IE11 Release Preview for Windows 7: 30% Faster than Other Browsers | Hacker News Reader

IE11 Release Preview for Windows 7: 30% Faster than Other Browsers

blogs.msdn.com

36 pointscleverjake47 comments

I'm sure this will gather lots of downvotes, but it really does sound a lot like this to me:

    IE is faster than the rest.
    -- Microsoft

If you're about to publish such numbers, at least have an external party publish it, paid or otherwise. 30% faster loading would be astounding, especially on connections with high ping times and more than 1 percent of packet loss (3 percent is really noticeable and yet common in lots of places, especially public ones). But of course they didn't mean 30 overall improvement, just the Javascript got 30 percent faster if you use certain things, right? Let's click that link.

...

> IE11 Release Preview is 9% faster than IE10

Alright so there goes the "30% faster than other browsers". Unless you don't count IE10 as a browser. Also IE10 is way slower in all sorts of benchmarks compared to other browsers.

To take just one recent benchmark: http://www.tomshardware.com/reviews/chrome-27-firefox-21-ope...

Here IE10 is 257% slower than the fastest browser 379pts vs 974pts). If IE11 is 9% faster, then how is it "30% faster than other browsers"? A 9% improvement is still 248% slower than Firefox, at least in Tom's Hardware's benchmark. Or is "other browsers" not "all other browsers?" Perhaps that is their reasoning: "we meant only Opera, Maxthon and $other_unknown_browser as 'other browsers'."

You linked to article where results in Sunspider were excluded because of too efficient dead code elimination optimization in IE. See page "Javascript and DOM performance".

Alright so there goes the "30% faster than other browsers". Unless you don't count IE10 as a browser.

I'm going to nit-pick, just because you did. The issue isn't counting IE10 as a browser, it's counting it as an "other" browser. Which arguably it is not.

Astounding amount of cribbing, when they do explain the 30% advantage here: "On Windows 7, IE11 Release Preview is 9% faster than IE10, which is a 5% improvement from the IE11 Developer Preview, and over 30% faster than the nearest competitive browser." And that nearest competitive browser is FireFox 23, because IE10 is still an IE family browser. Yeah let's ask a third party to run simple sunspider tests.

"Here IE10 is 257% slower than the fastest browser 379pts vs 974pts)" I would say something slower can be 0-100% of the faster thing. Being 100% slower means it does nothing at all. I would define IE 10 as being 61.1% slower and Firefox as being 157% (not 257%) faster. 100% is double the speed, 200% is triple the speed, which it's not.

"A 9% improvement is still 248% slower..."

That's saying that if A is 257% faster than B and B speeds up by 9%, A is 248% faster than B. I don't agree. 974/379 = 2.57. Increase the 379 by 9% -> *1.09 = 413.1. 974/413.1 = 2.35.

Which means Firefox is 135% faster after IE's 9% improvement.

Agreed, and it's really been disgusting me lately with MS. Not just IE, but virtually every MS product or service is 99% marketing-driven, to the point that it should be making politicians jealous. They don't try and prove why they're better, they just tell you that they're better. And no, cherry-picked results don't count.

All they do now is try to force a narrative about how they're "better" than some competitor in some wholly subjective (or at least, objective-yet-inadequate) way. Google vs. Bing? That entire fiasco was downright embarrassing.

When you don't support standards or commonly accepted conventions of modern browsers, and build a browser optimized for one operating system you SHOULD end up with a faster browser.

However, given the aforementioned limitation why would you want to use such a browser?

-----

for ie10 see

http://html5test.com/compare/browser/ie10.html

ie10 scores 320 vs firefox 22's 410

i'm downloading the ie11 preview to compare firsthand now

"You'll need to restart your PC before you can use your new browser" -- thanks IE

ie11 scores 350 on html5test (compared to firefox 22's 410 and ie 10's 320)

Have you developed for IE10 or 11? They are far more standards compliant than any other version. In fact IE11 surpasses Firefox in some areas (full support for Flexbox for example)

"... they are far more standards complaint than any other version ..."

"standards complaint" - Nice, I think that nails it.

Lately Microsoft complains all the time that they are fully standards compliant, but when it comes down to it, I always end up having to special case IE so it never really feels that way.

They are NSA standards compliant.

> They are far more standards compliant than any other version.

Of IE? Considering the other versions, that's not exactly a high bar to set.

I agree IE became an object of scorn in the past. Are there any clues yet which show IE11 will be different regarding standards and compatibility?

