applications they care about.

    I think it's unlikely that the IE team is going to take a page from the losers' playbook.

  237. Jose says:

    @Matt Don't go there… That is not the main reason Matt. The reason is IE was here first.

    The browsers are trying to crash the pluggin itself. Complain to adobe.

    Other browsers are winning users. It's because that they are doing something wright

  238. Prior Semblance says:

    I just read up on HTML 5 forms after reading Miguel Web Developers comment and they look amazing.  I hope IE9 supports at least a couple of the new form features, at the very least autofocus should be really easy and maybe search/email/url input.

  239. Haw-Bin says:

    Will the release version of IE9 be installable side-by-side with IE8? I think this would be a key feature to support testing of both on the same machine.

  240. must says:

    @IE6IsDead,

    I'm not sure what you are talking about "you're missing the point", since my point is exactly that, IE6 is far from dead here in Asia, and every website programmer has to cater to IE6 compatibility and can't do much about it. I have put banners to Chrome, Opera, and Firefox, and that still doesn't help the usage of those non-IE browsers. That's exactly my point, a LOT of stubborn people are using IE6, so IE6 is NOT dead yet.

    @mitch074,

    Since I'm living in Asia, the intended customers of my company are all Asians. And Asians don't have "different appreciations for blinking stuff, loud colours and page overloads", just that the ads are less regulated here, so you see more ads here like those in English porn sites :P

    The biggest difference is still that, most people here just stays with IE6 and XP, and web programmers here have to code sites with IE6 compatibility if they want to have a job at all :(

  241. hAl says:

    @Simon

    I tried your example and the dragging of the slider nobs on your example page is indeed a lot less smooth on IE9 than it was on IE8.

    There seems be regression in performance there.

  242. Jay says:

    I can't see the video. Can you provide a WebM version? If you're having trouble producing a WebM encode then perhaps an upload to YouTube would be an easy way to achieve it.

  243. MyTime says:

    How does IE team think about WebGL?

  244. Dick says:

    I hope to see it available for linux, bsd, osx and ios4 too (with same performance)

  245. jesper says:

    finaly the internet explorer 9 comes soon becuse other browsers sems faster and better so i almost want to leve explorer and mybe google crome but i hold ut and now 9 is here soon as an beta continue to work microsoft on internet explorerer 9  explorer has alwyas been the first browser in the world and the best so lets mace it to the best so microsoft dont fail with iex 9 now as you have done with 8 and the beta will come around 19 sep^^

  246. hAl says:

    @MyTime

    WebGL is not a web markup language nor any other kind of W3C webstandard.

    It is now an arbitrary browser extension of javascript/DOM to plugin 3D support into a browser whereas 3D support in the future should really be introduced trough normal grafical markup (like in a future SVG) or trough extending the canvas element  to 3D in HTML5 trough the normal standardization processes in W3C.

  247. Danny Moules says:

    Given your obsession with hardware utilisation from the web and HTML5, your lack of appreciation for WebGL is baffling. Your use of the phrase 'markup' to describe WebGL is also pretty confusing use of the term – I know people have interpreted it wrongly.

    @hAl WebGL provides web applications with tight access to GPU hardware through OpenGL ES, exactly the same as any C++ application would have. Also, it is a HTML5 canvas element context. It is not arbitrary, it follows standardisation from the Khronos group who developed OpenGL and all the (other) major browser vendors. The only people who aren't on-board are Microsoft – because they develop the competing library DirectX. The rest of the industry is moving with WebGL.

  248. hAl says:

    @Danny Moules

    Why would Microsoft implement 3d hardware accelleration for their browser trough OpenGL. That would be nonefficient if they hava a more advanced 3D hardware accelleration architecture in DirectX.

    If WebGL was handed over to W3C it could be standardized in a more generic way so that it would not just map almost 1-1- on OpenGL but would become indepent of the underlying hardware accelleration architecture. Then it would be usefull to support for Microsoft it with DirectX.

    If WebGL is not submitted to W3C I would expect Microsoft to support 3D developmnents in SVG and/or in Canvas and then support those with DirectX and in that case WebGL would quickly fade away as there is no use supporting some opengl architure dependant gfx solution when there is an architecture independant open W3C standard that supplies the same features.

