MIX – Microsoft, W3C and SVG

In addition to our engagement in the SVG Working Group, Doug Schepers, W3C Team Contact for the SVG Working Group, and I are going to be presenting SVG: The Past, Present and Future of Vector Graphics for the Web, at MIX 2010 in Las Vegas this coming March to share help developers understand where SVG is headed.  At Microsoft we have been investigating how SVG can deliver graphics for the next generation of the Web Development. Its inclusion in HTML5 promises many opportunities for developers to enhance their sites. We will provide an overview of SVG and how the standard is evolving to support a broader range of applications on the Web.

Patrick Dengler
Senior Program Manager

Comments (35)

  1. Anonymous says:

    It must be hard working on the IE team. Everyone hates them!

    Workers wants to write standard compliance engine, then management wants a monopoly, workers produce a monopoly engine. Piss everyone off and take years of crap!

  2. Anonymous says:

    Thanks for these articles, I enjoyed them!

  3. anon says:

    Customers have been enhancing their sites for years with SVG despite IE’s constant ignorance of SVG.

    Way to catch up to 3 years ago!

  4. mogden says:

    If this means we get SVG in IE9, that would be great news!

  5. OmariO says:

    Implementation on the top of DirectDraw must make it really performant

  6. I know marketing requires making announcements all at once for the strongest effect, but I do hope you include SVG in your IE9 plans, and not just the HTML5 requirements (which is only for parsing).

  7. I know marketing requires making announcements all at once for the strongest effect, but I do hope you include SVG in your IE9 plans, and not just the HTML5 requirements (which is only for parsing).

  8. hito says:

    >Its inclusion in HTML5 promises many opportunities for developers to enhance their sites.

    We already knows this.

  9. one word please says:

    One word answer please,…

    Will IE9 support SVG natively?

    Yes or No?

  10. onh says:

    Keep us up to date and keep up the good work

  11. GreLI says:

    Hope this will real standard implemention not just poor Micrososft point of view such as CSS2.1 in IE8 (failing tests on http://www.w3.org/Style/CSS/Test/CSS2.1/20100127/ )

  12. Great! 🙂 As a big SVG fan (and of standards, in general) I’m always happy whenever promising things like this evolve. 🙂

    Nevertheless, as already stated [1] [2], I’d still invite the team to work together and produce some sort of technology roadmap for the upcoming IE version(s), so the IE/SVG community can also (potentially) reposition based on the new data (same for CSS, HTML5, etc.). Again [2], I’m aware that probably this sort of information will not be disclosed due to strategic reasons but asking doesn’t hurt. 😉



    [1] http://blogs.msdn.com/ie/archive/2010/01/05/microsoft-joins-w3c-svg-working-group.aspx#9945088

    [2] http://blogs.msdn.com/ie/archive/2010/02/01/w3c-svg-working-group-update-for-january-2010.aspx#9957470

  13. hAl says:


    Hown many of those 7000+ W3C conformance tests on CSS 2.1 is IE8 failing ? Can you give some examples ?

    Can you name a resource that proves that other browsers have better css 2.1 conformance ?

  14. hAl says:

    Will Microsoft addres possible missing elements in SVG compared to the featureset of VML ?  

    Will Microsoft provide tooling to convert VML to SVG ? (http://svg-vml.net/)

    How will the W3C workgroups adress the issues with current SVG implementations not being fully identical ?

  15. blah says:

    Hey Microsoft, you forgot to uncomment your SVG code in IE7.

  16. I’m looking forward to mid-March…hm, not sure how to explain it; I suppose you could call it intuition. 😉

  17. Rob says:

    The only people who need to know where SVG is headed is Microsoft. You’re already years and years behind everyone else. Why are you speaking at the conference?

  18. Matthew David says:

    Which version of SVG will IE9 be implementing? Will IE9 also support CANVAS, the kissing cousin to SVG?

  19. Gyrobo says:

    A little off-topic, but there’s a discussion going on about Trident’s versioning scheme on Wikipedia’s layout engine comparison pages. We know that IE8 uses Trident 4.0 [1], but prior versions of Internet Explorer had no official Trident version number.

    Could you please release a statement of some kind chronicling Trident’s version history? I’m sure it would be a swashbuckling adventure.

    [1] http://blogs.msdn.com/ie/archive/2009/01/09/the-internet-explorer-8-user-agent-string-updated-edition.aspx

  20. carlos says:

    Forget about conferences. Just replace Trident with a decent engine: webkit, gecko and you will catch up SVG.

    There is no need to talk here, you should write *code*, this is what makes computers work.

  21. Saitir says:


    If you think that a senior program manager spends much, if any, time writing code then you really have no idea what makes computers work.

    You sound like the guys who were wailing about wasted time when they announced the new IE logo back in IE7 “Dude don’t waste time designing logos, write code!”  Because as we all know, in organisations like Microsoft, the coders do all the graphic design…

  22. Luke Price says:

    Overdue, yes. But better late than never.

  23. Jon says:

    Those of you asking for confirmation of SVG support before Mix are wasting your time. Dean Hachamovitch is giving a keynote on IE9, and he needs new things to announce and demonstrate. People pay $1400 to go to Mix on the understanding that it’ll be the first place to hear and see big news about Microsoft’s web ecosystem. I think Microsoft have done everything that can short of actually making an announcement that they’ll support SVG though. If I were to speculate, I reckon Dean’s keynote will focus on a few big things:

    – SVG support

    – Canvas support

    – Some CSS 3 support (rounded corners, etc)

    – Performance of JS and Direct 3D rendering

    If they can, I bet he’d absolutely love to top it off with a demo of IE9 passing Acid 3, but who knows if they’ve improved things that far? And finally I reckon they’ll release a CTP, maybe even a beta, of IE9.

