IEBlog

Ok, I'm on board, glad to finally hear the plan. Thanks for the honesty.

Now, between IE8 Beta 2, and the RC you plan to release in Q1,2009 what bugs have been fixed?  Has IE Feedback on Connect been updated to include all the internal fixes?

In particular I have 3 questions:

1.) Has the window.resize event firing been fixed?

2.) Has the z-index -ms-opacity issue been fixed? (broke in Beta 2)

3.) Has the XP theme regression bug on select lists been fixed?

Thanks

I kinda think that performance should be addressed (in addition to the memory leaks that are occasionally found).

The multiple-tab-different-processes gets messy sometimes, and the javascript engine is falling behind. (Seriously, please do something about this; Firefox and Chrome are having a drag race on this one !!)

Though, I must admit, bug fixes for the features that will push IE8 is necessary.

(But, personally, I'm a performance kinda person, so I think that should be worked on a lot.)

So, any timeframe? Minimum requirements? Anything on the *Canvas* element or Acid3 or HTML5 ;) ? (The smiley is for that I probably know that its not gonna happen.)

Javascript performance in beta2 is pretty good; much better than IE7. What really needs work performance wise is the DOM rendering speed.

In beta2, inserting a lot of content into an existing DOM with complex layout/CSS (aka the primary Ajax use case) can freeze the browser for 4-5 seconds. Firefox/Safari/Chrome are much better at this.

Before deploying a final product, we verify compatibility with current service offerings, notify staff about the upcoming change, and establish a date and notify staff of that date.

@Chase Seibert: We'd love to get the URL of a test case so we can take a look.

@Stan: We've done a bunch of performance work post-beta2.  As Chase noted, the script engine is often not the performance bottleneck.  To build a faster app, you want to focus on the bottlenecks.  The post outlines a timeframe (Q1 for RC).  IE8 does not include the canvas element.  IE8 does include several HTML5 features (as discussed on the blog previously).  The HTML5 spec is still under construction, and we focused our energies on the more stable areas that offer the greatest value for web developers (e.g. postMessage, DOM Storage).

@Chase Seibert: We'd love to get the URL of a test case so we can take a look.

@Stan: We've done a bunch of performance work post-beta2.  As Chase noted, the script engine is often not the performance bottleneck.  To build a faster app, you want to focus on the bottlenecks.  The post outlines a timeframe (Q1 for RC).  IE8 does not include the canvas element.  IE8 does include several HTML5 features (as discussed on the blog previously).  The HTML5 spec is still under construction, and we focused our energies on the more stable areas that offer the greatest value for web developers (e.g. postMessage, DOM Storage).

...And to those of you who are wondering, yes, we've filed a bug on the case where sometimes hitting Enter in a web form text box immediately after switching windows results in the "Submit" button being triggered (hence, the double entry above).  :-)

IE8 beta2 is still far from a stable product.

1. Switching tabs is extremely slow.

2. IE will crash very frequently, when I am typing Chinese in gmail.

3. The JS engine is extremely slow. Especially by running google applications.

- The spec is 10 year old

- All other browsers handle it

- IE does XML and HTML, why not xml+html?

And what about SVG?

What kind of time frame is between the RC, and the RTM (even presuming that the RC is absolutely perfect!) This is important because we are all waiting to know what fixes are going into the RC (that's why we were expecting a Beta 3).  I'm not making any code changes to handle IE8 until that RC/(Beta 3) comes out because of all the issues still broken in IE8 Beta 2.

I don't want to spend any time making workarounds for something you plan to fix in IE8 RC (if you had an up to date public bug tracking database I would know, but we've seen that that path ends in big disappointment)

As for critical bugs that need fixes for the RC, add these to the list please.

1. Page Zoom performance in IE8 is an Epic Fail

2. SVG Loading in IE8 (w/Adobe SVG plugin) performance is an Epic Fail

3. HTMLSelect/HTMLOption element modification via CSS or JS is an Epic Fail

4. UI Customizability in IE8 is still an Epic Fail

I hope the time between RC and RTM is at LEAST 3 WEEKS so that we have time to sync up with whatever you actually plan to ship.

thanks

Dean Hachamovitch, General Manager for Internet Explorer, has posted today on the IEBlog about what’s

Quirk 1:

1) open IE8 to about:tabs

2) click the "Start InPrivate..." link to open an in private session

3) resize this window, so that only the "abo" of about:inprivate is displayed in the address bar

4) close the inprivate session (we just did this to set the default window size)

5) RE-click the "Start InPrivate..." link to open an in private session

6) Expand the window... note the address is no longer "about:inprivate", but is now just "InPrivate"

7) click in the address bar and scroll back to the left and the "about:" portion re-reveals itself

Quirk 2:

When you resize an inprivate window to very small, (reduce the addressbar to about 14px) and note that the favicon overlaps the addressbar, even though it should be "inside" the addressbar box

Quirk 3:

resizing the window frame is painfully slow if the page is zoomed in

Quirk 4:

Compatibility view icon and "go" arrow icon disappear when address bar is shrunk down to almost nothing

Quirk 5:

shrink an IE8 window to quite small, then hover over menu items.  the statusbar text overflows the zoom text in the statusbar

@steve_web: Thanks for your help in filing bugs. The connect site will be updated with all of the bug fixes, and current status on the issues when the RC build is released. Kellie

NM, don't you ever wonder why the HTML5 spec project was begun?  XHTML is a failed spec. There's no point in wasting time on it.

quirky, I really hope the IE team has better things to do than make the browser pretty when you shrink it to a tiny-weeny size.

concerned, feedback that contains the text "Epic Fail" is an "Epic fail."  You need to provide reproducible test cases in order to add value.  Blaming the IE team for the Adobe team's unsupported plugin seems like a significant waste of breath.

stanleyxu, try running without plugins.  Use the -extoff command line switch.  Be sure you have the latest version of Flash installed too.

