Internet Explorer 9 destroys Chrome 6 in HTML5 speed test (video) | Hacker News Reader

Internet Explorer 9 destroys Chrome 6 in HTML5 speed test (video)

downloadsquad.com

42 pointsdanishkhan49 comments

Congrats to the IE team for getting back on the rails!

More competition in this area is a bonus - now we just have to ensure that microsoft continues to sound out valid reasoning to get the vast proportion of Enterprise that still is on IE6 updated (and thus shift their legacy apps from proprietary hacks to a far more future "proofed" (resistant really) standards base).

Using hardware acceleration in a contrived example, IE 9 is faster at rendering fish. Yawn. Nothing Microsoft does now can make up how awful IE 6,7, and 8 are and the fact that most of my day-to-day pain is caused by supporting their crap-tastic browsers. Even IE 8 leaks memory like a sieve.

"Using hardware acceleration in a contrived example, IE 9 is faster at rendering fish."

It's not a contrived example, as someone already pointed out.

"Nothing Microsoft does now can make up how awful IE 6,7, and 8 are and the fact that most of my day-to-day pain is caused by supporting their crap-tastic browsers. Even IE 8 leaks memory like a sieve."

This is entirely irrelevant.

Why is this comment being upvoted?

The trust has been lost with a lot of developers. I guess they are going to have to earn it again.

I don't think this example is contrived. I can imagine plenty of use cases where being able to draw lots of objects is really important. Hardware acceleration is exactly what HTML5 needs in order to compete with Flash.

True, true.

Wonder what the result would be if the video maker had tried using the WebGL extension... if this FishIE thing supports it.

Look, I really dislike Microsoft, but we're not allowed to discuss why on HN. :-) However, IE9 really can catapult the Internet to the next level by being highly standards compliant and having great performance. Let's give them credit on this one. Safari, Chrome, Firefox, and Opera competing against a strong contender from Microsoft is a great thing.

Lets give them credit when ie9 has 97%+ market share of the IE browsers.

Anybody can make a nice demo, it is getting people to use it that matters.

Am I the only one who's actually excited that Microsoft is actually making some very tangible improvements to the IE platform?

The "let's hate on MSFT for past browser wrongs" attitude is worrisomely common and accepted 'round these parts. It's a strange form of blindness.

From where I sit, IE9 looks like a genuine attempt by MSFT to get back in the race.

What's frustrating to me is that they could fix the problem of high usage of old IE versions. They just seem to be unwilling to.

They could add in IE7's and IE8's rendering engines into IE9 as compatibility modes, push out IE9 as a forced or automatic update and have HTML5 ready to use in a few weeks. But they seem to be more worried potential whining from update-averse users than holding back the technological progress of the entire internet.

Who to piss off: the IT department at almost every major corporation worldwide who pay us billions of dollars a year, or a bunch of bloggers who run Firefox on Macs and would still hate our bones if we cured cancer tomorrow.

Decisions, decisions.

Why would IT departments be pissed off if IE9 was fully backwards compatible? Honest question.

The entire point of the suggestion is to break compatibility with IE6 so that the rest of the world can stop having to code against it. Breaking compatibility with IE6 has freaking enormous switching costs for some users.

(Let me hum a few bars: you use a $3 million CRM which only supports IE6, and the company which bought the company which made it has since folded. This is hypothetical, but not very hypothetical, if you catch my drift. A forced free upgrade to IE9 would create an organization-wide emergency for that customer, instantly, and it would be a cold day in Hell before they every do business with Microsoft again.)

The entire point of my suggestion is that a newer browser doesn't need to break compatibility with older browsers to allow people to stop supporting them. Backwards compatibility makes it possible for Microsoft to do a forced upgrade without screwing anyone over. Once a significant majority people are using IE9, it doesn't matter if they're still relying on its backwards compatibility for existing websites, new websites can just target IE9's native rendering engine and ignore older versions of IE.

I didn't mention IE6 backwards compatibility, because no one running IE6 is going to upgrade directly to IE9; they don't run on the same OS. Even if IE9 was made a mandatory or automatic upgrade, people relying on IE6 wouldn't have to worry. IE6 would have to be dealt with differently than IE7&8, which is fine, it'll get to insignificant market share soon enough on its own.

The currently existing alternative is just to install Chrome or Firefox for the current web and leave IE6 in place for legacy internal applications. But then browsing the web isn’t seen as a useful part of people’s jobs in most companies.

> $3 million CRM which only supports IE6, and the company which bought the company which made it has since folded.

You can make it entirely realistic by saying the company that made the CRM now has a version that supports Firefox and IE7-8 but it costs $5M to upgrade the license. I've had so many of those cases at my current job.

IE9 does have compatibility modes for IE7 and IE8 already (somewhere on the official IE blog there's a flow chart describing how IE decides what rendering mode to use for a given page; it's terrifying)

My understanding is that IE's entire user-base consists of update-averse users these days - everybody who can stomach the idea of changing browsers is already using Chrome or Firefox.

Not quite. IE8 is making huge gains in market share, but almost entirely at the expense of IE6 and IE7. The overall IE marketshare is holding somewhat steady at 60%. I think what you're seeing is Windows users are perfectly happy to take an upgrade that Microsoft offers, but either afraid to branch out to a non-MS browser or else just don't know how to do so. Windows 7 shipping with IE8 has a lot to do with it, too, as people buy new laptops, etc, and never touch the browser it comes with.

Reference: http://www.eweek.com/c/a/Windows/Microsoft-IE-6-Share-Falls-...

