Internet Explorer 9 preview improves HTML5 compliancy

IE9HTML5

Microsoft is working hard on the next version of their browser, Internet Explorer 9, and plans to make this the browser the top dog on the block.  Early IE9 preview tests have been looking very promising for the browser, finally moving to be fully HTML5 and CSS3 compliant.

The Internet Explorer team posted an update on IE9' progress, showing that it's fully W3C Web Standards compliant so far.  Not only did the IE team post the results for IE9, but posted the test results for Mozilla Firefox 3.6.3, Opera 10.52, Apple Safari 4.05, and Google Chrome 4.1.  The tests were all performed with browsers running on Windows, shipped in May 2010.

The W3C Web Standards tests included HTML5, SVG 1.1 2nd edition, CSS3 media queries, CSS3 borders & backgrounds, CSS3 selectors, DOM level 3 core, DOM level 3 events and DOM level 2 style. Microsoft, with the help of W3C, performed a total of 192 tests on all the browsers.

Internet Explorer 9 scored 100% in all eight tests, while every other browser, except Firefox in DOM level 2 style tests, didn't score perfect in any of the test categories.  Internet Explorer 9 seems to be picking up its game, and will be a very tough competitor to all other browsers when it launches.

You can download the Internet Explorer 9 preview from here.

Thanks to Skwerl for the news tip!

Report a problem with article
Previous Story

HP announces Data Vault X310; Neowin goes hands on

Next Story

Twitter testing 'You both follow' feature

109 Comments

Commenting is disabled on this article.

Dalek said,
I'm sorry but who checks articles before they're posted?

Compliancy. Seriously?

Yes? Compliancy? What's wrong with that? It's in the dictionary...

codename.venice said,
IE 9 scores 0% in few other tests!
http://www.freeciv.net/yet-another-html5-browser-test/ (got link from slashdot)
The tests MS listed were ones they submitted to the official WC3 test case suites themselves. Yes, of course they pass them. Kinda stupid to work on and submit tests you haven't passed.

Specifically choosing pre-existing tests that IE9 hasn't passed is an even worse idea.

I'm an IE fan and user, but frankly I find these statistics to be deliberately deceptive, even for statistics.

Very disappointing.

While Acid3 is not an end all and be all, it should definitely be on the list. It is excluded from the list because IE traditionally does so very poorly against it. My chrome browser in the dev channel gets 100/100 but utterly fails to create the correct rendering it - so really it would not have hurt to include it in the list.

As for the SVG. My router displays bandwidth reports with it, and they still do not render correctly, so I have serious reason to doubt those 100% flags. Either they are wrong, or they represent an even smaller subset of real-life functionality than Acid 3 does. Either way, they are misleading.

>> "Microsoft wrote the tests, Microsoft passes the tests."

>> "W3C Web Standards"

Everybody knows that Microsoft own W3C... stupid little kid.

fenderMarky said,
>> "Microsoft wrote the tests, Microsoft passes the tests."

>> "W3C Web Standards"

Everybody knows that Microsoft own W3C... stupid little kid.

Yer what? Was that sarcasm?

Microsoft wrote the tests, Microsoft passes the tests.

Hooray for them, it has nothing to do with standards compliance.

Yay for fancy tables skewed in Microsoft's favor! They seem to be implementing and testing things no other browsers have, but when you look at the grand scheme of things, other browsers still support far more features as a whole.

Elliott said,
Yay for fancy tables skewed in Microsoft's favor! They seem to be implementing and testing things no other browsers have, but when you look at the grand scheme of things, other browsers still support far more features as a whole.

The reason MS is submitting these tests is because almost no web developers implement the full spec of HTML/CSS/SVG/etc. Another issue is that the HTML/CSS/SVG specs are designed independently of one another, meaning that some of those specs contradict one another.

FWIW, I prefer "compliance" to "compliancy". The latter just sounds odd.

Example:
Q: How does this batch of compliance taste?
A: Better, but it's still a little compliancy.

Yuck.

thealexweb said,
No Ogg Theora or WebM tho so back of the pack in that respect.

MS has said that they are more concerned with the h.264 however, they did say that they would roll out support for the others. But this probably be in the form of a plugin.

thealexweb said,
No Ogg Theora or WebM tho so back of the pack in that respect.