"Reopen Last Browsing Session" does not work for me.  It is always grayed out from within the menu selection and does not appear upon opening IE8 b2 as a selection from the first new tab that is opened.  This makes it a non-starter for me.  It is a shame as the feature worked fine (opening active tabs the next time IE is started- or whatever the wording) with IE7.

Microsoft hat sich vor der Veröffentlichung einer Beta-Version des kommenden Internet Explorer 8 (IE8) nun doch (einigermaßen überraschend) dazu entschlossen, die ggü. IE6 und IE7 standardkompatiblere Darstellung von Webinhalten standardmäßig zu verwende

Yes but...

Why keep the status bar visible in kiosk mode?

Why reverting to "old tabs" in the Favorites panes?

Why resizing the (docked) favorites pane causes the entire pane to flicker badly?

Why the return of the menu and favorites bars by default?

Why remove IE7's "Add to favorites" button (move, actually, but to the wrong place)? A showstopper as far as I'm concerned.

Why do you still waste vertical screen estate with a window caption? If you include the default toolbars, the whole "header" is more than 150 pixels high, you could just as well replace the whole thing with a ribbon UI perhaps?

Why so little contrast in the new address bar, the dimmed parts of the URL are barely readable - light gray on white, does this meet accessibility standards?

Why still a half-dozen unlabelled little areas on the status bar?

Besides those small quirks and those mentioned above about the performance in some areas, it looks good :-)

@EricLaw [MSFT]

Hi, here's a rather quirky way of benchmarking specific DOM performance:

http://nontroppo.org/timer/progressive_raytracer.html

WARNING: If in an IE Browser I would NOT click the 'Full Render' as the browser will freeze up for a very long time and use an excessive amount of memory. Just do the Basic Render to compare it to other browsers.

I did this test on full render a couple of months back but the results should still hold true on my computer. My computer has an AMD 3800+ X2 with 2 GB DDR 400MHz RAM:

Chrome - 29.69 seconds

Opera 9.6RC - 31.609 seconds

Safari 3.1.2 - 38.734 seconds

Firefox Nightly- 537.907 seconds

IE8 Beta 2 - CRASH - Pass 84/120 - 2269.468 seconds

Also Memory had increased to over 900 MB by the 84th Pass for IE8 Beta 2, note that each pass was taking longer than the previous pass, so IE8 would have probably taken 2-3 hours to pass the whole thing had it not crashed.

I understand it's very very specific and a performance gain here will hardly show up in the same magnitude on normal sites. But I still think it's something worth a quick look at :)

I've been using beta 2 for sometime now and here is what I've to say.

1. JS Engine needs performance improvement. IE 8's JS Engine is much better than IE 7's though.

2. Switching between tabs is sometimes very slow.

3. For some reason Vista's DEP feature kills iexplore.exe when exiting the browser. I have no idea why, but because of this the performance monitor's index has fallen to below 2.

4. Please let the favourites bar to expand vertically (that is to support multiple lines). Now I have to go through the small arrow button (>>) and look for feeds!

You guys have done a great job in beta 2. Hope to see much improved and bug fixed IE 8 soon!

Ted, please note that HTML5 is designed to be compatible with an XML syntax (i.e., XHTML) as well as a specific HTML syntax.  While you'll probably get less argument that XHTML2 is a failure, XHTML itself (HTML in an XML form) does have appeal in a number of applications.  Most of them are not on the Web, though, for various reasons obvious and not.

@Damian,

Basic render in 2.543 seconds.

Full render in 37.768 seconds.

IE8 B2 on Vista x86 Business Ed. 4 GB Ram Intel DC 2.20 ghz, NVidia 8500 GT (512 MB)

@Ted, please use your full handle when posting. "Ted - The Epic Fail" is much easier to decipher than "Ted".

yes, XHTML is a failed spec.  Wonder why? did like one of the major browser mfgr's not get on board?... oh wait, never mind.

you can complain that quirky is being picky, fine, but he/she is highlighting that the UI is not polished or fully tested.

concerned mentioned things that were worrisome and quite frankly I agree with every statement they made.

You Love Microsoft, and that's cool, but the rest of us want a reliable stable browser to code against, and ATM, that IS NOT IE8B2!

As for SVG, supporting the adobe plugin wouldn't be an issue if IE supported SVG right out of the box.  IE7 supported the plugin just fine without issue. IE8 should be a tenfold downgrade in performance when using the exact same plugin, in fact, it should run even better.

As for comments containing the phrase Epic Fail? yeah, you think they don't belong, but according to the rules of this blog we can't use the words that REALLY describe our feelings towards this non-standards browser.

Thus all we can do is indicate in plain language how utterly broken various aspects of IE really are.

Thus "Epic Fail" seems very appropriate.

Thanks again for your trolling Ted - The Epic Fail, you never fail to show your incredibly biased view of web browser development and a total lack of understanding standards and where the web is trying to progress to.

Please troll on a forum where users actually want to hear your FUD.

SVG would be a nice feature to get into IE8, thus i think that is already late for IE8 roadmap. Real shame...

