Blog Post

Dean Hachamovitch: IE9 - Questions and Answers

Play Dean Hachamovitch: IE9 - Questions and Answers
Sign in to queue




Not too long ago, a thread in the Coffeehouse was posted asking for your IE9 questions for Dean Hachamovitch, VP of Engineering for Internet Explorer. Thank you for asking such great questions, Niners! We didn't have time cover all of them, but we did manage to get through several of them (and there was some redundancy, too, so I took the liberty of choosing the questions. Also, I didn't ask any of the snarkier questions (though I probably could have, since Dean can handle it) and Dean did not want to see the questions before they were asked (which is commonplace among many executives). He's OldSchool9 that way and that's great!).

IE9 PP7 was released today and inside of it is one of the world's fastest JS engines (according to the SunSpider suite of JS performance tests, IE9 PP7 is faster than all other browsers out there...). Chakra, the IE9 JS engine, continues to evolve! Hats off to the Chakra team. Interestingly, Dean isn't overly excited about this news (though, obviously, he's thrilled) since JS execution speed is but one part of the overall performance story, which is actually divided into several pieces, not just two or three (see Jason Weber's quick chat on the overall performance characteristics that combine to form a user's experience of blazing fast web surfing, and read the IE team's blog post on the subject).

Here are the unedited questions Dean answered in this morning's conversation in the IE engineering building. Thank you, Dean, for taking the time out of your busy schedule to engage Niners' questions and concerns in an honest and open way. There is bonus footage at the end, too Smiley

I recommend watching all of this as the flow is nice and the questions lead into one another nicely. Great job, Niners! This is a 51 minute interview, which is a long time in VP Time. Dean's a big fan of yours, Niners.

[Questions begin at 00:14:15, asked in the order listed below. Click on a question to go directly to the answer in the video. I do recommend that you watch/listen to the entire conversation as it flows naturally from question to question with many questions and answers nicely building on those that come before and after...]


How you do personally feel about the way the HTML 5 standard is progressing? It's talking such a long time to form, it's slowing down the evolution of the uniform internet and causing more and more fragmentation between browsers. Is it not?


What if the HTML5 standard changes/is finalized within IE9's lifetime and before IE10 is out. What's going to happen? A patch?


Windows 7 had a tremendous amount of telemetry captured from the beta users and that information was immensely valuable. Was or is there any telemetry data collected from current or past testers?


How do you measure Internet Explorer stability and when it's ready for RTM? Does the beta have any kind of telemetry for latencies: when the GUI freezes, when a page takes a long time to load, etc.
How many test cases do you have for IE9 at this point?


Will we be able to write javascript or .net plugins for IE? We need an easier way to interact deeply with the browser but writing native plugins is too difficult and accelerators can't go deep enough.


Are you anxious to see if Mozilla actually sends in a cake to the IE team when IE9 ships?

Do you plan to have IE score a 100 in the acid3 test?


Will platform previews continue after RTW release so we can see where you are heading for the next release? You seem to be about a 6 week or so window for the platform previews currently.

Any possibility of smaller update window between releases? eg will we be likely to see IE 9.1 or 9.5 that is smaller in scope but quicker to market?

US Archer:

What improvements have been made around touch support?  Is the UI fully baked in this Beta?

What is the state of webslices, something you still encourage site developers to implement?  For IE9, it would be really useful if we could preview these from the pinned site icon.


Does the IE team still feel like they're playing catch up or are they now at a point where future versions of IE might start to drive new ideas once again?



Download this episode

The Discussion

  • User profile image

    Related to the touch question:

    It's very difficult to navigate the Print, File, Zoom and Safety menus in IE9. At least for me the menu items are too small, even if they get enlarged for touch optimization. I'd prefer icons and in total a Ribbon-like (maybe optional) menu.

    For the page view: It's not as smooth to zoom in as it is like in Safari where zooming is a smooth animation.

  • User profile image

    Terrible backpaddling from Dean in this video.

    Ever since IE9 was announced Dean and the IE team has been talking standards compliance. It has all been "standards, standards, standards!" and that the time of writing browser specific code is coming to an end.

    Now in this video when asked about what will happen if HTML5 is finalized during IE9's lifetime he's suddenly essentially saying that "IE9 as a platform > standards". I thought that kind of IE6 thinking was a thing of the past.

    Dean talks at length about not having us updating our websites everytime the standards change. What he's promoting is having us deal with browserquirks yet again. *sigh*

  • User profile image

    @sialivi: Not following your logic here... Dean made it clear that if standards do change during the lifetime of an existing browser that implements the current standards, well, you don't just break things with a patch, assuming breaking changes in the standard. In fact, no browser maker would do this. Only security, reliablitlity and stability requirements could demand shipping of breaking changes, as he stressed. Finally, HTML5 is the platform. IE9 just executes it on Windows (and it does so as close to the underlying hardware as possible, with a best in class JS engine to compile and execute JavaScript and all renderable HTML+CSS computed on the gpu...).

    Your IE6 comment is baseless. Can you elaborate?


