95 comments

[ 2.6 ms ] story [ 161 ms ] thread
:(

still no canvas or svg... jerks.

At least it's almost as fast as the slowest competitor...
Yeah, I love how they say IE is no longer an "outlier."

They need to use trickery in the graph so that it's hard to tell that it's still 3x slower than Chrome or Safari.

I am shocked you would expect an honest report from Microsoft, about a very important product during the heavy media coverage of the PDC
They did show how IE8 is orders of magnitude slower than its competitors - that display of honesty shocked me a little.
They admit the flaws when the next version is about to come out. Standard operating procedure for marketing.
That's the only reason we have to imagine IE9 will be released soon.

Had it been a "minor improvement" over IE8 and were it aimed squarely as a competitor to Chrome or Firefox, I would doubt it would ever launch. Just preventing companies from rewriting internal web apps to be more cross-browser-friendly would be sufficient accomplishment.

I'm looking for the trickery, and I haven't spotted it yet. Can you elaborate?
I think he's speaking of y-axis' scale. Using such a large scale and including older versions of IE makes IE 9 look downright fast.

If you plotted just the rightmost cluster (IE 9, FF 3.6, WebKit nightly) on its own graph with a scale that went from 0 to, say, 1500 instead of 16000, it would be very clear that IE 9 is still slow compared to it's peers -- nearly 3x slower than WebKit as noted by the earlier comment.

It's quite an improvement, but here's hoping they keep optimizing.

Also, Firefox already gets snippy with Webkit because they bail on dealing with difficult corner cases (the page showing the HTML5 spec itself being a recent example), then trumpet how fast they are. Let's see how fast IE9 is when it actually supports the standards properly.
No mention on if they actually, truly, really support transparent PNGs this time around, either.

(If you think they fixed that in IE7, try setting opacity on a transparent PNG....)

that's a bug in their implementation of opacity, not of png. they don't support existing alpha in their opacity calculations.
Whoops, you're right. I hadn't really read in-depth when I ran into this several months ago, and I was rather peeved at the time regardless.

It doesn't really matter though - the point is, they still have a long way to go before they can earn back the trust they've squandered amongst the web development community.

I wonder when or even if the attitude seen in the comments will spread out to the rest of the computer community. (IE is the devil)
Maybe I'm just an optimist but I'm really glad to see that MS is even looking at the Acid3 and SunSpider tests as important metrics.
Me too. Not that they have much choice. Even the non-geeks in my office have noticed that they can work faster with Firefox.

As more things move to the web, built on web standards, they can't afford to continue having the worst browser by far.

To be far. IE8 is vastly better then IE7 and IE6 but Microsoft is not aggressive about getting it's users to upgrade. Whereas Firefox, Safari and especially Chrome are.
If only we weren't going to have to live with IE6 until 2020. Then we might be able to use more of the new stuff.

(Apologies for bitterness, I spent a lot of today back porting a Very Large Web App to IE6.)

Microsoft's efforts towards (de facto or de jure) standards compliance are dismal.

I've been working on a client-side app since before IE7 came out. The much hyped IE7 release gave us some performance benefits, but apart from that it just added another IE release that had to be tested and tested.

IE8 made things even worse.

We now have scenarios that work fine in all recent (going back 1 to 3 major versions, you know, everything over the last few years) of Opera, Firefox, Chrome and Safari, but IE? Hahaha, you must be joking. It works in IE6 and IE8, but not IE7. Or just in IE8, or in IE6 and IE8, but not IE7...

And now they're threatening to unleash yet another IE version.

Why couldn't they just stick with IE8 for the time being and fix things there? (And don't tell me 'the version number is just marketing', we all know that major releases entail lots and lots of changes and 'fixing things we messed up before').

I like the effect of competition on Microsoft.
I'd like it much more if it made them irrelevant, or preferably bankrupt.
Yes, saying bad things about Microsoft is frowned upon.

OOXML & bribes, anyone? SCO? Patent threats? Netscape? Holding back the web for 10 years? Countless other acts of evil bullshit?

I guess this is the same kind of denial that makes HN throw a hissy-fit at trying to bring up the very real problem that is Islam (in Europe).

See no evil, hear no evil.. What a wonderful community of very intelligent people we have here, heads firmly lodged in their asses. Let's never talk about anything that really matters in the world.

