Microsoft Aims To Close Performance Gap With Internet Explorer 9 477
Barence writes "Microsoft has unveiled the first details of Internet Explorer 9, promising that it will close the performance gap on rival browsers. The major newcomer is a revamped rendering engine that will tap the power of the PC's graphics card to accelerate text and graphics performance. 'We're changing IE to use the DirectX family of Windows APIs to enable many advances for web developers,' explains Internet Explorer's general manager, Dean Hachamovitch. As well as improving performance, Microsoft claims the hardware acceleration will enhance the appearance and readability of fonts on the web, with sub-pixel positioning that eradicates the jagged edges on large typefaces."
Re:Sub Pixel rendering, really? (Score:3, Informative)
But hey, that's a view that's not rabidly anti-Microsoft...
Re:Forget performance (Score:3, Informative)
BTW, why does Explorer (not IE, just basic file list explorer) take up 40 Meg? What on earth is it doing with all that memory just to display a list of files?
Re:Forget performance (Score:2, Informative)
Re:Forget performance (Score:2, Informative)
Re:JS performance (Score:2, Informative)
These are the sunspider results. Link [winisp.net]
Re:Forget performance (Score:2, Informative)
Re:Forget performance (Score:4, Informative)
Most Firefox memory issues since 3.x are due to bad extensions, not the core browser. Firefox is doing well with memory nowadays. I've had 2 windows, one of which has anywhere from 2 to 20 tabs in it, running all week on XP SP3, and haven't noticed any slowdowns.
Re:Forget performance (Score:2, Informative)
The Irony of you using a memory hungry OS and complaining about an application that diplays MEDIA is clearly lost on you.
Personally I want to access my information as quickly as possible.
Re:IE (Score:5, Informative)
> Is this the price you pay for having each tab run in a separate process?
That depends on the OS. On some the price of creating a new process is very high. On others a process costs only a little more than a thread.
Re:IE (Score:5, Informative)
> The issue is with the kludge design for multiple-tabbed browsing - which does the equivalent of starting an entire, new environment and plug-in
You mean like Chrome does? That's the BEST feature of IE8 - no more one-tab crashing taking down all yoru other tabs with more basic browsers like IE7 and Firefox.
Re:IE (Score:4, Informative)
The performance really depends on the browser's architecture, which is comprised of a lot of parts and potential bottlenecks.
Chrome and Chromium, for example, are heavily multi-processed and handle large amounts of tabs\plugins very nicely. It certainly doesn't hurt that they were designed from the ground up for this kind of behavior.
Re:Help with history (Score:4, Informative)
Because Microsoft didn't invent the Internet. As a matter of fact they were very late to the game.
MOSAIC was first, then Mozilla/Netscape. Microsoft realized very late that the Internet was going to be important and threw something together.
The standards had already been well under way by the time Microsoft got into the game.
Re:Help with history (Score:5, Informative)
You're wrong. When web standards started, MS had 0% of the market share. Internet Explorer did not yet exist. The standards were there first; MS decided not to support them.
Re:Forget performance (Score:2, Informative)
Re:Help with history (Score:3, Informative)
Technically they did what they always do. Microsoft bought out another company's browser (spyglass I think it was) and redbadged as their own.
[John]
Re:Meta moderation (Score:2, Informative)
Effectively the new system is NOT meta-moderation. I have been asked to meta-mod comments that have no moderation on them. Instead of evaluating a moderation, I am asked to simply comment on the value of the comment itself. I fail to see how the new system actually catches abusive mods.
Re:Help with history (Score:3, Informative)
Nonsense. Standards bodies can codify existing practice as standards or for reference, but can equally define good practice by creating standards based on some specific, well-defined notions of what "good" is, which more often than not do not match existing practice.
Re:Performance gap but not Conformance gap (Score:3, Informative)
http://www.earlham.edu/~peters/writing/infapp.htm [earlham.edu]
The cardinality of a finite line segment, a line, and a plane are all the same.
Theorem 21. The set of all points in an infinite plane has the same cardinality as the set of all points in a finite line segment, namely, c.