    However since WebGL is currently a pluggable javascript feature that is inserted trough canvas anybody can easily make a WebGL plugin for IE. So we migth see some third party webGL plugin mayby.

  249. David S says:

    IE9 is very promising so far. It definitely has a strong set of features. My only concern is that it may not render these features the same way as other browsers. Especially in the case of CSS. I've looked at few CSS3 examples on the web and they render differently from browser to browser (not just IE9). Things like gradients fading from dark to light instead of the opposite direction.

    Also the really cool features like transforms and animation would be very welcome. Not everyone is going to jump into SVG and Canvas as these seem to be specialized technologies like Flash. Most web devs are going to want the full power of CSS/Javascript first and then resort to something else for fancy stuff. However having consistency across browsers is very important in the CSS arena. I hope a lot of attention is paid to CSS rendering.

  250. Jay says:

    "Real-world HTML5 performance often reflects the entire browser’s performance, not just the JavaScript engine in isolation. This video, for example, shows a side by side comparison of different browsers running an HTML5 canvas."

    A day later Download Squad conducted another test with four browsers instead of two. It's available on YouTube (I watched the WebM version in Minefield):

    http://www.youtube.com/watch

    Download Squad's conclusion is, to quote, "Firefox does look like the HTML5 champion". It's interesting to note even though IE9's development is narrowly focused on just one platform it is nevertheless out performed by Firefox, a cross platform browser. I'm surprised by that. You'd think that limiting development to just one platform would make optimizing performance easier.

  251. Bull says:

    The poor quality video on YouTube shows IE's performance destroying the other browsers, and anyone who runs all four browsers simultaneously and doesn't understand why that yields bogus results is plainly unqualified to review much of anything.

  252. Jay says:

    @Bull: The video is available in 720p. I'd hardly call that "poor quality". It's interesting you say that running browsers simultaneously can only yield bogus results. The comparison video Microsoft linked to was running two browsers simultaneously. So you're saying that Microsoft is trying to deliberately mislead readers of the IE blog?

  253. hAl says:

    @Jay

    That rendering performance test needs to run full screen or at large windows so it uses high resolutions for the test. Not in 4 by 4 mini windows. You can easily try yourself that IE9 pp4 is te fastest in those test then.

  254. Klimax says:

    @hAl: And I'd add that only one browser/test should run at a time and have focus (so its threads get proper priority)

  255. adamtki says:

    @the_dees

    We are investigating all the issues you’ve reported including the XML Lang attribute. We’ll follow up directly in each Connect bug.  The feedback you provided has been of tremendous value. We really appreciate your continued contribution to enhancing Internet Explorer.

  256. Indrek says:

    Hell yeah, that is serious improvement, can't wait to see it out already. Finally some serious buttkick from MS.

    Instead of other browsers forcing IE to be up to date, it will be the other way around very soon. I hope those last 5 failing acid tests will also pass when it's shipping-ready, holding thumbs for full 100 ;)

    How long until we see the final version of IE that can replace older versions of IE's ? From current progress I'd guess one more platform preview and then RC1 RC2 and then final ? Will IE team keep delivering incremental updates to IE after final release ? Like 9.1, 9.2 etc ? And make the updates possible without restarting ? As I understand the JS DOM is now built into IE so it does not have any huge cross-dependencies to other system parts that would require restarting for smaller updates ?

    Why I'd enjoy to see same commitment from IE team in the future is that whenever specs get updated, IE would be updated and it would force other browsers to play catch up more intensely and create decent ongoing competition.

    Anyways, short summary: GREAT WORK!!

  257. 불닭먹자 says:

    와~~ 진짜 대단합니다. IE9가 파이어폭스 4.0도 이기는군요…

    잘하면 크롬보다도 더 빨라질 수 있겠습니다.