Aww.. You're so predictable!

Do I get more denial-groupthink downvotes with this one?

I believe the more cliché 'group think' reaction would have been to up-vote your Microsoft bashing.
I believe I mentioned denial. Take a look at that video of angry muslims, by the way :)

I wonder if this will get me banned. I'm not trying to get banned, but it might happen. It'll be interesting to find out.

Just say "hitler" already so we can know the conversation is over and prune this ridiculous tangent of a thread from the top of the page...
Yeah, or you could give things some thought, find out for yourself, and join the choir. Instead, you're proudly taking in the scenery inside your rectum.

I was just trying to spread awareness. It feels weird to even say that, but something has to be done. Of course, it could be argued that this is not the place for warning the world about the insanity that is Islam, but what is?

There's no forum in the world for talking about how dangerous Islam is, where people wouldn't already agree.

HN on the other hand, is full of intelligent people who certainly should be able to handle coming across material that is unpleasant to digest, and adjust their world views accordingly.

There really was some kind of protest full of enraged muslims in London, who were pretty much saying that they'll destroy Europe, and it sure wasn't the only enraged protest they've made. They really genuinely did kill a man for making a short movie that talked about Islam. Shouldn't this kind of stuff at least raise an eyebrow or two? Can you really dismiss all of this as Crazy Talk? Should you? Or should you maybe find out for yourself? Go on, confirm that I'm crazy. I dare you!

There's a blogger in Finland who's been talking about Islam & (muslim) immigrants for years, but only became famous last year when enough people noticed. He became the victim of a nation-wide witch-hunt lead by the local media, and eventually even got fined for something he said on his blog. You see, in Finland we have freedom of speech as long as we're careful not to say anything unpleasant. I was somewhat surprised he didn't get thrown in jail.

What the blogger has done requires giant balls of steel. What I'm doing here is the only thing I'm aware I can, however misplaced or aggravating it might be.

Also, "hitler".

Saying "lololol M$ i hope they die" is frowned upon. Post something with content, support your whining, talk about it.

Microsoft have opened their internal development to the world over the past few years - this blog entry is an example - from all kinds of development teams, we have enormous amounts of insight into many of Microsoft's major products and access to give feedback to people who care.

Don't just focus on the 'evil bullshit' (much of which is probably not 'evil', just the inertia of a huge company).

You can't just drop "Islam is a very real problem" into the middle of a rant about a software company and expect people to think you're not a nutter.

Post something with content, support your whining, talk about it.

Content? I did mention a few examples, didn't I? Well then, would you care to tell me why they're all invalid? I've done this kind of thing before, and every time someone disparages MS here, he's treading on thin ice. There will always be MS-apologists around.

Microsoft have opened their internal development to the world over the past few years

They sure have made a show of being more "open" to please/distract programmers, that's all.

You can't just drop "Islam is a very real problem" into the middle of a rant about a software company and expect people to think you're not a nutter.

I realize this is something that people don't want to hear, and can't process without adjusting their conceptions of things, but that doesn't make it any less true.

Want some content? Here's a video of angry muslims in London: http://www.youtube.com/watch?v=qoMeUcC_M20. One of the signs they're proudly displaying says "Europe, you will pay. Your annihilation is on its way!!!".

Does that sound a tad threatening to you? That protest is probably related to events in 2004, when they actually killed a Dutch film director, Theo Van Gogh, for releasing this: http://www.youtube.com/watch?v=1tvH_uowO-0 .. Yes, they killed a man for making a short movie.

The average muslim you meet will most likely tell you that Islam is a "religion of peace", but what they don't mention is that if you say otherwise, they'll kill you. If it's such a mellow religion, why is there a protest full of its followers waving signs saying that they'll destroy Europe then?

Western people believe the line about a religion of peace because they want to, because we should all just get along and coexist in peace and harmony. That's just the humane way of life that we as Westerners are wired to like and want, but that's not what muslims have in mind.

----

Update:

Quick! Someone call PG and get him to delete this thread so no one else will see what I'm saying here!

Everyone else on this page is posting interesting observations on Microsoft, development, and the web. It makes me angry that you've come into my little garden of contemplation and started yelling about irrelevant subjects.
Content? I did mention a few examples, didn't I? Well then, would you care to tell me why they're all invalid?