I wish Chrome would take more from IE than it is from FF, but it is good to see IE6 drop so much. It's under 5% now.

Note: these are US numbers. If I recall correctly, the global numbers are shaded a little more non-IE.

No they certainly could not. Have you ever worked in a big company before? The end users get their updates from in house sources. Introduction of MS updates is strictly controlled. If they can't quantify what the change will do then it's not going in. Of course security patches get priority and something like a browser version change has broad sweeping effects that are extremely difficult to quantify.

So if MS did what you ask tomorrow, nothing would change in 99% of big corporations.

They campaigned extensively for users to switch to newer version of IE. They don't force it because lots of company still have IT policies of using IE6 (or whichever comes installed with the OS, I suspect).

I forget where I heard it, but in some interview a Microsoft exec claimed no one at MS likes IE6 and they all want to see it die, but that they have signed long running service agreements which basically makes them contractually bound to keep it alive. Now this could all be bullshit, but that appears to be the official line.

Isn't there some old anti-trust thing that doesn't let them push IE updates? That was my recollection.

Edit, security ok, version upgrade not ok.

I'm disappointed they didn't make a three-way comparison with potatoes.

What a misleading headline.

"Internet Explorer 9 beats Chrome in hardware-accelerated canvas rendering, fails to beat several other browsers (video)".

Much better. But hey, gets less clicks right?

I work with large 3D models in my day job.

Hmm, maybe I should switch the competition's product to use software rendering, and I'll use OpenGL/DirectX, for side-by-side comparisons.

Wow, I'm "destroying" my competition...

TLDR/W: Hardware accelerated IE9 HTML5 canvas beats non-hardware accelerated Chrome 6 development HTML5 canvas.

http://www.downloadsquad.com/2010/06/24/4-way-html5-speed-te...

It beat hardware accelerated chrome as well.

Firefox 3.7 actually came out on type by a bit in the end at a 1000 fish.

Too Long Didn't Read/Write?

Too long didn't read/watch, I'd guess.

too long didn't really want (?)

Worrying about canvas graphic acceleration without having a standard canvas implementation may be jumping the gun a little, although traditionally that is how HTML has been advanced. Its a case of premature optimization, but then there is potential for payoff in being first to market with a winning strategy...

^This. Hard to read a video, but I wasn't sure people would get TL:DW without actually going to the site, which would have made a TLDR unnecessary.

TLDR/W = I hate Microsoft because I think doing so makes me cool in this community and I will get freebie points for bashing them.

Lol. I don't hate Microsoft, just almost all of their software, and some of their past business decisions. But there are plenty of decent folks working there, including a few I know, that I don't hate.

Does Chrome have HW acceleration otherwise? Serious question.

Pretty cool, now we only need to bring hw accel to linux and firefox :). My main browser will allways be a open one since we really shouldnt trust corporations in keeping the we open.

//Anders

All the other browser already have hw acceleration in various stages in their "preview" builds, in released builds to some extent too.

But not on linux/*bsd right? Guess you have to have a quite new mac for that too ( my 2008 Macbook wount cut it I guess :( )

Depends which browser/feature you're talking about. The WebGL stuff I believe is hardware accelerated in Linux Firefox, but some of the video stuff is Windows only for now and some other stuff is Vista & Windows7 only currently.

For the Mac are you talking about H.264 acceleration? Apple have historically not exposed the hardware acceleration features of the graphics cards they shipped, often not even using it themselves and just using software decode. They've changed this recently by making use of some (but not all) of the Nvidia (but not ATI) decode capabilities but it's not the hardware that's missing for many older devices, it's access to the hardware that only Apple can provide. These capabilities are accessible if you're running Windows or Linux on that same hardware.

In the Mr Potato head example video demo Chrome has 25fps before you click anything i.e. just displaying a static canvas doesn't get up to the 60fps seen for IE9. It also returns to only 25fps after everything stops moving again after you shoot the gun.

I thought that was a bit strange, so I tried to reproduce it. In the live demo the fps meter only appears after you fire the gun, and (apparently) stops with the last value after the items stop moving.

What's going on there?

It's one of those things that you don't want to be true... But there it is. Ben Parr was at a press event today for IE9 and confirmed to Mashable staff that, according to the tests, IE9 is gonna be hella fast.

...why do you not /want/ it to be true?

It's a good thing that all other browsers have stopped development. Or maybe a better test would be comparing browsers with hardware acceleration against each other? Does anyone compare two games, one with hardware rendering and one with software rendering? No, of course not. That would be crazy.

Chrome is currently the fastest at this; what else would you have them test against?

The current developer build of Chrome/Chromium has hardware acceleration (switch enabled). The 3.7 developer build of Firefox also has it (and it's supposed to be faster than IE9). Why didn't they compare those, since IE9 is also a developer build?

Chrome's acceleration is not of the entire page. It is only acceleration of things like 3D transforms and a few other things.

They can test against whatever they want, but the fact still remains that IE9 is basically using 'specialized hardware' to reach that speed. What happens when Chrome also supports this specialized hardware? [i.e. I hope that the IE Team doesn't feel like they can skimp on performance elsewhere by just throwing hardware at the problem, rather than fixing/improving their engine.]

Chrome is currently the fastest at this

No it isn't. Firefox in software mode (release builds) is generally faster than Chrome when it comes to canvas. Firefox in hardware mode (pre-release, Direct2D) is around as fast as IE.

A competitive marketplace is proving better than stagnant monopoly. Long live the competition.

Reply on news.ycombinator.com