  258. Mitch074 says:

    @blah: considering how many Chinese laptops I had to fix (I used to work in a university's IT service, and students would come to me to help them make their system work with our systems), I'd say that in China, it's harder to find a legit system with a real Windows license than any other. Heck, even boxed laptops use bogus license numbers and come preinstalled with pirated crap. And of course, those are never updated – and nobody bothers to install IE 7/8 if it's not present in automatic updates. Some did, instead, install Firefox, or had Safari as default (they didn't change the default after installing iTunes, back when Apple forced it along and hijacked user preferences).

    So when I say IE 6's prevalence comes from pirated software, I'm not faking it. And, making a website that works in IE 7is no more 'difficult' than for IE 6 (layout tricks remain the same, it's only 'advanced' CSS table-less tricks that go up in flames).

    @must: my wife and her friends spend a lot of time browsing Chinese websites (she's Chinese). Not one one of them was exempt from blinking stuff, even with ad blockers set up and working. I also roam Taiwanese hardware websites looking for drivers and BIOS updates – the same. And when my wife and I get asked to build a website, we get the following requirements:

    – looks like other Chinese sites,

    – must have a minimum of 30% of total area covered in red, gold and/or blinking stuff, 'like these websites: -insert list-'.

    I'm not making that up. And I don't make those websites.

  259. Miguel Web Developers says:

    @Mitch074 I'm with you…

    Please solve this problem as engineers that we are all.

  260. Miguel Web Developers says:

    @IE6IsDead … Solving IE6 problem is out of MS hands. Forcing the browser IE9 to make updates (and not the S.O.) will prevent IE6 mistakes

  261. PleaseEngageBrain says:

    Mitch: IE8 is in automatic updates (as was IE7). Only one of many of your typical factual errors.

    Miguel: You've completely missed the point. The same reason IE6 still exists is the same reason that IE9 will not force updates– Microsoft's customers demand the ability to run legacy versions. That's a simple fact, and nothing in the world has changed that over the last decade. If you don't understand that, you should stop commenting here.

    Indrek: They've already said they'll never get the last 5 ACID points since the features they test are becoming deprecated. JavaScript was never the reason that IE required a reboot for installation of updates (look at your patches and you'll see that JScript.dll gets update like once a year, but IE patches are 6 times a year). There will be no IE9.1 and 9.2 and so on for the same reason there was no IE8.1. and 8.2 etc– site owners and customers don't want it. (The 0.01% of real customers who read this blog don't represent what the market as a whole wants).

  262. Olivier says:

    Please give us IE9 for XP!

    My colleague has Windows 7 on a Core i7 laptop with 4GB RAM and it is much slower than my XP on a core2duo E8500 and 3GB of RAM. There's no way I'll install Windows 7 because it'll be so slow that I won't be able to work efficiently. And I can't change my computer just for a browser.

    I'm stuck with IE8 for a long time :( and I won't use others browsers because they don't work that well :(

  263. Anti-Luddite says:

    Correctly configured, a Core i7 on Win7 will smoke a E8500 running XP, Vista, or Win7. Your colleague's computer is misconfigured, likely either running unneeded utility software, or a massively underspec'd hard drive. One key thing to remember about laptops is that they can be configured for aggressive power-saving (and many are, by default) which significantly impacts runtime performance.

  264. Miguel Web Developers says:

    I will try to explain my point view. Without automatic updates (done by the browser itself) the evolution of HTML5 will be slower.

    This updates should done and forced by the browser by default (not the .O.S. – Windows Vista or Seven). Like more Google Chrome or Firefox are going to do. Without this, I believe, that the evolution of HTML5 will be incredible slow.

    I want to things move on… I love the web

  265. StillMissingThePoint says:

    You're still completely missing the point. The point is that the majority of corporations, governments, and other organizations of any size will not install a browser that changes the behavior of the platform in daily/weekly/monthly updates. Every time a patch is released, these organizations have to retest all of their mission-critical sites– and that's when the IE updates deliberately *aren't* making platform changes. That's why IE continues to have marketshare that's 200% of all competitors combined.

    Yes, we can all wish that this wasn't a fact, but it is one of the basic facts that everyone, including Microsoft, has to live with.

  266. MyTime says:

    Please support WebGL.

  267. mitch074 says:

    @PleaseEngageBrain: please live up to your name. Or learn basic logic: if AU is disabled, then IE7/8 won't show up in AU.