Thanks for the timeline Dean.

We now know where we stand. Roughly, at least!

PLEASE PLEASE PLEASE add a download manager, this is something a lot of people have asked for, for years! And still it doesn't reach you guys, why isn't this feature in already? What is the reason, it would be nice to hear from Microsoft why they haven't done this yet.

Now it is pretty annoying that you have to start multiple IE's to download multiple large files or otherwise you can't download them at once as 1 IE only supports 4 concurrent downloads.

I work for a fortune 50 company.  With hundreds of internal-facing web applications and somewhat fewer external-facing web applications, we will probably only take immediate action to fix any incompatibilities in the external-facing we bapplications.  For the internal apps, they would not be fixed until a project can be allocated in the next fiscal year.  It could be 2010 before everything is compatibility-tested -- there's just far too much out there to do a full regression test without diverting resources on other mission-critical projects.  For internal apps where we control the deployment of IE to desktops, it is hard to justify spending money on all the testing and fixing of all those applications until it's absolutely necessary.

Also critical to consider is the testing on embedded browser controls in desktop and mobile-deployed fat clients.

In any case, while all new development is standards-compliant, there is a major issue with all the old legacy code out there.  But that's how software development goes...

Most wanted features:

- Download manager

- Customizable (skinnable?) GUI

- Support for CSS "opacity" so proprietary filters wouldn't be needed

- More speed! Just compare a default IE with Firefox, Safari or Chrome. What will you see? #1 IE takes much longer to start up, even with Superfetch in Vista. #2 Open a new, empty tab. Again: much slower. #3-#99 Left out.

Nevertheless IE has improved hugely from IE6 over IE7 to IE8 - but that does not mean that there wasn't much left to do to be on the same level as the other browsers.

I vote for *FULL* CSS 2 compliance.  Make the web designer's job easier PLEASE!

Also how about Microsoft paying for all IE6 users to be flogged in the town square...??!?

"We listen", "we are listening", "we've heard you", and others stupid marketing sentences...

You've just heard nobody. Where's beta3 ? Beta2 was unusable and crashed all the time, so we can't test it. Please give us a testable beta before a release candidate.

@Nick : forget the download manager, they haven't listened to us. For IE7, they said they will consider the download manager for a future release, now they say the same for IE8. We'll see in IE9... or IE10...

@Thales : about SVG, you can get the same answer as Nick...

I am not sure if this is specific to my network connection. IE 8 most of times gives "Bad Gateway Error" on my XP machine. On my home PC I have Vista installed and IE 8 works fine. I have no problems with it at all.

If pseudo-elements and generated content is fixed in RC, I'll be happy; if reflow problems (reset all scrollings, PNG images disappear on DOM modification) are solved, I'll be glad.

As for SVG support, I'll simply put a link: "can't see this image in high resolution? Use a modern browser." using the 'object' fallback method.

IE is slow? That's what multi-Gigahertz quad cores are for: rendering web pages faster, so stop complaining.

We really really really need a download manager. Everybody wants it. It's the most important missing feature. I'm sure Microsoft can implement it in 1 in 2 weeks... So why is it still missing?

@ Damian

Full run finished in 67,8 seconds on Lenovo Thinkpad T2300.

980Mhz, 1GB RAM...

So there might be something wrong with your computer / browser...

@sebastien

What's so important about a download manager? If you need to resume downloads, just click on the same download link. The download will resume where you left off if possible.

Hey Now Dean,

I love IE8 beta 2. One thing that I would really like to see is a hotkey to open multiple tabs. If a user goes to the favorite cetner (alt+c) then arrows down to the folder (ctrl+enter) should open the sites in the folder but I've had no success in IE7 or IE8. When you are in the favorite center then hover over the folder with the mouse then on the right of the folder there is a little blue arrow. If this arrow is clicked (no hotkey w/ mouse I mean) then the group of tabs open good. That is a small thing but it is something I think of often.

thx 4 the info,

Catto

christophercatto@hotmailNOSPAM.com

Besides the usual feature requests, I would like to point out that I've noticed IE8B2 becoming unstable on quite a few occassions (hang/crash). It is very important that the stability is improved for the final product release. I'll gladly accept a few feature cuts if this results in improved stability.

@Disk4mat

@ajo (also are you sure you did the Full Render for that test? And it did all 120 passes)

Are you sure that the result rendered correctly and computation finished? The basic render looks like a pixlelated version of the full render which is a purply sphere.

I ran the test again on this computer and after 15 mins of running it was clear I was getting the same kind of performance results.

So I went and installed a fresh copy of IE8 on a different computer (AMD 64 single core 2.4GHz, 2 GB DDR2 667MHz). As of pass 66 (still running I'll post the final result later) it's currently taken 448015ms

Both computers run other performance benchmarks as expected, they are kept very clean of bloat ware, malware etc... IE is almost never used so it's also kept quite clean. Other people have confirmed my similarity bad performance.

thanks for detailed timeline around IE8. we like the IE8 BETA so much so that we have created a little community around it at http://www.spreadie8.com

we hope you like the community. we would be glad if you wanted to share any feedback around the same at admin@merawindows.com

@Disk4mat

@ajo

So yeah, the test on this other computer (AMD 64 single core 2.4GHz, 2 GB DDR2 667MHz) with a fresh install of IE8 Beta 2 crashed in the same place pass 84 and took 1912625 ms

The crash did not automatically recover, I had to terminate the iexplorer.exe process.

