IE9 Outperforms Other Browsers for HTML5 Compliance | Hacker News Reader

IE9 Outperforms Other Browsers for HTML5 Compliance

readwriteweb.com

31 pointsGarbage39 comments

Open articleView on HN

IE team has been donating test cases to W3C, and of course they have 100% compliance with test cases they've published.

Historically nobody wanted to write test cases. I think Microsoft/W3C now discovered incentive to do so…

Well, that should give other teams an incentive to write test cases, shouldn't it?

Unless someone is saying that the MS-submitted ones are incorrect, I fully endorse their submission.

Many of them were incorrect when first published (by Microsoft themselves) but I assume that's been fixed.

Source?

http://www.infoq.com/news/2010/06/HTML5-Compliance-Dispute

quote: "While we are glad that Microsoft is keen on using test suites to measure standards compliance, these particular tests on MSDN only cover a very limited spectrum of capabilities, and statements about browser compliance made here are frankly misleading.

The SVG-in-HTML tests cited by MSDN are known to be broken, do not conform to the HTML5 parsing rules or the DOM specification, and in fact make use of an IE bug. Microsoft's engineers know about this bug _and_ the fact that the test itself is wrong [1][2], so it is surprising that this test is touted as an important one in furthering interoperability between browsers. Firefox's beta builds, on the other hand, actually do have an HTML5 parser in place [3], and nightly builds [4] will pass legitimate tests for SVG-in-HTML parsing.

Additionally, some of the CSS tests listed on MSDN are inaccurate, and browsers that actually pass these tests are not compliant with the standard. There is an official CSS3-selectors test suite where we pass all the relevant pieces [5] and we think that for web developers, sites like caniuse.com [6] are much more useful when discussing HTML5, CSS3, and SVG support in browsers. Mozilla will continue to contribute to open and fair test suites that adhere to the specifications in question [7], and encourage other browser vendors to do so as well."

If people keep writing test cases, eventually we might know what HTML5 is!

Funny how chrome gets 100 on the acid3 and ie9 gets a 95.

http://ie.microsoft.com/testdrive/benchmarks/Acid3/Default.h...

Funny indeed. I mean look, IE9 is definitely a BIG and most excellent step in the right direction. If only it was supported on Win XP. If we can get rid of IE8 and have the few top-notch browsers start going head-to-head with features and getting updated regularly, it would be a good day indeed.

Both MS and Mozilla have explicitly stated that they're not going to get 100 in Acid3.

The results table is also extremely misleading. There are many tests with no results for a particular browser, and in some cases the table gives a 0% score for no results in a category, and in other cases 100% for, say, two passes plus a bunch of "no results". At this point, you should probably treat the percentages as random numbers, not meaningful metrics of anything.

See here and followups for discussion: http://lists.w3.org/Archives/Public/public-html-testsuite/20...

I don't think this so called test/benchmark has everything that is under the HTML5 umbrella. It lacks WebSQLDatabase, Geolocation, offline storage, CSS3(debatable) and many more.

So since this test is not complete, I call this article: false advertising.

WebSQL has been dropped in favour of IndexedDB (both are separate specs).

Geolocation has been removed from the main HTML5 spec.

CSS3 is not HTML5 and that fact is not debatable at W3C.

  CSS3 is not HTML5 and that fact is not debatable at W3C.

I don't understand: the ongoing efforts seem to be centralized at http://www.w3.org/Style/CSS/current-work

In marketing materials (like Apple's HTML5 showcase) "HTML5" is a buzzword that means any technology that makes cool things in a browser.

However, W3C does not follow buzzwords. When W3C says "HTML5" it means http://www.w3.org/TR/html5/

CSS is developed by W3C, but it's a separate working group, and CSS is orthogonal to HTML5 (i.e. you can have 100% HTML5-conformant browser that does not support any CSS at all!)

> CSS3 is not HTML5 and that fact is not debatable at W3C.

HTML5 is like AJAX, it's just an umbrella term in reality. When you hear "AJAX library" you probably think of a library that does a lot more than just HTTP requests. Same goes for HTML, it's almost useless without some CSS.

> Geolocation has been removed from the main HTML5 spec.

InnerHTML wasn't in the spec originally, but everyone used/had it anyway because it was a good idea.

> WebSQL has been dropped in favour of IndexedDB (both are separate specs).

It was dropped only from Firefox, just like h.264, but that didn't stop anybody from using it. Some browsers have IndexedDB, some have WebSQL. IndexedDB was just a work in progress last time I checked, WebSQL already had existing implementations.

> HTML5 is like AJAX

Yes, but we're talking about W3C documents using W3C terminology, not colloquial misnomers.

> InnerHTML wasn't in the spec originally, but everyone used/had it anyway because it was a good idea.

I was mistaken about Geolocation. It hasn't been removed (as many other similar features were), this one been developed as separate spec from the beginning.

Separation of HTML5 and other HTML5-ish specs is decided by way W3C works, internal politics, HTML group charter, and inter-dependencies between the feature and rest of HTML. It doesn't mean that something was or wasn't a good idea.

> It was dropped only from Firefox, just like h.264, but that didn't stop anybody from using it.