IE versions have rapidly become dated compared to the rapid-release leading edge of Chrome and Firefox, so IMHO the most representative comparison would be in the middle of the slower IE release cycle, maybe 3-6 months after launch when Chrome and Firefox have been updated a few times.

But there are some clues, yes.

Firstly IE versions from 7 to 10 have improved a lot each time. (or sucked less if you look at it that way). IE11 should continue this trend.

Secondly the IE release cycle is getting shorter. Not as much as Firefox or Chrome, but there is some response to this pressure.

Also, it's not like the IE11 feature set is unknown at this point.

That site is somewhat biased. Where are Pointer Events for instance, which IE would score highly on and Firefox not so much?

I don't have much confidence that those numbers are actually meaningful. IE10 counts as a "modern browser" in my book.

Yeah, well, when the standard changes are basically equivalent to Blink diffs, then it's difficult..

It's great all these companies are competing to make their JS engines faster and faster with each new release. Its creating exciting possibilities for the web!

I do wish there was a bit more aggressive a push to get older users to update their browsers as well though.

There has been a big push. As best I can tell, there isn't much more aggressive they can get, aside from ignoring the user and literally forcing computers with old browsers to update.

They could develop IE9+ for XP.

That would just make the problem worse. XP is EoL, and propping it up only makes it harder when the day comes.

The problem is not that XP can't run IE11. The problem is people are still running XP.

You share Microsoft's opinion.

I share the opinion of anybody who's already had to support a product for twice its intended lifetime

Yes, IE6/7/8 really need to go. I don't know why they can't force update everyone similar to Chrome.

A bunch of old megacorp in-house webapps would break if the IE would move from 6 to 7+.

Because a number of users (IIRC, South Koreans especially) rely on the incompatibility of IE6 to view a number of their websites. (Something about certain ActiveX controls being government mandated)

Force upgrading people from these older versions will break a significant fraction of the web, at least for some users. It might even be a legal issue in South Korea.

ActiveX still works in IE11 though.

Microsoft hasn't developed IE9+ for Windows XP. This might be to differentiate their newer OS'es.

I don't use Firefox anymore, but I feel that this is more or less because of them and for that I am eternally thankful. As for the lack of updating, I really don't think you can blame Microsoft for that. The problem lies with corporations.

I never trust code that Microsoft writes to make up statistics about how awesome their technology is.

How well does it handle real world examples rather than some contrite example of theirs?

It would have been worst for them to pay a lab for the test. Now that it is out, we shall see and test. MS threw the first punch with the claims

FF user and very happy here.

I tried to install it on Windows 8 on my desktop (not 8.1) and was informed that it is only for Windows 7 SP1. I could help but roll my eyes.

Yes, they want to force you to install Windows 8.1. I'd also prefer to have just the IE 11, at least until Windows 8.1 SP1 thank you Microsoft.

They are really strange: if I had Windows 7 I'd be able to install IE 11, but since I have Windows 8, I'll have to wait until October 18 and then I'll have to install the whole Windows 8.1.

The main problem with IE has never been its speed. I hope it dies a long agonizing death for all the hours I have wasted when developing two versions of each page - one for all the other browsers and one for IE. Its market share finally reached a point where I don't care about it... Good riddance!

It's disappointing that so many tech videos (in my experience) pay such little attention to the audio (bitrate and/or volume). My volume had to be at 100% to hear this comfortably.

Another example are published webcasts wherein the presenter is using their telephone (on speakerphone!) to give the presentation. That is just ridiculous.

If you do presentations don't use a system that doesn't allow you to use a microphone (and don't forget to buy and use the microphone).

The only reason I have not been able to switch to IE is browser extensions. Does IE have that yet?

I said the same thing about Chrome when it first came out. I happily use it now.

IE had ActiveX support for ages. There are other types of addons too, they're not exactly new. e.g. http://www.iegallery.com/Addons

But no one develops for them, at least not in the same way they develop for chrome or FF.

The new development tools look really nice, I must say. The dev tools in previous releases have been shocking.

Tell me about it, even in 10 I find the dev tools extremely lacking and slow.

What's up with "Reimagining the Net" - I don't want a browser to mess up with my image of the .net :) - do I?

Well the sad thing is the pirate test barely runs in IE10 and actually renders things wrong.

Haha, IE11 won't install on Windows 8. It requires Win7 only it says...

I was hoping for 31%. Are you certain it's not 29?

they have only 1 significant figure of accuracy

Let's just hope it's not an order of magnitude wrong, then.

Has to be tightly bound to hardware acceleration, no?

So what about Server 2012 R1?

Reply on news.ycombinator.com