They are invalid in a discussion about IE9 developments because they aren't about IE9. They are valid in a discussion about web development in general... except they are arguing against your citing 'netscape' and 'holding back web development' because this blog is about MS not holding back web development and not destroying a competing browser. In other words, Microsoft has moved on from those (e.g. this page) so those aren't good support for your complaints anymore.

People in this page are saying bad things about Microsoft, about their javascript interpreter speed, their lack of canvas support and rounded corners and font handling and they aren't being voted down. Any guesses why?

I am an MS apologist. I have worked with their stuff for the last six years and find it (mostly) better than I previously thought, and not deserving of quite as much of the mudslinging as actually happens online.

Islam

Don't know anything much about Islam and couldn't give two whoots. Particularly don't want to discuss religion or country/world politics on HN, even more so in a thread about an IT topic. By the way, I also don't want to discuss dishwasher detergent, pie recipes, good books on optical illusions or how to build a canoe.

They are invalid in a discussion about IE9 developments because they aren't about IE9.

Well then, why did you insist that I should "support my whining"? I'm pretty sure my "whining" was just as not related to IE9 as it is now.

Besides, if you squint and look really hard, you'll notice my goal here was not really to discuss the finer points of IE 9. You've obviously got a problem with that, but how much would you surmise I care?

They are valid in a discussion about web development in general... except they are arguing against your citing 'netscape' and 'holding back web development' because this blog is about MS not holding back web development and not destroying a competing browser.

I take it that MS discussing IE 9 on a blog means that they actually never did any of the things I listed?

People in this page are saying bad things about Microsoft, about their javascript interpreter speed, their lack of canvas support and rounded corners and font handling and they aren't being voted down. Any guesses why?

Wouldn't this be related to "holding back web development"?

I am an MS apologist. I have worked with their stuff for the last six years and find it (mostly) better than I previously thought, and not deserving of quite as much of the mudslinging as actually happens online.

Glad you came clean. Even I realize that they've been working hard to improve the .NET platform. They have to, because their grip on the world is dependent on developers, developers, developers.. Did you see what they did to improve IE while they weren't worried? Oh that's right! -Nothing.

Don't know anything much about Islam and couldn't give two whoots.

See, now that's the exact same problem I'm trying to solve. You should know that there are people following a 1300-year-old sacred manual on how to behave that says "kill everyone who's not a muslim", and it's happening in 2000-fucking-9.

You don't see Christians going on Crusades anymore, but you do see enraged muslims shouting that they'll destroy Europe. And Europe's collective reaction? -To apologize, look away, and hope that they calm down.

You've obviously got a problem with that, but how much would you surmise I care?

I guess I'll say the same about your terrorist scaremongering and call it a day then. Nice talking with you.

The difference is that you really should. Oh, and I'm not talking about terrorists. I guess that's what it looks like to an American.

Are terrorists known for holding protests on London's streets where they shout they'll destroy Europe? No? What are those people then?

You call it scaremongering, I could call it trying to dislodge heads from asses, or just "trying to spread awareness" as I said earlier.

I'm sure you had good counter-arguments for everything I said too.
> "perceiving the differences may be difficult on real-world sites"

Geee, I wonder why?

Could it be because noone can use javascript properly on real world sites because IE6, IE7, IE8 have effing 1/100th the javascript performance as any other major browser????

Eh, I don't buy that. I've got a client site that does fade-in/fade-out effects on multiple simultaneous elements, ajaxyness, and horizontally scrolled lists of images. It also uses jquery-like css selectors to target elements, and it intelligently re-renders (in realtime) selected page elements when the page is resized.

I can do all of that in IE 6, 7, or 8, without maxing out a CPU core.

What I can't do, and really would like to do, is offload more of the page rendering to Javascript, and the only reason I can't do that is because of the various noscript users out there.

Regarding hardware-accelerated drawing of text:

I thought that IE just like any other Windows program is using Windows core APIs like GDI+ to display text. And those APIs, in turn, are hardware-accelerated as much as possible. These days the entire desktop is essentially a "game" running on top of DirectX, no?

Even back in 1995 most of GDI 2D functions were hardware-accelerated, assuming you had a good card with good drivers.

What am I missing? (I haven't done any Windows work in 5 years).