It'll use WebM if the codec is installed and, let's be honest almost no one is going to use Theora if WebM is available.

thealexweb said,
Why is Microsoft comparing their alpha builds against other browsers stable builds, why?

Because if they compared it to unreleased builds you'd whine that they're works in progress and not suitable for comparison. You can't have it both ways, troll! >

I personally use Chrome but am a long time user of Firefox. I havent used IE since Firefox was in its first betas. I use it at times though as IE is the best browser to be using to browse Microsoft products (no surprise). As an administrator for many company networks, IE also has much unrivalled Group Policy configuration, something every other browser lacks.

If they make the browser support extensions, then I guess, Chrome+FF will get a big kick in the butt. Then its the other side that will have to improve.

Well, lets wait and see.

I just tested the preview on a website where I have a jQuery cycle slide show on the front page and the GPU rendering really does make a huge difference. With the latest release build of FF, Chrome, Safari, and Opera, the slide show spikes one core of my CPU between 70% and 95%. However, with the preview of IE9, the CPU core spike is only ~ 5%. Offloading the rendering to the GPU is a great step forward and I hope the other browsers do so as well.

I have the latest Webkit and want to see how it compares to this. Where the hell is the test !?

PsykX said,
I have the latest Webkit and want to see how it compares to this. Where the hell is the test !?

They already compared it to Safari which uses Webkit ?

Edited by kInG aLeXo, Jun 3 2010, 6:55am :

kInG aLeXo said,

They already compared it to Safari which uses Webkit ?

Webkit is updated daily… and Safari uses an old version actually.

It's so hilarious to see people react as though they're being personally threatened when the product that they use (one that they don't even freaking pay for no less) might be bettered by another product (another one that they wouldn't even freaking have to pay for to use). Download it and test it out yourself. Or don't. Why do you feel the need to trash the benchmarks? Any moron writing a tech blog could run some test to dispute the results if these benchmarks were truly rigged. Do you think Microsoft is really that desperate to make their browser look good? Sheesh!

I remember a few days ago when someone was suggesting that MS didn't do agile development - and here we see they're clearly using TDD - may even be BDD.

It's even more clear that they haven't a clue about software development when they say "Sure it passes the tests NOW - but it's still in development!" Laughable!

That's why they make tests, and contributing them to the rest of the community is really throwing down the gauntlet. It'd be great if the Firefox, Chrome, and Safari teams all contributed so many tests. May the best browser win!

Deathray said,
The browser needs good plugin/addon support like firefox... IE NEEDS it

Aren't badly written addins what causes browsers to eat CPU cycles or crash completely? I'd rather have native support, thank you very much.

kryten said,

Aren't badly written addins what causes browsers to eat CPU cycles or crash completely? I'd rather have native support, thank you very much.

Well I don't expect to see NoScript in IE natively. Even though they tried something close to it. I'm using Firefox because of the addons I can add to make the experience match my needs

kryten said,

Aren't badly written addins what causes browsers to eat CPU cycles or crash completely? I'd rather have native support, thank you very much.


On Firefox moreso than Chrome, because Chrome has a more restrictive extension API. However, this also implies that they can do slightly less on Chrome. It's stability at the cost of freedom. But I'd take that kind of extensions over the Internet Explorer BHO's any day.

yardman said,
Keep up the good work IE team.....

Keep up ? They have just started the “good” work

I can't understand why people are so personal about their browser. It's actually OK that IE is improving, don't have to feel threaten by it. Go worry about what kind of food you gonna eat tonight, or who you gonna hang out with, etc.

I'm quite happy to see Microsoft trying their hardest to bring IE up to date - and they're doing a good job. But this is nothing more than marketing, and cheap marketing tactics at that. Of course Microsoft would pass all the test that they make. Do they support all the things they are testing better than other browsers?Quite possibly, yes. But the way they put this table together is atrociously misleading, and quite frankly insulting to people who know better.

Raa said,
Ahh, but can it pass an Acid3 test?

Who cares, check the html5test.com and prepare to be disappointed of this reporting.

html5test.com... 19 of 160. W. T. F.

