Internet Explorer Performance Lab: reliably measuring browser performance

re: Internet Explorer Performance Lab: reliably measuring browser performance

Drewfus — Thu, 23 Feb 2012 14:00:47 GMT

Not much to argue with in this post. Thanks for all this info.

"To add hardware breadth, we have additional machine pools that run the spectrum of consumer scenarios. Good performance on these machines ensures that IE uses the underlying hardware effectively across the PC ecosystem."

No AMD or Via? Core 2 Duo seems over-represented. Shouldn't the testing be biased to current generation chips, which are more representative of the CPUs that IE10 will be used on? Bias should also be towards low-end chips, especially mobile CPUs where battery life is critical. High-end CPU browser performance is relatively a non-issue.

"Loading web content is also the only category that touches most of the browser’s eleven subsystems."

The IE Blog article linked to is fascinating. So much time dedicated to Javascript! So browsing might be much faster with scripting completely disabled, but this would kill the functionality of too many sites. Perhaps a really quick way of enabling Javascript for a site - both temporarily and/or permanently - might be valuable?

"Synthetic benchmarks – Rarely forgotten but often overstated are synthetic benchmarks like WebKit SunSpider. Benchmarks can be a useful engineering tool as they are designed to stress individual browser subsystems and accentuate differences between browsers. However, in order to maximize those differences, benchmarks often resort to atypical usage patterns or edge cases."

This made me wonder what the *least* synthetic benchmark might be. Perhaps comparative ecommerce sales per time-on-site or time-on-product-page? Can't find a good link but i believe web design researchers like Jakob Neilsen have found a strong inverse correlation between page loading times and web sales. If that is true, faster browsers might show improved ecommerce results. If not true, then browser speeds are not the bottleneck, and at least in regard to online sales outcomes, would mean that all the major browsers are essentially identical in speed.

"Idle tasks are a way for Windows and other developers to schedule non-critical work to happen at a later time when the user is not competing for resources. OS idle tasks include prefetching or SuperFetching, disk defragmentation, updating search indexes, and others, depending on OS version and configured services. To ensure that no idle work is done during the tests, the idle task queue is flushed."

Presumably this is done via the command '%windir%\system32\rundll32.exe advapi32.dll,ProcessIdleTasks'. Anyway, this section got me thinking. Firstly, would running this command regularly myself stop the near-constant harddrive chatter that some non-SSD machines seem to experience? If yes, then maybe giving users an easily accessible method of running this command (Computer right-click menu?) would be a nice touch? (The idea being to run the command when stepping away from the PC for several minutes.)

Secondly, the IE cache does not as far as i know, have any predictive type functionality like that of Prefetch and Superfetch. The only stuff in the cache is from sites the user has already visited. Could this be improved on by speculative behavior that was determined by the user's browsing history, combined with the site ranks of related sites? More specifically, the files within the Temporary Internet Files folder might not only relate to already browsed sites, but would also include content from the worlds most popular sites, and sites that were both popular and had strong browsing association to the the user's Favorites and History, but had *not* necessarily been visited by the user. In this scenario, the IE cache might talk to Bing, and download web content as a background task. In a sense, Bing would then become a quasi-proxy server for all IE users.

I hope MSFT UI developers have a good understanding of the subjective aspects of UI/UX performance. This excellent post covers UX time scales. www.useit.com/.../timeframes.html It would be fair to note that the move from the Start Menu to Start screen involves moving from the 0.1 to 1.0 second bracket - a subjectively large retrograde step from a UI responsiveness PoV. This will surely require attention prior to the release candidate stage. More generally, i wonder if these subjective power of 10 timeframes are used as benchmarks within Windows development? Apparently, moves across these boundaries are much move significant for a user than moves with the boundaries.

"Many third-party applications depend on Trident, the network stack, and other IE components. Extensions like BHOs and toolbars load within the IE context. Other applications, like security software, can inject themselves between IE components. These applications become part of the IE stack, and can lead to poor performance."

Yuk. I support the move away from extensions and add-ons.

Re the remarks @Paul Coddington, the readability issues of maximized and fullscreen browser windows should be given serious consideration. Half-screen browser windows are the way to go. Put it this way; What scenario would provide the most immersive experience - a webpage in a fullscreen browser session with excessive line length (from a readability PoV), or a half-screen session on a black background, displaying the same page with much closer to optimal line length? The second scenario might also improve battery life.

Any other post on Windows development and testing would be welcome. Having said that, i'm curious to hear more about the Metro Contracts technology.

re: Internet Explorer Performance Lab: reliably measuring browser performance

checho — Thu, 23 Feb 2012 13:02:01 GMT

better provide a video with those long posts

re: Internet Explorer Performance Lab: reliably measuring browser performance

hamakaze（from japan） — Thu, 23 Feb 2012 07:42:09 GMT

IE10 is please for 7 also support it. We ask, especially emphasizing the safety and balance the lack of speed and memory usage. The official version is very fun. Win8 and simultaneous release is a good us official version.