The part of the post about the font rendering is BS anyway. The picture is supposed to show "sub-pixel positioning" and it only shows antialiasing (not even sub-pixel antialiasing) which as you've pointed out isn't handled by a browser anyway.

"Sub-pixel positioning" on the other hand would indeed need to be handled inside of the browser, so that the engine knows how much space something takes up so it can position the next item- But the picture accompanying the post is wrong/irrelevant.

Not sure I agree - most fonts are cached (as pixels) at a few offsets, so the ability to place a character is constrained by 1-pixel increments (or 1/3-pixel increments in the case of ClearType).

A true vector placement of each character gives you the ability to show type at any subpixel offset and then anti-alias, and I think there's a good chance they're doing this.

See this Loop/Blinn paper for more details: http://research.microsoft.com/en-us/um/people/cloop/loopblin...

Summary: still a long way to go.

No announcement yet on @font-face, canvas, video. Basic performance improvements.

The @font-face spec doesn't mention a font standard, and Microsoft have already said they won't implement TTF or OTF because the font foundries don't like it.

It's possible that IE will implement the upcoming WOFF (Web Open Font Format) standard in IE9 but I wouldn't rely on it.

Sigh.

Thanks a bunch for the info.

This is nice - I don't think I've seen Microsoft look at themselves this candidly before.

Next thing you know we'll have commercials saying "IE9? That was my idea!"

It'll be people dressed up in Firefox, Safari, and Chrome costumes. "One day, I had the idea that browsers should follow web standards. And guess what? 10 years later Microsoft started following them! I'm an open source web rendering engine, and IE 9 was my idea."
Honest question: what does Microsoft stand to gain by doing all this work instead of just using WebKit?
Control and influence. Backwards compatibility. And they probably still believe they can do it better.
I'm not sure that they can't. Give Microsoft a compelling business reason to build a great browser that also supports IE 8,7, and 6 rendering modes, and I bet they'll do it.
Isn't reclaiming the 20-30% market share they lost to Safari/Firefox/Chrome reason enough?
They're trying to support more standards -- by definition, they're just reinventing the work others have already done. So it's grunt work, only useful for their future influence and power.

They could ship IE9 with WebKit in addition to custom rendering modes for IE<=8, and consumers wouldn't even need to know. It'd save MS a lot of work, that's for sure.

I say let them worry about the cost/benefit ratio. If they can afford it, another engine with completely separate codebase is a good thing for the market.

Just think how hard it would push webkit/mozilla/chrome if they did manage to beat them on speed. It would push them to get even better. Compeition is always good as long as 2 or more projects stay alive.

...another engine with completely separate codebase is a good thing for the market.

You've just made an enemy out of every web-dev on HN :)

Probably not. I suspect that as long as IE is the default browser on Windows, their revenue doesn't change that much year over year. People who don't change their default homepage don't change their browser either.

What is a compelling reason is revenue loss in their cash-cow franchises like Office. If they come to believe that they need a version of Office that works in the browser or they will lose revenue, then developing a fast version of IE is very important. If they believe that people won't use a computer that doesn't provide them with great experiences on the sites they love right out of the box, then developing a standards compliant version of IE is important.

Yeah, but if they destroy all other browsers again, then they don't have to do any work on improving browsers. Making browsers actually goes against their business process, they want everyone on the desktop, not the web. They have been dragged into the web, and hated every step of it.
I think part of the problem is that they don't seem to view building good products as a 'compelling business reason.'
I think the different rendering modes in IE8 is a serious defect. For example, when testing locally IE8 defaults to IE7 mode instead of IE8 standards mode. You physically have to go into the developer toolbar to switch it. My friend(who's a .Net developer) ran into this exact problem last week and couldn't understand why his site was failing in IE8, looked good in IE7, and worked well in FF & Safari. I've run into this problem more than enough times.

If people need to use a particular older browser for a very narrow application, then they should keep an installation of that browser around for that specific purpose.

Its the same idea with modular programming methods. I build my js in modules. My comment system is separate from my login system. In the last couple of days I have been updating the Facebook Connect logic in my login system and I haven't had to worry about updating logic in my comment code.

IE is deeply integrated within windows. Rewriting everything for webkit may be costly.
If I recall correctly there's some weird incestuous relationship in the HTML rendering in Outlook, Word and IE.

I agree it's almost definitely technically infeasible.