Are you sure you both got up to pass 120 of the full render AND it looked like  purple shaded sphere? I find it odd that I can run to separate tests on different computers with different OSes (XP x64 vs. XP) and get the same result and it be wrong...

I normally don't post to these things, but I'd like to make you guys aware of a bug in beta 2 that is driving me nuts...

For some reason, after using IE8 the system does not close open processes and just leaves them there, running... it's aggravating because after a while there'll be like 9 - 15 iexplore processes when IE8 will undoubtedly crash (it's fine, we all understand it's a beta)... the thing is that the application needs to be restarted... that involves me having to manually go to Task Manager and close the 15 processes, which is a bothersome and slow process, before I can open it again...

Also, if I need to install a program that needs me to close IE to continue, I also either have to restart previously or manually close those dozen processes... VERY aggravating...

@Chase Seibert

That's like saying that a Ford Edsel is a pretty good car because it's slightly faster than a Ford Model-T.

By todays standard the performance of both, IE7 and IE8 beta 2, is simply unacceptable.

The only reason why Google even bothered to create Google chrome is due to IE's horrific Javascript and DOM performance which is holding back the web as a platform.

IE is so far behind in performance and standards compatibility that I don't know why Microsoft even bothers any more.

I'm wondering if IE8 is going to fix the regression in dealing with offsetParent?  If you load this page http://dump.testsuite.org/2006/dom/style/offset/023.htm in IE7 and in IE8, IE8 will show that it failed while IE7 didn't.  IE6 passes as well.  This bug in IE8 makes it pretty difficult to support IE8.

Hi there

I know it has been like this for ages and it's probably too late to address this issue at this stage.

However, it is a pain in the ... to have the "script debugging" option enabled, when you browse 3rd party websites (I know, it may come as a surprise that I use the browser both as a developer AND as an end user :-)

I would though like to have the Script debugging option enabled when browsing/debugging my own websites - but I don't really bother about other developers script errors.

So well, what I am basicly asking for is an option to restrict script debugging to only be enabled for specifik websites (perhaps Zone based). If this is somehow already possible, I appoligize.

This has been said many times before, so I'll make it simple...

We want a Beta 3! Beta 2 was no where near the quality we expected. Before getting to an RC, we want to get the last set of bugs reports before you get to RC1.

Closing the door now would be a horrible mistake.

Please make sure IE8 ships with the following:

SVG + Canvas + CSS2.1 Full compliance. Pretty please?

My biggest complaints from IE8 Beta 2 are

-Lack of spell check

-Page zoom is ridiculously slow. This become more important if you have set Windows to a high DPI setting... IE8 automatically adjusts the page zoom to 125% which in turn is very very slow.

Also I sure hope you add the ability to drag tabs into new IE containers, ala Google Chrome.

How do I prepare for a new software release? I start by installing the beta or RC, but until it can sit alongside Visual Studio 2003 I can't do this. Have you fixed that yet?

@Mike: Spellcheck is available from at least 2 add-ons (IESpell and IE7Pro).  Performance issues with Beta-2's Zoom feature are known and we've been working on performance of Zoom (and across the browser).  I'm not sure what you mean by "new IE containers", but if you hit CTRL+N, the current tab is opened in a new browser instance.

@Jason Ashdown: As mentioned previously, our goals for IE8 include full CSS 2.1 compliance, which we aim to deliver.  We do not plan to natively support SVG or the proposed Canvas tag in IE8.

@Daniel: We've received quite a bit of feedback about the script debugging prompts, although most of it is from non-developer users who simply want an easier mechanism to turn it off.  You can, of course, reenable script debugging using the dev tools (F12). Stay tuned for the RC.

@Trevan: Thanks for the test case.

@Gabriel Golcher: Do you see the same problem when IE is run without add-ons?  Beta-1 had some serious problems with process management, but most were ironed out before Beta-2.  I haven't seen any issues on current builds.

@AccessDenied: Fear not, reliability and performance both remain key areas of investment on the part of the IE team.

@Catto: You can also middle-click the group to open them all.

@sebastien: There are various 3rd party download manager addons available for IE if you haven't tried them.

Unfortunately, integrating a proper download manager than handles all of the myriad ways that file downloads take place takes a lot longer than you might think.  As syb notes, download resumption has been improved in IE.  

@Jagannath: Are the Tools / Internet Options / Connections proxy settings the same on both computers?

@Dave: As noted in various places, our target is full CSS2.1 compliance.

@mynetx: You should take a look at what browser add-ons you have installed.  The new Tools / Manage Add-ons UI will show the load time for each of your enabled addons.  

On a few machines I've tested, both IE and Chrome start almost instantly, and Safari and Firefox take several additional seconds.  Slowness in creating new tabs is caused almost exclusively by slow add-ons.

@Nick: IE8 supports 6 simultaneous connections per host.  Adding more connections usually results in a slower overall experience.  But, if you want more, it's trivial to increase the limit.  See www.enhanceie.com/ie/tweaks.asp in the "Speed tweaks" section.

@Hammad: If IE experiences a DEP crash when closing the browser, this is a signal that you have a buggy add-on that does not shutdown properly.  When IE destroys the add-on, the add-on attempts to access already-freed memory and crashes.  Remove (or update) the addon and this problem should go away.