One other, please let me block the ads here and there. I want to thank you for thinking in your company or other companies make the add-on or something for that. Even though Firefox or Chrome, you can with IE can't hate. So don't shove wrong click, and to avoid the serious....

re: Internet Explorer Performance Lab: reliably measuring browser performance

temp — Wed, 22 Feb 2012 16:59:45 GMT

Give us what we want in a browser but not what you think is good for us (like useless features pinned websites in IE9...)

re: Internet Explorer Performance Lab: reliably measuring browser performance

temp — Wed, 22 Feb 2012 16:57:20 GMT

It is clear that most of us are complaining about IE and this is going to continue unless Microsoft decide to take a big step forward innovation for IE. And notice that only power users come on this blog. I also have the same feeling, IE is not designed anymore for power users. It still have a good market share only because it is included with the OS. But this won't last forever. I hope MS will wake up before it is too late!

re: Internet Explorer Performance Lab: reliably measuring browser performance

HELLO_WORLD — Tue, 21 Feb 2012 20:07:25 GMT

Most of what you folks are complaining about with regard to IE performance is caused by Add-Ons and security software hooks into IE. Put IE9 and MSSE on a clean Win7 install and it screams. Try IE9 on Windows Phone 7, it kills mobile Safari and Chrome.

MS is fixing the problem, it's called IE10 on ARM based devices. No more security software, no more add-ons...

Also, in regards to standards, the problem these days is often not IE9. Many times IE9 is implementing the final version of a standard while other browsers had their own implementation before the final version was settled and never modified their behavior to the final spec...

re: Internet Explorer Performance Lab: reliably measuring browser performance

BRAZIL — Tue, 21 Feb 2012 14:31:05 GMT

Best Suport to all CSS and HTML5 Atributes!!! to all!! 100% msdn.microsoft.com/.../cc351024(VS.85).aspx

re: Internet Explorer Performance Lab: reliably measuring browser performance

Mayur Prayag ( India ) — Tue, 21 Feb 2012 08:04:27 GMT

Current IE Team is not brilliant ,talented.They can't think from user's view.Steven ,Please Remove IE development team ,appoint new team like Windows 7 team or give project to Google ,they will develop IE like Chrome and later you tell world that we(Microsoft) developed "the IE".....but finally we will get better IE ....Please reply me or guaranty me that future IE will be a benchmark .....the best.....better than Chrome......

Below are IE's failure Reasons:

1) IE's UI feels very heavy when first opened.

2) IE is not simple means back button and forward button is not light.

3) IE is slow

4) IE takes more time when loading Facebook and Twitter.....!!!!!!

5) Give IE's development to "Google" and they will tell you how to develop a web browser.

6) Please,copy the chrome UI ,responsiveness , simpleness.....then and then only IE will be great otherwise whatever mini Internet,millions test you do ,it is of no use......

7) chrome feels very light and IE feels like ton of load on our head..

8) when we open any site IE ask for default browser in a very sluggish manner...

9) IE smart screen filter very slow takes much time

10) Remove status bar from IE

11) Take some development lessons from Google Team so that you can develop OS like Android and browser like Chrome....

12) Within 3 years Google showed your position in market(Browser)...

13) Microsoft is not very serious about IE...

14) Current IE 9 is like Windows Vista ,make some strict steps and give IE 10 like Windows 7,8 performance .....

15) Make proper Acid 3 tests,because when user sees Acid 3 tests after release on various sites,they give very less marks to Internet Explorer....

16) Make UI clean,again I'm giving warning otherwise within 5 years you will lost completely..

17) When we Right Click on empty space,a lot of items appear.....Remove many items(Translation etc) from right click

18) Keep less items for Right click .....we want that

19)Current IE Team is not brilliant ,talented.

20)They can't think from user's view

Most Importantly, All will agree that, " Last thing ,we all people wanted that IE must beat chrome.......when we double click chrome icon on desktop ,it opens smoothly and feel softs,smoother .....whereas IE feels like scratches on our Screen,skin etc......!!!!!!!!!!!!!!!!.."

I hope you will take a lesson from above .I'm an end user.

re: Internet Explorer Performance Lab: reliably measuring browser performance

Mayur Prayag — Tue, 21 Feb 2012 07:55:41 GMT

Below are IE's failure Reasons:

1) IE's UI feels very heavy when first opened.

2) IE is not simple means back button and forward button is not light.

3) IE is slow

4) IE takes more time when loading Facebook and Twitter.....!!!!!!

5) Give IE's development to "Google" and they will tell you how to develop a web browser.

6) Please,copy the chrome UI ,responsiveness , simpleness.....then and then only IE will be great otherwise whatever mini Internet,millions test you do ,it is of no use......

7) chrome feels very light and IE feels like ton of load on our head..

8) when we open any site IE ask for default browser in a very sluggish manner...

9) IE smart screen filter very slow takes much time

10) Remove status bar from IE

11) Take some development lessons from Google Team so that you can develop OS like Android and browser like Chrome....

12) Within 3 years Google showed your position in market(Browser)...

13) Microsoft is not very serious about IE...

14) Current IE 9 is like Windows Vista ,make some strict steps and give IE 10 like Windows 7,8 performance .....

15) Make proper Acid 3 tests,because when user sees Acid 3 tests after release on various sites,they give very less marks to Internet Explorer....

16) Make UI clean,again I'm giving warning otherwise within 5 years you will lost completely..

17) When we Right Click on empty space,a lot of items appear.....Remove many items(Translation etc) from right click

18) Keep less items for Right click .....we want that

19)Current IE Team is not brilliant ,talented.

20)They can't think from user's view

I hope you will take a lesson from above .I'm a end user.

re: Internet Explorer Performance Lab: reliably measuring browser performance

Huki — Tue, 21 Feb 2012 04:32:53 GMT

There should be a way to circumvent IE's compatibility mode. Any page without a proper DOCTYPE or X-UA-Compatible tag is forced to render in a legacy document mode. Not to mention the blacklist database which can only be turned off from the hidden menu bar. It wouldn't hurt to add one option in the advanced setting to force the use of IE-Edge document mode (considering there is even a setting to turn off automatic crash recovery).