Northgrove said,

Who cares, check the html5test.com and prepare to be disappointed of this reporting.

html5test.com... 19 of 160. W. T. F.


Yeah....that's what I'm getting in IE8, though. It site doesn't actually work in the latest Platform Preview I have.

Edited by lordcanti86, Jun 3 2010, 7:23am :

Raa said,
Ahh, but can it pass an Acid3 test?

Stop using Acid 3 as some kind of "be all, and end all" test. It isn't. It's tests a small subset (100) of features from the latest HTML5/CSS3 standards. The results Microsoft submitted test a much larger (192) and different set of features from HTML5, CSS3 and DOM.

So, who's test is valid? Neither. All this shows is that Microsoft passes at least 192 features of HTML5/CSS3/DOM and the others pass at least 100 features of HTML5/CSS3. Until there's a test that checks every feature in those specs, it'll be impossible to say which has better standards compliance.

At the end of the day though, as long as the sites I visit render fine in my choice of browser, does it really matter?

Thx Giga, surprise that no a single reply in the first 20 doubt the results. Which is untrue, unreal, false marketing and i lost words for it. The test is wrriten, and done by IE Team only.

RGBA colors work. I'm trying to get box shadow to work but I couldn't figure it out (if it's in there at all). I tried leaving it un-prefixed and I also tried using the prefix -ie (that was just a shot in the dark, I'm not sure if there is some different prefix IE uses or if they even use prefixes for unfinished features at all).

geoken said,
RGBA colors work. I'm trying to get box shadow to work but I couldn't figure it out (if it's in there at all). I tried leaving it un-prefixed and I also tried using the prefix -ie (that was just a shot in the dark, I'm not sure if there is some different prefix IE uses or if they even use prefixes for unfinished features at all).

IE uses the -ms prefix.

This is great and all but since IE9 is a way off from release, anything could happen. Google/Apple/Opera/Mozilla etc aren't just going to wait twiddling their thumbs for IE9 to release, they'll be working hard to make their respective browsers better. What this will do though is put the pressure on the others to work hard keep their users.

I really do hope that IE9 makes out to be a good "THE" browser. For the sake of all the uninformed Windows users out there.

Recon415 said,
I really do hope that IE9 makes out to be a good "THE" browser. For the sake of all the uninformed Windows users out there.

Just because people use IE doesn't mean they're all uniformed. I use IE8 primarily, even knowing the other browsers I can use.

It looks like it does. I just tried the preview and went to neowin and it had rounded corners. However it crashed when I went to login.

I'm glad Microsoft put people who actually care about standards into the IE team. I'd love to be able to finally recommend IE among the browsers a user can have, and even use it in lieu of Firefox if the IE team can prove their commitment to the HTML 5 spec.

billyea said,
I'm glad Microsoft put people who actually care about standards into the IE team. I'd love to be able to finally recommend IE among the browsers a user can have, and even use it in lieu of Firefox if the IE team can prove their commitment to the HTML 5 spec.

Microsoft's IE people have always had some of the better standards people (even the authors of XHTML etc). It is the prediction of what standards get accepted and keeping rendering compatible with the IE 4.0 general of handling web content when they were defining many of the standards and handling 'errors' better than Netscape.

IE fails things like ACID 3 easily because it always has added 'intelligence' to the rendering, so if a ending TABLE tag was not present, it would display the table anyway, where Netscape would show a blank page. Back in the 4.0 days when people were using dial up and had flaky connections, this 'intelligence' of the expect page rendering was a better solution than giving users a blank page, and is the main reason IE 4 became the leader in browsers of the time.

Even today, the network quality/connection is something IE manages better than other browsers. If you work from a tethered phone or crappy connection, IE is usually the one that obtains the page and content better than Firefox or Chrome, and again is why some people prefer IE over the other browsers since they don't have super fast solid connections.

IE9 was a long time coming, and IE7 and IE8 were good shoves for developers to get them away from IE rendering modes.

PS Standards are a tricky thing, because just as in the IE4 days, not everything is approved when the browsers ship and this creates a lot of conflicts in rendering. Even today the standards with CSS and how SVG are handled work very differently in Chrome and Firefox becasue some things are not finalized and each browser team handles them differently.