  268. meni says:

    StillMissingThePoint, you just summarized why the web is so fined up, and why everyone is, rightly or not, blaming  Microsoft for this situation. My advice to these organizations that want stability: Please please please go have an orgy with Microsoft development tools (they are nice, i jest not) , and develop your app with silverlight/MFC/WPF/Winforms/Access/LiquidMotion/VB/Office/CopiedTechOfTheDayAtMS, trust me it's much better the HTML ;-) .    Oh, how could i forget, you could also use Front Page coupled with IE6.

    me, just a fanboii of openness and standards

    IETeam, what's holding you from releasing a statement on WebGL?

  269. hAl says:

    @meni

    WebGL is not part of any W3C webstandards thus there is no reason to implement it.

    It is an plugin/extension on javacript based on OpenGL hardware driver architecture.

    A good 3D standard should not be based on OpenGL but should work independant of underlying architecture.

    So more likely for 3D you should think of a 3D extension of a standard like the Canvas element which could then be rendered thru OpenGL or thru Direct3D. Microsoft could then support that 3D Canvas (HTML) standard. (a 3D extension of SVG could also be an option) and use Direct3D hardware acceleration for it.

  270. meni says:

    hi hAl, thanks for responding. My request was directed to the IETeam. IETeam, either way please respond.

    "A good 3D standard should not be based on OpenGL"

    i naively thought opengl was an open, patentless standard. yes, it's different then directx. so you are saying there should be a third api? why? is there ego involved? or would you feel your world shuttered if you'll have to learn opengl, after being in so many dx vs opengl flame wars?

  271. meni says:

    @MSDN Blogs, how about a revolutionary idea: you have a form element with action attribute, when the user click "post" you submit the form using the POST method, then you see post displayed :-))))

    All this, and no AJAX too, amazing! Oh, and it will work even when the user, heavens forbid,  disables JS. Yes I know it's controversial today, but someone told me it might be all the rage a few years from now.

    (I blame asp.net for all the these unnecessarily complicated  sites, well maybe not, also YouTube has this sickness)

  272. must says:

    @ Mitch074,

    yeah, I'm living in Asia and I have built tons of websites and none of them are asking to have blinking or red/gold stuff all over the place. The only "blinking stuff" I can think of are the flash ads.

    Some examples :

    www. hao123.com

    www. china5e.com

    www. taobao.com

    www. 163.com

    those are some of the largest and most popular websites in China and none of them have much blinking/red/gold stuff except some flash ads banners.

    I don't know, maybe those who asks you to build websites are those scam sites or warez download sites, I'm not saying you are making things up, but you can't just compare some scam/warez sites in Asia to normal sites in English. You know, porn/warez sites in English are full of those blinking non-sense craps too.

    And you don't tell someone who's been living in Asia for over 30 years and making websites for Asians for more than a decade what Asians like in their websites, it's ridiculous that you try to tell Asians what Asians like and don't like here, and I'm telling you that you are clearly wrong here. Asians, just like anyone else in this world, like clean, easy-to-navigate, ads-free websites.

  273. must says:

    @ Mitch074,

    yeah, I'm living in Asia and I have built tons of websites and none of them are asking to have blinking or red/gold stuff all over the place. The only "blinking stuff" I can think of are the flash ads.

    Some examples : hao123, china5e, taobao, 163, etc. etc.

    those are some of the largest and most popular websites in China and none of them have much blinking/red/gold stuff except some flash ads banners.

    I don't know, maybe those who asks you to build websites are those scam sites or warez download sites, I'm not saying you are making things up, but you can't just compare some scam/warez sites in Asia to normal sites in English. You know, porn/warez sites in English are full of those blinking non-sense craps too.

    And you don't tell someone who's been living in Asia for over 30 years and making websites for Asians for more than a decade what Asians like in their websites, it's ridiculous that you try to tell Asians what Asians like and don't like here, and I'm telling you that you are clearly wrong here. Asians, just like anyone else in this world, like clean, easy-to-navigate, ads-free websites.