That's not related. W3C test suite covers W3C HTML5, not Firefox (and H.264 vs Firefox is a completely different problem).

WebSQL suffers from very real problem — it has only one DB implementation, which is SQLite. This means that sites will end up denending on that particular SQL dialect with all its quirks (e.g. sqlite storage is weakly, dynamically typed) and in long-term will require to be bug-compatible with SQLite.

I encourage everyone not to use it, even if it's available now, because we'll miss last chance to remove it.

(I love SQLite in my apps, but API for the web needs to be much smaller and tighter than that to withstand decades of people copying & pasting code they don't understand).

The article links directly to the tests on W3.org - where they are clearly endorsed by the HTML Working Group.

If you want tests to cover other areas then feel free to submit them, or encourage others to do so.

Would seem to be more of a test of canvas than HTML5 in its entirity, given that over three quarters of the tests (179 out of 233) are canvas related.

Also, I'm not sure how seriously I can take an HTML5 conformance report that doesn't actually validate correctly (<center> element!?):

http://validator.w3.org/check?uri=http://test.w3.org/html/te...

Two important notes: current stable version of Chrome 7 so beta of IE9 should be compared with Chrome 8. Say it takes 6 months for IE9 to go gold, in that time Chrome stable will have had two more releases bringing it up to 9 which will probably contain a number of enhancements to close the gape if not over take IE9.

Yea every time I look at MS's site, they compare apples to rotting discarded oranges. "Look, our apples are quite superior!"

The charts are showing only the tests IE9 excel at, which is a fraction of the picture.

The tests also compare IE9 vs current stable releases. I want to see that comparison against beta releases.

Oh wait lets also write down what operating systems IE9 is compatible with:

Google Chrome: Windows XP, Windows Vista, Windows 7, Mac OSX, Linux

Firefox: Windows XP, Windows Vista, Windows 7, Mac OSX, *nix

Opera: Windows XP, Windows Vista, Windows 7, Mac OSX, Linux

IE9: Windows Vista, Windows 7

Oh look IE9 runs on only the latest MS operating system. They don't even support their own older OSes.

Lets not even get into the security aspects.

OS support is a little more nuanced than that.

The latest version of Firefox for OS X requires at least version 10.4 (released 2005). Ditto for Opera.

Google Chrome requires at least 10.5 (released 2007), and doesn't support PPC (the transition to Intel was `completed' in 2007).

Windows Vista went to retail in 2007.

Firefox, Opera and Chrome all still work on Windows XP though.

What fraction of the market is underserved by Firefox not working on OSX 10.3 and earlier?

Meanwhile what fraction is underserved by IE9 not working on Windows XP?

I have no objection to your numbers as long as we also compare apples to apples and quantify the impact.

Why did ReadWriteWeb omit the needed subtitle:

"IE CSS3 compliance continues to bring up the rear"

How come they didn't test a WebKit nightly as well? All other browsers are either latest betas or alpha(!).

That sure is an enterprise test runner they've got there, couldn't they come up with something more like http://html5test.com/ ?

This is very good, I'm glad the team have done such a good job - but I don't think the IE team will manager to 'win back' developers unless IE9 can replace IE6/7 and 8.

Ideally IE9 needs to fix the problems that the older legacy browsers created. Because it can't be run on Windows XP, IE9 can only complement the mistakes that were previously made.

My suspicion is that part of the motivation for IE9 is encourage upgrades from XP. Otherwise, there isn't much of an economic basis for its development because unlike Chrome, IE doesn't send every keystroke in the address bar to a server for analysis.

I agree - that's the main motivation .. but realistically, a lot of people will stick with Windows XP.

Microsoft rarely seems to think that altruism can align with good business sense, and I think it suffers because of this.

HTML5 compliance is fine and good, but I'm mostly concerned with how often I'll have to fix a broken layout grid etc.

If I hadn't checked myself into Compulsive Domain Buyers Anonymous already, I would grab ie9hacks.com and ie9bugs.com

If you want to see how browsers stack up on finalized web specifications, I find http://www.webdevout.net/ is the single best place to go.

What is IE9 score on http://html5test.com/ ?

Not that this test is official or anything, but IE9 beta gets 96. Preview 6 gets 106.

The chart shows: "Internet Explorer Preview 6".

That's right, IE9 Platform Preview 6 is the release that was rolled out at the PDC last week,

http://www.betanews.com/article/Microsoft-launches-Internet-...

Get back to me when IE passes a significant battery of Firefox-submitted tests. W3C is owned by Microsoft at a time when Microsoft has lost all right to call themselves a browser vendor for any purpose other than to harm the browser market.

Get back to me when IE 9 runs on any of the thousands of machines in my company, where Windows Vista/7 is a non-starter at least for the next two years. Get back to me when it is able to run on any machine I have at home. It never will. With the possible exception of Safari, every other browser tested runs everywhere required, I think.

If Mozilla want to submit tests then there's nothing stopping them doing so.

> W3C is owned by Microsoft

It is?

Microsoft has an interest and is a member of the W3C. They do not own them though: http://en.wikipedia.org/wiki/W3c

In other news, record low temperatures were recorded in hell.

Reply on news.ycombinator.com