Nope, Outlook now uses the Word rendering engine, since it's part of the Office family. Totally makes sense right?
That was it, thanks for the clarification.

I still can't believe they're maintaining two distinct HTML renderers...

Considering the politics of such a large company, they're lucky if they only have two.
You'd be right about that. There's another one in the Media Center division, and it happens to be much more strict with respect to standards, although perhaps not as full-featured.
mind elaborating on this? I'd love to hear about it.
A friend and former colleague used to work there. He was pretty parsimonious with the details, unfortunately.
Describing what the Word engine does with HTML as "rendering" is probably being charitable.
Well, if Outlook uses the Word rendering engine, then at least it's somewhat decent, looking at some HTML formatted emails that fill my inbox.
Only because a lot of HTML email is tested to make sure it’ll appear at least somewhat as expected in Outlook.
This is a good point, but having just designed an HTML formatted email don't discount the fact that Outlook is popular enough to gain bend-over-backwards-to-make-it-work status. Having said that, it wasn't too difficult to get it looking good.
There's at least 4 independent codebases — modern IE (which includes 3+ renderers), MS Word, archaic Mobile IE, and the new IE6-based Mobile IE they have in development.
I suspect the other replies are way off base. Legal concerns are probably the largest blocker by far. Apple and Google don't have enterprise customers for WebKit, but Microsoft has many for IE. Since buying Microsoft software is effectively like buying job-insurance for IT staff, Microsoft's risk accounting is wildly different from Apple and Google. I'm not saying deciding not to use WebKit is the right risk/reward trade off, but IANAL.
What do you mean by "enterprise customers" for IE? I don't understand the point you're trying to make.
People who have written IE only, active-x infested applications are the "enterprise customers" of IE :)
Is it still conspiracy theory to suggest that they are the ones with the most to lose from a universally reliable JS and HTML implementation, so they should want to do enough to keep IE competitive, but not enough to make it fully compatible with everyone else?
I think they will way the possibility of maintaining control through developers choosing their implementation over one the other engines choose to user. If they see a chance for developers to pick their standard they'll push it, if not they'll back off and adopt the other one. It can be seen in the history of the engines, and will continue that way. Like evolution of standards, the most awesome ones according to developers live, the others die.
Microsoft walks a fine line between having a browser that's broken enough so that web apps written and tested for it break on other browsers and a browser functional enough not to drive users to move to better products.
ActiveX support and other backwards compatibility with 5/10 year old Intranet systems people have bought.
The most positive thing that came out of that keynote was that they focused on talking about achieving feature and performance parity with competing browsers, not adding useless features like Accelerators and Web Slices.
Here is what I want them to do: Release an IE shell that wraps WebKit. Leave all the Trident DLL's there since they're heavily baked in for help file viewers and other OS bits and parts.

No one (except for us nerds) care what rendering engine they use as long as things look ok and work. What people do care about is the UI the use to get to content. So the IE can go balls to the wall crazy with whatever UI they think users will enjoy, but just leave the heavy lifting to a way more capable machine (this is essentially what Google did with Chrome).

But then they couldn't render with Direct2D and DirectWrite, which has potential to be a valuable improvement.
Problem is, a lot of custom software depends on ActiveX controls and other plugins that work in IE but not anywhere else.

Until WebKit or Gecko start wrapping activex (not likely and doesn't seem like something they should be spending time on anyway), IE will continue to have significant usage - at least within companies.

Also, there's the pride issue of dropping their own rendering engine for one sponsored by Apple.

Right, thats why I'm saying leave all the IE rendering engine bits and pieces, but the IE UI itself is built on webkit. Didn't they lift some BSD networking code for NT back in the day? So clearly they have no problem with doing it :-) (indeed its encouraged by the license). And developers the world over would celebrate!
Border radius!
Sweet Moses yes. Rounded corners should be a setting, not a project.
The thing that shocked me the most was reading that Direct2D which can be used in place of GDI+. IMO, Microsoft should have made available a hardware-accelerated 2D library a LONG time ago ... before this, you had to use Direct3D to attain hardware acceleration, which was always far too much of a hassle just to render some 2D graphics.
32/100 is a little underwhelming. If it's a focus on standards then it's still dreadful. Am I missing something?
This is good. The end-users/web devs will benefit. We have more options, more competition and hopefully better browsers because of the competition.