Actually, the ACID tests test each browser not only on how they handle standards based code, but also how they handle malformed code.

thenetavenger said,

Microsoft's IE people have always had some of the better standards people (even the authors of XHTML etc). It is the prediction of what standards get accepted and keeping rendering compatible with the IE 4.0 general of handling web content when they were defining many of the standards and handling 'errors' better than Netscape.

IE fails things like ACID 3 easily because it always has added 'intelligence' to the rendering, so if a ending TABLE tag was not present, it would display the table anyway, where Netscape would show a blank page. Back in the 4.0 days when people were using dial up and had flaky connections, this 'intelligence' of the expect page rendering was a better solution than giving users a blank page, and is the main reason IE 4 became the leader in browsers of the time.

Even today, the network quality/connection is something IE manages better than other browsers. If you work from a tethered phone or crappy connection, IE is usually the one that obtains the page and content better than Firefox or Chrome, and again is why some people prefer IE over the other browsers since they don't have super fast solid connections.

IE9 was a long time coming, and IE7 and IE8 were good shoves for developers to get them away from IE rendering modes.

PS Standards are a tricky thing, because just as in the IE4 days, not everything is approved when the browsers ship and this creates a lot of conflicts in rendering. Even today the standards with CSS and how SVG are handled work very differently in Chrome and Firefox becasue some things are not finalized and each browser team handles them differently.

+1 Thanks. This explained a lot to a novice like me.

Benjamin Rubenstein said,
Does this mean Neowin's nav bar will have rounded corners in IE?!?! [excited for those who, for some off reason, still use IE]

its a browser, theres no "off reason" to use one over the other. Use what you use and let us use what we use.

Benjamin Rubenstein said,
Does this mean Neowin's nav bar will have rounded corners in IE?!?! [excited for those who, for some off reason, still use IE]

A Red Name asking this? Really? You know you can download the test preview yourself.

Oh, and BTW, the rounding corners and how even style lines are dealt with is something they showcase in the preview site, as there are still problems with consistency between Webkit, Firefox, etc on how they handle line styles and corner rounding.

If you follow the developer blogs, MS are the ones asking the questions to get the rounded corners handling standardized between all browsers as the specifications don't clearly state how they should be handled. This is one of 30+ specification problems that are not clearly defined they are trying to get everyone to agree on.

PS Neowin does look better with rounded corners, so that I agree with you on, and am glad to see IE handle them properly in IE9.

Benjamin Rubenstein said,
Does this mean Neowin's nav bar will have rounded corners in IE?!?! [excited for those who, for some off reason, still use IE]

Well there's bias if I ever saw it

Jokes aside, IE has made a lot of improvements in recent years and is an excellent choice of browser. I'm hoping Chrome will give it a run for it's money!

Benjamin Rubenstein said,
Does this mean Neowin's nav bar will have rounded corners in IE?!?! [excited for those who, for some off reason, still use IE]

don't care about rounded corners

security (thanks to IE sandbox which also sandbox risky plugins like flash and adobe reader), autonomy ( http://www.anandtech.com/show/2834/3 ) and stability is way more important than rounded corners for people who use their computer to work

dagamer34 said,
Does this include stuff like geolocation and canvas?

There is an entire list of everything they tested inside the first link. It's also separated into different categories.

Why didn't they post the latest google chrome results? The stable version is 5.0... Where do I have to visit to run these results on my browser?

Evolution said,
Why didn't they post the latest google chrome results? The stable version is 5.0... Where do I have to visit to run these results on my browser?

Chances are these results were compiled before 5.0 was released to the stable channel.

Evolution said,
Why didn't they post the latest google chrome results? The stable version is 5.0... Where do I have to visit to run these results on my browser?

Because Google recently played dirty by giving up on using Windows for their work computers

Evolution said,
Why didn't they post the latest google chrome results? The stable version is 5.0... Where do I have to visit to run these results on my browser?

I personally would like to see Google Chrome 6.0 go up against IE9, but as it's still a beta, Microsoft is only testing release builds.

Andrew Lyle said,

I personally would like to see Google Chrome 6.0 go up against IE9, but as it's still a beta, Microsoft is only testing release builds.

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.

DrCheese said,

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.

People are going to complain that it's not a public build so all the kinks haven't been ironed out.