# Proof. Think of the plane as marked off into an infinite number of square cells, like graph paper. First we show that there will be denumerably many, or À0, such square cells. Pick one cell arbitrarily, and number it 0. Go to the cell above it and number that cell 1. Go one cell to the right and number it 2. Continue in this way to circle the "0" cell. The result will be a spiral that would eventually cover the plane. Yet each cell contains a natural number. Hence the cells and the natural numbers can be put into one-to-one correspondence. Second we note that each cell contains c points, under Theorem 18. Therefore, the number of points in the infinite plane is the number of cells, À0, times the number of points in a cell, c (by Theorem 18), which we know is equal to c (by Theorem 15).
Re:IE (Score:3, Informative)
Chrome doesn't seem to have a problem doing this (on Windows).
Re:Performance gap but not Conformance gap (Score:5, Informative)
"Acid3 [acidtests.org] is the third in a series of test pages written to help browser vendors ensure proper support for web standards in their products.
Acid3 is primarily testing specifications for “Web 2.0 dynamic Web applications. Also there
are some visual rendering tests, including webfonts. Here is the list of specifications tested:
Here's an abbreviated history (Score:5, Informative)
You're wrong. MS was a huge supporter of web standards back in the mid to late nineties, back when they were the underdog browser. They were extremely active in the development of XML, HTML4, DOM, and CSS. They proposed and implemented VML, which was combined with PGML to produce SVG. They were the first to begin implementations of numerous standards, including DOM, CSS and SMIL. That's a big part of why Microsoft won the first browser war; because they had a genuinely superior product to Netscape.
In 1997 Netscape started development on Gecko, in an attempt to leapfrog Microsoft's Trident engine. The problem is that Netscape couldn't get a product to market in a reasonable amount of time. Without a competitor, Microsoft took over the market, peaking at 95% share in 2003. The die was cast in 2000, however, when Microsoft saw that they'd won browser war. That's when they started moving IE into maintenance, and migrating the top developers over to .NET. This left the web stagnating for years with partially implemented standards and no viable competitor to IE.
Fast forward to late 2004, and Mozilla finally had a polished product built on Netscape's Gecko engine. Firefox emerged as a genuinely superior product to IE, and Mozilla relentlessly proclaimed the web standards mantra. They chipped away at Microsoft's market share until Firefox reached around 10% at the end of 2005. Meanwhile, companies like Google provided really compelling services based on the web standards supported by Firefox, and eventually other browsers. And of course, there were all the security fumbles with IE, while the competing browsers were (mostly undeservedly) considered safer. At that point, Microsoft finally got worried and pulled IE out of maintenance in early 2006.
So, now IE is back in active development, and MS is returning to the features they started roughly a decade ago, which places them well behind competitors like Firefox, Chrome, Safari, and Opera. And Microsoft still doesn't consider IE to be a very important product, because the team today is just a shadow of what they were at their peak in the nineties. That's why the improvements are progressing so slowly, and they're continuing to lag even farther behind the competition. Meanwhile they're hemorrhaging market share at a rate of about 7% per year.
TL;DR: MS cared about standards until they were on top; once they owned the browser market, they did nothing to improve it. Now that they're losing the market, they're making a half-hearted attempt to compete again.
Re:JS performance (Score:5, Informative)
> Two things surprised me here. One is that Chrome and Safari are 3x faster than FF
There are a few things going on here:
1) The public sunspider benchmark has a bug in that it uses a Spidermonkey-specific
extension in one of the tests that slows it down in Firefox only. Apple has fixes the
bug in their revision control system but is refusing to push the fix out to the public
site.
2) Chrome and Safari are in fact faster on sunspider than Firefox. Firefox is up to 5x
faster on other JS benchmarks. Depending on exactly what you're doing, you might have
better performance with one or the other.
Re:Forget performance (Score:3, Informative)
Are there particular sites that trigger this? If so, can you please file a bug and cc ":bz" on it?
Re:Performance gap but not Conformance gap (Score:3, Informative)
IE8 has pushed to a number of boxes where I hid the update and said, don't show again. Microsoft keeps pushing it over and over again as an automatic update. IE8 defaults to Bing. Many users don't know how to, nor care to change defaults.