    And yes, I’m aware I’m being optimistic here!

  24. Atli says:

    Congrats on finally catching up on SVG support! I just hope you conform with the standards. (My code is already bloated enough with IE CSS and JS fixes. I don’t need SVG fixes as well.)

    I only hope IE9 (which I assume will support SVG?) gets adopted faster than it’s predecessors, or we will still have to wait like a decade before we can actually start using this :/

    Not that I want to sound discouraging or anything. I think it’s great to see IE working with the W3C! – Finally a small glimmer of light at the end of the tunnel 🙂

  25. carlos says:


    Like it or not, you have to code to make SVG work in a program, in this case IE.

    if you go to conferences, someone has to do the hard work: coding.

    There is no need to reinvent wheels here. Just implement what the SVG *current* standard says in IE9 and ship it.  No brainer, isn’t it?

  26. This is awesome, I’ve always avoided using SVG because of it’s wonky compatibility in browsers.

    One question though, will IE9 support inline SVG? Because at this point that’s what really matters with HTML5.

  27. Paul McKeown says:


    I imagine that MS will implement SVG and some (fairly large hopefully) parts of CSS3 as you suggest; I’m not sure about canvas, though, as they seem to be saying that "hmtl 5 is not standardised yet".  I would suggest, though, that MS will finally implement xhmtl, just to provide full support for svg, including namespaces.  I think you will also see jscript support inching closer to more recent ecmascript/javascript versions (huzzah!)

    Can’t get everything, but on that basis I think IE will be welcome back amongst modern browsers.

    It’s funny, but the last few months or so, non-technical friends have asked me from time to time what browser is best.  Hand on heart, I just cannot recommend IE at the moment (although IE 8 is admittedly a vast improvement on those ugly dogs, IE 6 and IE 7).  I think if IE 9 manages to carry out the prospectus above, then I would be able to say that IE 9 is "fair enough, not really the very best, but the differences won’t be obvious for non-technical use".


  28. Jon says:


    Not so long ago, Microsoft were in the W3C mailing lists offering to take over editorship of the 2D Context spec, if it were to be split out from HTML 5. Whilst conspiracy theorists leapt on this as an attempt to gain control and stifle Canvas, I preferred to take it as a more encouraging sign that they were at least considering an implementation. I agree though that it’s the least likely of the above to feature in IE9, but hope springs eternal…

  29. Cartelett says:

    If a VML file size can be untill 10% of the size of the SVG equivalent.

    If a VML graphic display can be 10 times quicker than the same SVG.

    How to accept SVG?

    Is W3C really serious?


  30. Kirby Files says:


    "If a VML graphic display can be 10 times quicker than the same SVG.

    How to accept SVG?"

    Are you really trying to convince anyone that IE’s render speed for VML is faster (by a factor of 10, no less) than SVG render speed in Webkit or Firefox? Real world performance in recent browsers says otherwise.

    Sure, SVG can be a little larger (though the examples on http://www.svg-vml.net rely on poorly written SVG vs well-written VML (I mean really, wrapping a g tag around each path of the SVG tiger to apply the same style, and no use elements?).

    However, IE’s terrible DOM insertion and VML rendering time are what make cross-browser vector graphics so vexing. By comparison, SVG and XAML rendering are usually 2-4 times faster, depending upon browser versions. And with IE8, VML rendering is so slow, it’s close to useless.

  31. Cartelett says:

    @Kirby Files

    Yes, I am trying to convince that VML DOM is really well made.

    But I shall not repeat what is notified on our Website http://www.web-vector-image.net/

    VML only runs on IE. But this browser is not a technical advantage for VML. See IE9 and its performances next month.

    But small sizes and high speed are especially the result of many elements and sub-elements you don’t find with SVG.

    Among them: VML purpose a third gradient type called "gradientTitle". You need three times more scripts to get the same effect with SVG.

    The 3D extrusion allows light effects and the depth developement spares quantity of shapes.

    These are two main reasons of VML superiority you can see on the last comparisons of svg-vml.net (particularely about the phone drawings)

    Here is another example using these 2 properties:


    In spite of its svgz compression. VML is always superior to SVG.

    PS: SVG and VML tigers have strictly similar conditions scripts and VML version displays largely quicker on IE than SVG one on FF.

  32. Cartelett,

    From what I can tell, other than the third gradient type, VML saves its bytes because it integrates with HTML4 (not requiring quotes around its attributes or closing tags).  Is that true?

    If so, then you  might be interested to know that, because SVG will be allowed in HTML5, that you will one day be able to write SVG in a similar manner (in fact Firefox 3.7 nightlies ship with a HTML5+SVG parser built-in but disabled by default).

    The chief thing that’s wrong with VML is the following:  No web browsers are rallying behind it.  Not even Microsoft.  

    I’m sorry that you’ve invested so much time in the wrong format, but for 2D vector graphics format, SVG is the future.  Why not spend your time making suggestions (like new gradient types) to the SVG Working Group?

  33. Cartelett says:

    @Jeff Schiller

    "VML saves its bytes because it integrates with HTML4":

    of course but partially!

    "not requiring quotes around its attributes or closing tags":

    this is a detail!

    "because SVG will be allowed in HTML5":

    This could be very interresting indeed!

    "No web browsers are rallying behind it.  Not even Microsoft":

    Waiting for others!

    "I’m sorry that you’ve invested so much time …":

    Time is never lost!

    "Why not spend your time making suggestions (like new gradient types) to the SVG Working Group?":

    This blog is making for!

    …But all this already existed since 1998 with VML. Finally…SVG is progressively looking for to be like VML (15 years have been lost).

Skip to main content