@Damian Shaw: I hope you'll agree that creating 59000 DIV elements isn't really a good benchmark for overall DOM performance.  :-)  The problem with optimizing for contrived benchmarks is that the benchmarks often don't map to real-world performance problems.  As noted, we focus our performance investments on real-world sites (e.g., GMail) to isolate and remove bottlenecks.

@Howie: Have you configured IE to delete browsing history on exit?  That's one reason that "Reopen last session" could be unavailable.

I do not care how you do it. Just get this browser out and done well. Make it give me tools I need that I cannot get from Mozilla, Opera or any other browser. If you can not, another browser will. Your job is to make my life easier, not yours! Andale!

Wasn't this product meant to ship end of year *this* year? Just checking...

Dissapointed that there isn't going to be a beta 3 soon.  Beta 2 isn't nice to use for more than a few minutes, and I don't think you are going to get much more useful feedback on it now, just the same bugs repeating themselves.  Then by the time we get the next public release, you will only be acting on a small number of critical issues as you said, so the opportunity to fix other issues will have been missed.

Not sure if someone has mentioned this or not (because I'm not familiar with the technical reasons for rendering issues), but I have noticed that on one website I help maintain, a logo in the header is consistently not showing until I refresh the pages. On another website I have dealt with, random items on the page do not show up and upon refreshing, they might show up, but others disappear. I have disabled all add-ons but no change. Has anyone else experienced this?

Le blog de MSDN l'a annoncé hier : la sortie de la version finale d'IE 8 sera retardée, puisqu'elle sera précédée d'une version RC au cours du premier trimestre 2009.

I've kept using IE8B2, despite the many flaws.  I am glad to see you recognize that it wasn't close to done, but am disappointed that we won't see another update before Q1.  What about making somehting available via MSDN, since folks paying for that are likely all professional developers?

One thing I have is that the debugging just doesn't work.  I expect that it will be as easy to use or better than Firebug on Mozilla, *without* needing to buy or use some VS tool.  Getting some update on this would be nice.

I've found sesison restore to be flaky and unreliable (yes I know it is btea, but this code is aplha at best).  Please publish where the session files are kept and format.

When there is some content blocked, I get an info bar, but it doesn;t tell me what the issue was.  Please provide details so advanced users can see what the issue is.

I'd expect that MS's QA team and employees will try the browser on the top 1000 web sites and will make sure it work on those sites.  You have enough staff at the company to do this and your user base deserves this.

B2 was a lot better than B1 and so I am looking forward to the RCs and final.

I guess they took my strident feedback about Beta 2 seriously: IE 8 pretty much breaks the Web at this

Javascript performance. It's no good comparing yourself against IE7 and claim averything is great and ok.

In our organization we are starting to move away from IE to other browser choices specifically for standards compliance AND javascript performance

I know you are capable of giving us killer Javascript performance but are reluctant to compete with other MS technologies.

--jw

I just ran the basic render test in 60.125 seconds. This is on a plain P4 2.6Ghz with 2mb of PC400 RAM and using the motherboard Intel graphics. Running XP SP3, IE 8 B2. The test ran to completion.

Try running in no add ons and retry your test.

Cookies. Some sites logout, I don't know why.Also when XP crashes for any reason, it seems that IE deletes cookies and Temp files I suppose for security, but thats very anoying as I preserve my cookies

Images. Some sites are unable to save images in the IE cache :

http://agaudi.wordpress.com/2008/08/11/sawdust-diseno-grafico-e-ilustracion/

"save Picture as" tries to save it as bmp

Tabs. MMB opens a new tab in the extreme right of the group. it should open it aside the current page, still inside the group

refresh button, it sometimes does not work, using the address bar and pressing ENTER is usually better

Favorites bar. is it possible to make it like a real menu instead of buttons. and a shorcut to hide/show easily

Restore last session. It needs more depth. and be independent of IE cache, because deleting the history makes it obsolete right?

Popup blocker. Uses Ctrl+Alt to override, but interfieres with IE menu

adress bar.sometimes typing a single word attempts to load it as as url, instead of search

Scroll Bars. when the page is loading it looses focus when you are scrolling, I cannot wait for pages to load 100% its a waste of time, can you make it more independent?

Animated Gifs, etc. When running IE slow down a lot, for example links, etc

Saving Images. IE does not remember the last directory correctly, some tabs do, other tabs suddently remember a different directory.

My Pictures. Always opens as thumbnails,anoying!!

I suppose plugins are not independent, that makes tabs slow, but IE8 is much better than IE7 but still very beta, even google needs to run in compatibility mode

Bill Veghte earlier this year said IE8 was coming out by 12/31/2008.  So much for that promise.  I guess Microsoft blew through that deadline like a bulldozer running a picket fence!

ok so today microsoft said ie 8 well come out in 2009 after the rcs version well what happends to the ppl that have beta 2 installed on there pcs will we get a update from windows update.....

aww ie 8 is so much better then ie 7 the loading times r much better its faster and i have no problems with it......

Wow, awesome!

IE8 Beta 2 is already a major improvement from IE8 Beta 1, can't wait till the RC is out :)

Great Job!

Seems there's a bug in the favorites menu, bookmarks (and subfolders) with ampersands doesn't display correctly.

Back in July, Microsoft indicated that there would be one more beta of Internet Explorer 8 and that the final version would ship before the end of 2008. Beta 2 was duly released in August, but yesterday, Microsoft’s Dean Hachamovitch revealed tha..

@sialivi: I can't repro the ampersand issue in the current build.  Can you provide exact repro steps?

@Lori: What's the repro URL?