Re:Help with history (Score:3, Informative)
+1 insightful. I went and looked it up on wikipedia.
The World Wide Web Consortium (W3C) was founded in 1994. Microsoft's Internet Exploder was not released until a year later, and then it went hog-wild to ignore the W3C standards. (In fairness, so too did Netscape Navigator with adding new extensions to HTML.)
Re:And this will make a difference because? (Score:3, Informative)
> it seems to me that the actual rendering of the browser window isn't the bottleneck
It really depends. It's a bottleneck for scrolling. It's a bottleneck any time interesting graphical effects (transforms, opacity, svg, etc) are being used. In Gecko's case, it's not uncommon to have the actual painting taking 30+% of the user-perceived time. From my profiling of webkit nightlies, the numbers are similar there. Things are even worse for video (e.g. for full-screen video color-space conversion is one of the main bottlenecks!).
I believe all the browser vendors are looking at making serious use of hardware-accelerated rendering at this point; it's the only way to get acceptable performance on some of the graphical effects people are using more and more.
Re:IE (Score:1, Informative)
Processes on windows (as far as as 3.1) could have more then one window, or no windows, or whatever. The feature isn't unique to Apple OSes.
Re:Sub Pixel rendering, really? (Score:2, Informative)
That they want to make the browser faster, and can do so using D2D over GDI? It seems pretty common to me for a Vista / 7 desktop to have better "gaming graphics" processing than "desktop," if you go by the Windows experience index, or whatever its called.
Re:Help with history (Score:5, Informative)
Microsoft licensed the NCSA/spyglass MOSAIC which was the dominant browser at that time (1993-94).
Then Microsoft got sued for giving-away the browser for free and thus not making royalty payments to NCSA/Spyglass (no sales==no profit sharing). Microsoft used its economic muscle to force Spyglass to accept 8 million dollars in one-time payment, and kept the code for themselves.
Embrace. Extend. Extinguish. "Business is war." - Jack Tramel
Re:Help with history (Score:3, Informative)
>>>Actually, the W3C and IE appeared almost contemporaneously with each other
False. W3C == 1994. IE == 1995. There were standards put forth by the W3C, but both Microsoft and Netscape were ignoring them (and being criticized as well). I remember it well.
Re:Here's an abbreviated history (Score:4, Informative)
>>>MS was a huge supporter of web standards back in the mid to late nineties, back when they were the underdog browser.
Not true. W3C has been criticizing Microsoft since day 1 for not following their recommendations. (They also criticized Netscape.)
.
>>>That's a big part of why Microsoft won the first browser war; because they had a genuinely superior product to Netscape.
I don't agree, but even if we assume IE was better, the MAIN reason it "won" was because IE was free and Netscape cost $30 at the time (I remember; I paid to get the shiny new Navigator 3 in a box). Free almost always wins in a battle.
Re:IE (Score:5, Informative)
When we talk about process creation being expensive, as opposed to thread creation, we're usually talking about it taking milliseconds rather than microseconds. From the perspective of the computer, process creation is expensive, and that means we can't use software design which relies on rapidly creating new processes, but if we're talking about the creation of a SINGLE process to service a new tab, it's absofuckinglutely irrelevant. From a user perspective, 1ms might as well be 1us. They both fall into the 'imperceptibly short' bin.
Re:IE (Score:3, Informative)
I use Chrome. It's an excellent browser that managed to easily sway me away from Firefox despite the fact that I use some FF extensions on a regular basis.
I have read all about the process per tab design of chrome but I must say that 95% of the times when Chrome crushes, it takes down the whole browser.
Re:Sub Pixel rendering, really? (Score:3, Informative)
I can understand that. Why would business want to deal with the mess that is Vista? Most likely they'll just hop directly to Win 7.
If corporate america was smart they'd also get rid of IE. I just ran Spybot Search & Destroy yesterday. Every piece of malware it found was connected to Internet Explorer. Nothing was tied to Firefox or Opera. IE is like an open door.
Re:IE (Score:3, Informative)