  274. must says:

    @Mitch074,

    "I'd say that in China, it's harder to find a legit system with a real Windows license than any other. Heck, even boxed laptops use bogus license numbers and come preinstalled with pirated crap."

    heh, that's because a lot of people intentionally replace the legit Vista and Win7 systems which came with the computer to Win XP. Almost every brand name computer in China, desktop or laptop, actually came with legit Windows systems like anywhere else in the world, it's just that the users almost always replace those legit Windows systems to Win XP because they think Win XP is more comfortable to use.

    And you are definitely making it up by saying "boxed laptops use bogus license numbers and come preinstalled with pirated crap", even in China almost all boxed laptops come with preinstalled legit Windows systems, it's just that too many people asks the retailer to replace the legit Vista or Win7 systems in their laptops with XP, that most end up with pirated XP.

    "IE 6's prevalence comes from pirated software"

    Not exactly pirated software, rather than IE6's prevalence comes from people's unwillingness to switch from XP. They are so unwilling to switch from XP that they even replace their legit Vista and Win7 systems that they have paid for (when buying new PCs) to XP.

    "must have a minimum of 30% of total area covered in red, gold and/or blinking stuff"

    heh, I'm living in Asia and I have built tons of websites and none of them are asking to have blinking or red/gold stuff all over the place. The only "blinking stuff" I can think of are the flash ads.

    Some examples :

    www. hao123.com

    www. china5e.com

    www. taobao.com

    www. 163.com

    those are some of the largest and most popular websites in China and none of them have much blinking/red/gold stuff except some flash ads banners.

    I don't know, maybe those who asks you to build websites are those scam sites or warez download sites, I'm not saying you are making things up, but you can't just compare some scam/warez sites in Asia to normal sites in English. You know, porn/warez sites in English are full of those blinking non-sense craps too.

    And you don't tell someone who's been living in Asia for over 30 years and making websites for Asians for more than a decade what Asians like in their websites, it's ridiculous that you try to tell Asians what Asians like and don't like here, and I'm telling you that you are clearly wrong here. Asians, just like anyone else in this world, like clean, easy-to-navigate, ads-free websites.

  275. must says:

    @mitch074,

    "eventhough I don't recommend it, because it would push away all those that use more recent/free browsers in Asia – and there are"

    Since I'm not doing IE6-only sites, but W3C compliant sites with IE6-hacks to ensure IE6 compatibility (to make the sites behave correctly in IE6 as in other browsers), and I have put download banners for alternative browsers on the home page to encourage people switch from IE (albeit with little success so far), so I don't think that'll "push away" anyone that use more recent/alternative browsers.

    "Should you double development time to support IE 6?"

    Since around 60% of the visitors here are using IE6, one has no choice but to support IE6 compatibility, even if one puts download banners for alternative browsers on the home page. Unless you want to get yourself fired and starve.

    "IE 6's prevalence comes from pirated software"

    Not exactly pirated software, rather that IE6's prevalence comes from people's unwillingness to switch from XP. They are so unwilling to switch from XP that they even replace their legit Vista and Win7 systems that they have paid for (when buying new PCs) to pirated versions of WinXP.

  276. Andy Babiec says:

    IEPP4 was getting a score similar to IE 8 (4569ms vs. 4541ms) on a dual core Win 7 64-bit 8GB laptop.

    I then went into Control Panel > Internet Settings > Advanced and checked "Disable Script Debugging (Other)"

    "Disable Script Debugging (IE)" was already checked.

    Ran the test again and got a score of 394ms

  277. EricLaw [MSFT] says:

    @Andy: When Script Debugging is enabled, the JavaScript engine disables the just-in-time compiler and runs in Interpreted Mode. It's interesting that you still didn't see a better score with IE9 than IE8, however, since the new interpreter is generally quite a bit faster than the IE8 interpreter.

  278. hAl says:

    @Meni

    A good 3D standfard should be based around 3D functionality and should be independant of 3D hardware support architecture.

    So you would design for example a 3D Canvas standard purely around the functionality that 3D designers and 3D users would need in that standard and not around the functionality that OpenGL (or directx) architecture provides.

  279. ropi says:

    Oho finnally smth good! please release it asap!