@PCause: Can you please be more specific about what problems you've encountered when debugging?

Simple list of needed fixes.

1) Fix Java Script

2) Make sure Automated Crash Recovery works all the time.  Also, you could show an icon showing that IE is trying to recover.

3)More web standardization.  You guys need a higher score on the ACID 3 test.  If you could pass it, no one could doubt that Internet Explorer is awesome.

4)Just an Idea.  Take a look at IE7 Pro.  Those features make IE great.  If Microsoft were to show that site a few grand.  IE would take on a whole slew of great features.

is there any hope for a more reliable and usable favorites panel?

problems:

- sometimes it simply rearrange the custom order alphabetically (after a crush or some ie update)

- after some time, wrong favicons are displayed (it simply mix them) , only workaround is to delete the temp files

- bookmark import is also lost the custom order

- during rearranging with d&d the scrolling is extremely slow (any other win app would speed up scrolling according to the cursor distance from the panel)

- positioning the links with d&d is very hard on vista (on xp there is a thin line showing the current cursor position)

- there is no support for d&d from tab to favorites (it is just more handy than using the addressbar icon)

more:

- the addressbar should keep the newly typed(but not yet submitted) text during tab switch

- there should be a close icon on the last tab too, it should change the tab to about:blank or about:Tabs or to the first homepage (an option for setting this would be nice)

I wait for the final release and then i will compare the IE 8 with my other browsers.

Side by side install of different versions. Having to run IE6 on a virtual PC is a really annoying.

@concerned

I agree that the lack of a customizable UI is a problem, but don't hold your breath waiting for one.  I was just browsing some of the old IE blog and found this:

># re: Security strategy for IE7: Beta 1

>overview, Beta 2 preview

>Thursday, August 04, 2005 11:27 AM by redxii

>Is there any chance Beta 2 or final will >allow total customization of every toolbar >position? I myself prefer the File menu below >the title bar.

... so MS has known about this for over 3 years, when IE7 was in Beta 1, and now they tell us it's not happening in IE8 (well they are throwing us a bone by letting us move 2... count 'em 2... buttons).  

Now let's take a look back at 2005...

... Practically no one had ever heard of Barack Obama

... Seattle was about to start a Super Bowl season

... and IE had 87% market share

Times were really good

Since there will be no canvas support in IE8 I would like to see the IE team working with the ExCanvas guys to make sure ExCanvas will function properly in IE8 standards mode.

https://connect.microsoft.com/IE/feedback/ViewFeedback.aspx?FeedbackID=364096

On another topic, it would be nice if ticket comments were not signed anonymously, "Best regards, The IE Team", but with the actual name of the team member writing the comment.

It would be nice if tickets were closed as they are resolved, not waiting until the date of the release itself.

I would like to see nightly releases.

As someone else mentioned above, I also think a beta3 would be a good idea. I am very much concerned about VML continuing to function properly until canvas is supported in a future version of IE.

@Trevan

That's because there is an implicit <tbody> (you can see it with the Developer tools). Therefore, when applying the algorithms given at

www.w3.org/TR/cssom-view/#offset-attributes  ,

then the nearest static-positioned ancestor of <tr id="t"> is the implicit <tbody> and not <table>.

Obviously the algorithm given at www.w3.org/TR/cssom-view/#offset-attributes will have to be tweaked and cover the case of implicit and explicit declaration of <tbody>.

What should happen when

<tbody> <tr id="t">

is explicit?

Regards, Gérard

With multiple tabs open the task manager gets loaded with "Internet Explorer" descriptions/processes. Unfortunatly since tabs in IE crashes surprisingly quite often it would be nice to know which tab I need to "end process" on to get rid of the non-responsive tab since you can't close it from the browser itself.

Knowing which tabbed process is which would help from having to pick and guess to hope you get the right one.

Other than that I'm pretty happy. Stability seems to be the biggest issue for my IE8b2

In IE8 Beta 2 the default rendering mode for WebBrowser Control is IE7. To force IE8 rendering mode the HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Internet Explorer\Main\FeatureControl\FEATURE_BROWSER_EMULATION must be set to 8. Will this behavior stay the same in the final IE 8 release? Are you going to change this in the future?

@MSFT - What is going to happen from this point forward to solve the lack of transparency with IE.

1.) The public bug tracking site sucks, and you'll likely shut it down again when IE8 goes final, once again making the whole exercise pointless.

2.) The feature roadmap is non-existent for any release.

3.) The milestones are not identified up front.

4.) No information about bugs being fixed is released until the next public build goes out.

5.) Developers got 2 betas to test/develop against, but both have such major regression bugs that it is almost worthless even trying, and now you tell us that there won't be a Beta 3 (the one we've all been waiting for)

6.) EricLaw is almost the only MSFT team member that replies to comments on this blog. Every other MSFT IE team member needs to follow his lead

7.) Monthly IE chats (although much appreciated) fail horribly because they always happen at the same day/time. In case you didn't notice, the developer audience is GLOBAL.  The question box doesn't allow pasting, and the question size is limited to anything that would be useful. This would all be fine but every question is responded to with a lame blanket statement to the effect of: "thats an interesting idea, but at this point we have locked down...   ...thanks for your feedback"

8.) Commit now! Is SVG on the roadmap for IE9 or not?

9.) Commit now! Is CANVAS on the roadmap for IE9 or not?

10.) Commit now! Are proper EventListners going to be handled in IE9?