  • User profile image

    @Charles: So lets make the ridiculous assumption that HTML5 is finalized a week after IE9 RTM and contain severe breaking changes, probably a year or two before IE10 will be done. We are then to continue making our new websites compatible with IE9's then non standard compliant implementation of HTML5 for months or years, alongside browser specific versions for other browsers. We will once again, to save us the hassle of updating our old sites, be stuck designing our future websites around browserquirks instead of standards.

    I assume your HTML5 is the platform comment was a response to the "IE9 as the platform" part of my previous comment. Yes, HTML5 is the platform. It should be the platform, that was my whole point. It was Dean who referred to IE9 as the the video. IE9 shouldn't be the platform and the fact that he sees it as such is troubling.

  • User profile image

    Secret Key mentioned is E

  • User profile image

    sialivi here is an idea: don't use a feature before it is a standard and all targeted browsers support it.

  • User profile image

    @Stilgar: You're missing the point.

  • User profile image

    For people who want to know Gorgio Sardo's point of view:

    We interviewed him at TechEd 2010 in Berlin.

    Great Video Charles! Great content, and thanks for asking my question!



  • User profile image

    Nice interview and I'm pleased with how IE9 is going. 

    Bug with touch (or pen at least) which has existed since IE8:

    Pen flicks not working to scroll the page, as a Tablet PC user this has driven me insane for too long. Smiley  During the IE8/Windows 7 beta it was closed as postponed, if you could squeeze it in it would be welcome!

  • User profile image

    Thanks for the answers Smiley

    I don't see anything controversial in what Dean says. He just says, as I get it, that new versions get more compatible but the team will not go back and provide updates to old browsers that break compatibility except if they have very good reasons for it - such as security holes.

    That's very sensible and understandable. It's like the realm of functional programming: you don't do destructive updates and redefine what IE6 is, instead you create IE7, IE8, IE9, etc.

    Then there's a challenge getting people to update their browsers but that's really in peoples own hands.

    Godspeed with the progress!

  • User profile image

    Considering that the HTML working group is dropping the Version aspects. after the full ratification of the spec known as HTML 5, and it will just be known as HTML at that point.  and given the Editor of HTML 5 is expecting final ratification of html 5 sometime around 2020. 

    Please explain your point sialivi.

    Most of the items from the HTML 5 spec that have been implemented are at least in the stable state and already implemented (sometimes poorly) in other browsers. (some areas IE still needs help) 

    The HTML working group is already to start splitting the spec much the way that CSS 3 has split. after the base of HTML 5 is complete. 




  • User profile image
    Dean H [MSFT]

    @andreasbalzer: ah - I understand the feedback, thanks. Will get it to the right folks.

  • User profile image

    @Charles/Dean, thank you for making this Q&A video.  

    Dean's response to ZippyV's question regarding plugins was insightful.  My personal opinion is that the traditional plugin model (for any browser) is a dead end.

    I should have been more specific with my question regarding touch.  IE9's chrome was the intended target.  So thanks Dean, you did the best he could with a open question like that.  If I had to reframe the question, it would be around usability testing and ui design philosophy.  I could be very wrong, but I don't think the IE9 team does enough testing when it comes to touch.  Now many of the same issues (ex. context menus, scaling) affect the rest of Windows as well.  I just thought with 60% of PC usage centered around web browsing and the coming wave of slates/tablets that IE would be the precursor (similiar to Office Ribbon or Zune UI) of what eventually will need to be done with Windows v.Next and touch.

    The question about webslices rose from the removal (hidden) Feeds icon in the default IE9 Beta chrome.  That suggested to me that Microsoft was backing away from this.  Sure you can move your mouse over content to see if there are webslices to be had.  Though thats not the way I would want to discover content.  It will diminish webslice uptake.  Maybe there is a better way, so I was glad to see Dean churn for a brief moment on the idea of previewing webslices from a pinned site.  I actually think pinned sites are the way to go long term.  Possibly leading to removal of the traditional favorites bar, etc... Instead let the act of pinning a site subscribe the user to related RSS feeds and webslices in one shot.  To do this, pinned site jumplists would need a way to interface with the IE Feed Store/Sync.  An additional user benefit, users wouldn't need a full brower session just to preview updated webslice/rss...providing an enhanced app like experience.  I don't pretend to have thought this through (your job IE team) but I believe there are more ways to leverage pinned sites.

    Re: Bookmarklets:  RoboForm is one of the better ones out there.  Personally, I'd get more out of having accellerators listed in an easy to access icon (on the default IE chrome) instead of being accessed through a context popup menu.

    Thanks again!!