DrCheese said,

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.

Very true. Good point tho.

Andrew Lyle said,

I personally would like to see Google Chrome 6.0 go up against IE9, but as it's still a beta, Microsoft is only testing release builds.

I found in Chromium 6.0.423.0 (48639) the same results as Chrome got.

DrCheese said,

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.

Not really. Then there is people who would say "Its still in beta/dev thats why it scorred lower that IE9"

Its a perfectly valid test. 1 Beta product outpreformed all the lattest builds when most fo the general users would be using. Only people who know what their doing use beta/nightly versions of software,

Evolution said,
Why didn't they post the latest google chrome results? The stable version is 5.0... Where do I have to visit to run these results on my browser?

Version 6.0 (Currently Beta and Dev) renders all of these things over 80% (Except for DOM Level 2 style, I have no idea what that is). Plus these results are probably fake, Microsoft does it all the time (Every major company does even if they say they done). Also the test where probably made for IE 9 to do well on them, they probably recognize that its IE 9 and purposely makes the score better. While the other scores are based on how well those browsers did at mimicking IE, but they can never do it because none of the other browsers above are crap, IE is crap. How can a browser created to be amazing mimic crap? Also IE 9's Acid 3 Test that MS is showing off is terrible. If you just look at the score you can tell that those results in the chart above are fake. Chrome 4.1 scores 100/100 and IE ~64/100 so how does it get lesser scores? Same with peacemaker

Aaron7pm said,
...
1. Learn what the Acid3 test actually is.
2. Show how the results listed are fake, or how the tests don't follow the spec.
3. Actually try the IE9 preview.

Edited by Kirkburn, Jun 3 2010, 2:28am :

DrCheese said,

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.
It's not about being fair.


If they get the latest build it could result in huge difference depending on what night they chose to get the nightly so that may not even be fair anyway. The latest build for any of the products could be awful depending on where in the release cycle the dev teams at. Anyway, they listed version numbers so they were up front about what was being tested.

Edited by Smigit, Jun 3 2010, 2:41am :

DrCheese said,

which is cheating somewhat given that IE9 isn't a release build. To make this fair they would have to test the latest dev build of each browser.

It's not the IE dev teams job to do so. They're just highlighting their work. Leave it to a tech site if they want to get into comparing dev builds, etc.

Aaron7pm said,

Plus these results are probably fake, Microsoft does it all the time (Every major company does even if they say they done).

how could these results be fake since you can reproduce this at home and get the same result?
plus, these tests have been approved by the W3C to be included in the HTML5/css3 test suite.

Also the test where probably made for IE 9 to do well on them, they probably recognize that its IE 9 and purposely makes the score better.

lol XD
then check the source code of these tests, it's just html, css and javascript.
there is NO conditional script which recognize it is IE9.
these tests represent real features of html and css.
they just designed IE9 to be compliant with the parts of the standards tested within these tests.

While the other scores are based on how well those browsers did at mimicking IE, but they can never do it because none of the other browsers above are crap, IE is crap.

what you're saying IS crap!


Also IE 9's Acid 3 Test that MS is showing off is terrible. If you just look at the score you can tell that those results in the chart above are fake. Chrome 4.1 scores 100/100 and IE ~64/100 so how does it get lesser scores? Same with peacemaker

Acid3 is testing 100 features among the tens of thousands features included in HTML5/CSS3

the test suite submited by MS is testing 192 features among the tens of thousands features included in HTML5, CSS3 and the DOM

neither the microsoft test suite nor Acid3 contains enough tests to test every feature of these standards (it would need tens of thousands test to cover every aspect of these standards)

however, the microsoft test suite is testing twice as much features as Acid 3. That makes it slightly more representative of the quality of the web standards implementation among non IE browsers than Acid3.

IE9 currently doesn't have 100/100 on Acid3 since it doesn't implement yet all the features tested by Acid3. However, IE team said that they plan to reach that level when IE9 will be released.

Anyway, it's interesting to see how demagogic the Acid3 test was:
every browser company have fixed the bugs that needed to be fixed to pass acid3 without fixing the thousands of other untested/unfixed bugs (which is harder), and then claiming to be the best standard compliant browser!