11.) Commit now! Will all the HTML form elements be fixed in IE9? (checkbox/radios firing onchange properly, not hanging the browser when clicking them, select elements supporting events on the options, or say innerHTML? or say styles? File uploads rendering a file selection box from this decade, etc.

12.) Commit now! Tell us if MSFT plans to have a proper public bug tracking database that is updated AS THE BUGS ARE FIXED

13.) Commit now! Is IE9 going to have a customizable UI? or are we in for a new pile of the same crud?

I am curious if the IE team has any changes to their plan on bug fixing the rendering engine, in case of any. In IE6 and IE7 I see MSFT not going to touch the rendering engine after release, leaving all security unrelated rendering flaws and bugs as they are until the next major release to avoid that so-called "break BC" which already gave enough countless pain to web developers.

IE8 is claimed to commit to standards AFAIK. I really think MSFT should evaluate the possibility to continue fixing rendering bugs that are not comply to standard even after the product goes gold.

the auto recovery may ask the user to recover page or not before it tries to recover.

A typical website: crashie.com, which contains some html that can corrupt/full load IE(any version, in XP but not vista).

Problem is if i open the website, and find IE not responding, I may close it through task manager (well, another problem, it may be difficult for me to identify those iexplore process belongs to main UI(parent?) or tab). If I correctly "End Process" of that mal function tab, in general IE will recover it without enquire me. So the problem will loop back, IE hangs again.

I agree with the others that are asking for a beta 3. Because of the zoom issues (repainting takes multiple seconds when zoom isn't 100%, and zoom is constantly being reset to 125% in 120 DPI mode when I want 100%), I cannot use IE8 as my default browser. I am pretty sure that anybody using high DPI mode feels the same way. Further, it seems like the developers and MS's testers are not using IE8 much in high DPI mode either--if they were, these problems never would have made it to beta 2.

I applaud your attempt to reach full CSS 2.1 compliance. But, it will be almost impossible to get there with the current plan. Are you committed to fixing all CSS 2.1 bugs that are reported after RC? I doubt you will be able to, because some issues will likely need risky fixes. You really need a beta release that you believe is 100% CSS 2.1 compliant, stable, is stable, and has usable performance *before* RC in order to reach that goal.

Finally, I just reported the "what is the current tab" usability regression. Every time I use beta 2 I feel the productivity loss from this problem, and the issue has been reported repeatedly by others. All reports so far have been dismissed as "won't fix". What is the logic in that? Why spend so much time improving usability and productivity in other areas without fixing the regression here?

I *am* very happy with beta 2's CSS 2.1 capabilities. IE8's CSS 2.1 support is notably better than Firefox's CSS 2.1 support.

BTW, what happens when somebody finds a CSS 2.1 compliance bug after RTM? Will there be hotfixes or service pack updates that fix compliance issues?

Please add at least a "No to all" in addition to the "Yes" and "No" choices in the ActiveX / active content prompt dialog. Please also add information on the particular active content in question, such as its name and its certificate, etc. I am referring to the dialogs that are used when IE is set to ask for permission for every single control, such as when going to say, Youtube to watch a video and IE would ask if it's okay to allow ActiveX to run.

Another issue is the same kind of prompts showing up on top of pages that they do not originate from. Say you're browsing a news site and you open one of its articles in a new tab. Chances are that the ActiveX controls in the new tab will show up while you're viewing the first page. This is confusing, even for pages that are grouped / related.

A third issue, also with ActiveX or active content is when IE seem to want to run active content from say, a news site, when I have just navigated away from the news site and onto a pure HTML / CSS page on my hard drive that has got no risky content what so ever. I'm getting this information bar at the top of the page telling me my page wants to run active content but was stopped. Doesn't make sense. This issue is particularily worrying if I want to sell similar minimalistic pages to customers, it would turn into a support problem as this is too complicated for most customers to have to learn when it should not be necessary.

I have mentioned all three issues before. I thought at least the last one was fixed now, but it turned out it was not.

Other than that, I'd like to say that I really like IE 8 and the improved standardization. Keep up the good work!

IE 8 beta 2.

@Damian

For certain I get all 120 passes. It slows down after pass 90 but still completes. Just tried it again and finished full render in 41.46 seconds.

It may be that on your 2 systems you have 3rd party apps/security/filters that are affecting IE's rendering.

Here is a screenshot I snagged at pass 115. Showing open tabs and current progress for pass 115. http://img81.imageshack.us/img81/219/ie8b2rendervw6.jpg

my main issue is the lack of a simple ingerated download manager, spell checker and if i am not mistaken, there is no way to view saved passwords.

all of the above are simple but effective features from firefox.

@Sialivi: Issue confirmed.

@EricLaw: To reproduce create a new folder or bookmark with a single & symbol in the name. Example: MS Windows & Vista

From the favorites menu the bookmark (or folder) will display an under score. So MS Windows & Vista would display as: MS Windows _Vista

The work around is to use a double &&. Example: MS Windows && Vista

Screenshot: http://img509.imageshack.us/img509/1557/ampersandvi9.png

Note: This bug is also present on the favorites bar when selecting and item that has child items (folder on fav bar that contains bookmarks)

Make Launch Fast...

for e.g. Windows Media Player Light version in windows & beta opens at blazing speed.

Normall web surfer has minimum 10- 15 tabs open. Under this, it should perform best. It is still sluggish like vista taking more time in opening windows explorer.

Windows 7 did great job in speed..and responsiveness..users expect same from IE 8

Finally, keep it light, fast and responsive....