  • User profile image
    Russell Leggett

    This was a great discussion, and I was happy to get some really candid answers. (The more open the better) The only issue which really frustrated me was the statement that IE will not get any feature updates until IE10 after IE9 ships. I actually completely understand and agree that breaking changes should not be made (other than the security type ones you mentioned) until the next major version number - that's just standard practice - but I don't understand why additional features can't be added. There are still plenty of goodies that are available in other browsers but not yet in IE9 - a lot of the newest CSS3 stuff for example. Or WebGL, or App Cache, or Web Workers, or IndexDB, or Web Sockets, etc. I get that it can't all be in the initial release, but I don't see why it can't added 6 months down the road. Every other browser has Web Workers and Web Sockets right now - shouldn't that be enough reason to add the support, even if it has to be in an IE9.1?
    I'll go ahead and say that IE9 is currently setting the bar for hardware accelerated graphics, the other browser may need to take the entire next rev to catch up, but they all do backwards compatible minor version updates on a pretty healthy cycle, and the updates are usually automatic by default too. If you want to prevent another IE6 disaster, that is how you do it. Developers willing to use bleeding edge stuff in HTML5 know how to do feature detection - that's the only way to do it right now, and tools like Modernizr can really help with this. I think we as developers can handle it, and doing an autoupdate every 6 months should hurt the average joe either - you clearly said that you'll do it for security issues. Nobody should be developing for IE only anymore - that is the point of HTML5 and standards. The web has grown up, and a lot of developer resources are going towards making progressive enhancement feasible across browsers, but unless there's at least a hope of getting to progressively enhance important features for all modern browsers, the features just won't get used as much, and that slows everyone down across the board.

  • User profile image

    Excellent video, really interesting to hear from Dean again. Not sure if I answered my own question or if Charles did, but good answer all the same. Now off to download PP7 and give it a try!

  • User profile image

    @AndyC: I guess I did, but so did Dean! Smiley Great question, Andy.


  • User profile image
    Ken Jackson

    I want little Yoda!!!

  • User profile image

    Very cool, they chose my question, thank you.  I like this interview format, questions from real users and not having the interviewee see the questions before hand.

  • User profile image

    "Thank you, Dean, for taking the time out of your busy schedule to engage Niners' questions and concerns in an honest and open way.":

    Here's a couple concerning questions I didn't hear asked:

    1) Is the proven fuzzy text a concern that you or Direct2D team is looking into?

    2) What about the concerns for less room for tabs and address bar than previously?

    3) Google maps may run faster but the actual use behaviour seems worse than in IE8 even in  Preview 7. Could be a problem in their code but it also points out that IE9 can make existing stuff behave worse. I tried all the 7,8,9 doc modes and dragging the map is jerking it visually around much more than in IE8 -> user for sure will think IE9 is slower than IE8!

    Just in case there's a separate interview for the left out snarkier questions, here's one more.

    1) Both above issues 1&2 could be addressed by upgrading to larger, widescreen, high DPI displays. Do you have plans to capitalize on this upgrade & disposal process? eg. Atleast by getting some good deals for new displays and disposal services for the old ones from China?


  • User profile image
    Deactivated User

    Comment removed at user's request.

  • User profile image

    Really informative. Thanks guys!

    Do you have a link to that blogpost Dean mentions in response to my question? Maybe I missed it in the video description?

    Also, it's Holland. But close enough. Tongue Out

  • User profile image

    @Bas: Sorry about getting your country of residence wrong, man Smiley

    In terms of the blog post Dean referenced, it's the one on vendor prefixes:


  • User profile image

    All questions listed in the description are now jump-links directly into the point in time when the question was asked during the conversation. You should watch the whole thing, however, as it's a conversation and not just a question and answer pony show. Smiley Still, I love the new feature. Thanks Duncan and team!


  • User profile image

    @androidi: It wasn't just snarkiness, Androidi, that determined the cut... I had to limit the number of absolute questions so that we could end in less than an hour (Dean is incredibly busy, as you can imagine). The questions chosen made for a conversation that flowed nicely (at least in my mind).

    That said, maybe we can get the IE team to answer some of the questions that didn't make it into the filmed conversation.


  • User profile image

    @Charles: right, I couldn't quite make out what Dean was saying there. Vendor prefixes. Thanks, I'll check it out.

  • User profile image

    What about x64 bit support? Browsers normally runs in x86 because the plugins for the (Silverlight, Flash, Java, etc.) are mostly x86. Are you going to make x86 plugins run in some sort of isolated environment so the browser can take advantage of the x64 bit support?

  • User profile image

    I would like to have an adblock plugin. The one I'm using now with IE8 (simple adblock) doesn't work in IE9. Can I create an adblock plugin using a "bookmarklet"?

  • User profile image

    I have a few bookmarklets I have pasted into a few pastebins that you can all look at as examples of how I use them, which is great cross-browser functionality which I keep synchronised among Opera 11, Chrome Dev, IE9 Beta and Firefox 4.0b8 nightlies.

    Enjoy, as you ca just paste in a line starting with "javascript:" into your URL bar and try one out to see what It does.

    PS Your spam filter sucks thinking my links are spam... Sad



  • User profile image
    Dan Wellman

    @Stilgar: if no one uses standards before they mature and are adopted, how are those standards supposed to mature and be adopted? You can't just say 'don't use a feature until it becomes a standard' and restrict developers like that because the innovation in this industry will simply die.
    And what about all of the 'features' that IE/MS implemented in their browsers? If we follow your advice there would be no VML, because that itsn't a standard. Aside from the fact that VML is broken in IE8, it is a useful, non-standard feature that can and has been used to great success.

Add Your 2 Cents