@EricLaw [MSFT]

I totally agree creating 59000 DIV elements isn't a good general benchmark for DOM performance and I was very careful about never stating such :-).

Still, it does highlight bugs with the IE layout engine and it also impedes how creative web developers can be. No one will ever do anything similar for a web page, not necessarily because it's not a good idea but simply because IE can't handle it.

@Disk4mat

Thank you very much for the information, I simply can't explain why it works well on some but not on others. The only particular difference I notice is the use of Intel CPUs vs. AMD CPUs. I'd find it odd that IE work so radically different on them though.

(P.S I'm 100% sure I have no interfering security applications or anything which is explicitly plugged in to IE other than Flash and Java)

Regarding XHTML being a failed spec, it really isn't.  Just because IE doesn't parse the stuff, that doesn't mean that it is a failed spec.  Many people are doing things like switching to Linux, which helps Firefox's case mostly or Mac, which helps Safari.  In addition, even people who are sticking with Windows are switching.  I'm not saying it is a mass movement, but there are certainly fewer IE users than there were a year ago.

As for the status of XHTML 2.0, it appears to be picking up the pace again.  They're beginning to figure out what to do with the language by figuring out exactly how certain features should be decided.  In my opinion, XHTML 2.0 would be quite beneficial to developers, but I'm only one person...

Even if XHTML 2.0 doesn't get implemented by any user-agents, there will always be XSLT...  ^_^

About the download manager : clicking on the same url to resume a download is not always possible. Actually most of the sites I use have automatic mirrors. And sometimes, downloads are corrupted, so you may want to have a donwload manager which lists URLs you already downloaded.

I tried several 3rd party download managers, none works 100% for all the links.

Well, I will still use IE but it's weird to add gadgets while a basic functionality is still several years late

Full CSS2.1 compliance is not enough to gain the respect of web developers and designers, it's important to fix as many rendering bugs as possible before final release. Otherwise we'll be worse off than we are now.

Spend some time on this page [1], and fix every bug you can, much like you did with PiE.net for IE7 [2].

1. http://www.gtalbot.org/BrowserBugsSection/MSIE8Bugs/

2. http://blogs.msdn.com/ie/archive/2006/08/22/712830.aspx

EricLaw: Thank you for the prompt reply, looking forward to see what you have planned for the RC.

Noticed a small bug where text is able to wrap inside an <input type="submit"> button.

FireFox: http://temp.frenz.dk/dev/ff.jpg

IE8: http://temp.frenz.dk/dev/ie8.jpg

Just to let you know, if you wasn't aware of it :)

Top News Stories SharePoint Migration in the Hands of the Content Owner? (ITWorld) Let Business Users

Suggestion: IE8 drop down menu are too cluttered and some function have too many launch surface (Accelerator can be accessed via 4 places)

I have uploaded my suggestion to Scribd

http://www.scribd.com/doc/8247186/Internet-Explorer-Menu-Bar-Improvements

hope you will concern

Internet Explorer 8 Release Candidate erscheint Anfang 2009

Please stop fixing your terrible layout engine and just use Webkit or Gecko.  You make web developer's life a nightmare on a regular basis.

Something that would be really helpful to me as a designer/developer would be to have a split view browser for tabs with two widescreen monitors.

In otherwords like in Visual studio you can have File one next to File Two as a splitscreen.

With Tabs already an option in IE, being able to see 2 or even more would be super slick. Instead of having to open up two different IE browsers.

In Win7 you go half way with the idea of wanting a split screen where you can quickly set two apps to split the screen width, now just bring it inside the browser.

@EricLaw: I have a repo for the performance issue. Where would you like me to send it?I already submitted it to the "Email" link on this blog.

@Daniel: Can you provide your input type=submit testcase?  Those look like checkboxes to me.  Which spec specifies the wrap behavior for such tags, HTML4.01?

@W: A significant number of the "bugs" on the page you cited are not actually bugs.  Unfortunately, corrections or feedback on that page have not been accepted.  Bugs in IE should be reported through Connect for proper tracking and analysis.

@Damian: I don't understand why you think there's a "Bug" in the layout engine, as the page in question works perfectly.

@Disk4mat: Thanks for the clarification.  Our test team reports that the ampersand issue was fixed between Beta-2 and current builds.

@Steinar: Very very few users have altered their Internet Explorer configuration to introduce prompting for all use of ActiveX and/or script.  Such prompting does get tiresome, which is why it's disabled by default.

Please keep in mind that "Active Content" includes any binary behaviors (such as Filters) in addition to Javascript, VBScript, ActiveX, and CSS Expressions.  

@Brian: On the contrary, MS is testing high DPI extensively.  Quite a few issues have been fixed so far.  We encourage you to file appropriate bugs on the issues that you've noticed.  

@wai: When IE recovers, it recovers each tab into its own process.  It also stops attempting to recover a given tab after 2 failures in a row.  

@Oliver: The IE team is hard at work delivering IE8.  IE9 planning will begin in earnest AFTER IE8 is delivered; hence, no, no one from Microsoft is going to speculate on the final plan.  For cases where you believe you've found a bug (e.g. #11), please be sure you've provided a verifiable test case, and preferably filed an issue in Connect.  

@Florin: The default rendering mode for non-IE hosts of the Web Browser Object will continue to be controlled by the Feature Control key.  We have no plans to change this.

@Zebb: Are you on Windows Vista or Windows XP?  Changes in Windows Vista allow for better recovery from hangs.