Please test your sites with IE7

The information published in this post is now out-of-date and one or more links are invalid.

—IEBlog Editor, 22 August 2012

I’m very excited we’ve released a public preview of beta2 that everyone can download. I’m also very happy that we’ve opened up a couple of different avenues to take your feedback. As Dean mentioned, there is a great checklist for site developers on the IE7 preview site; I wanted to call out a specific item on that list.

Markus posted a while ago about the demise of CSS hacks in IE. We heard your feedback in the comments to that post, and we’ve put out a public preview release of IE7. I’d like to reiterate the call to action – if you are a web developer, please check your usage of CSS hacks. In addition to all the improvements in our CSS standard support that I talked about in July, and the other stuff I mentioned at the PDC, the IE layout team – developers, testers and program management – has done some amazing work since my last update and fixed the core of our CSS box model overflow problem (this means content overflow does not need to cause boxes to grow in size). We’re still refining our solution, but this means, for example, that simultaneously 1) the Holly hack won’t work properly anymore, and 2) it probably won’t be applied, because as we said before, the Star-HTML hack is gone.

It’s a difficult challenge to keep compatibility with sites and apps but break compatibility for standards compliance, and we would appreciate you checking your sites and making necessary changes where you’ve hacked in non-standard stuff for IE in the past. We would particularly like detailed feedback on the work we’re doing in this area.

We’re all very excited to start releasing. We’re going to try to restrain ourselves from posting dozens of blog posts today, and stop at three. 🙂

 - Chris Wilson

Update: We now have a MSDN article that walks you through CSS compatibility issues:

Comments (233)
  1. Anonymous says:

    Great news, and congratulations on getting through the door what must have been a very significant amount of refactoring work in CSS compliance areas alone.

  2. Anonymous says:

    Great work! I’ve installed IE 7 on 2 Windows XP machines and the install went without a hitch.

    It takes a bit to find the locations of icons and new features, but all in all it looks great and so far, works great!

  3. Anonymous says:

    Well, I’m very confused: underscore hacks and others are gone, bad box model is gone, but bad margin model is still here. So how can I tell IE7: pls, use good margin model?

    I think that now is time for conditional comments.

    Margin model: margins collapsing when overlaps.

  4. Anonymous says:

    Will installing IE7 overwrite IE 6, 5.5, and 5.1 on my machine?

  5. Anonymous says:

    When will be released some version for Windows Server 2003?

  6. Anonymous says:

    So are you guys ever going to offer a standalone version of IE7? How do you seriously expect people to use this program if it overwrites their most important web programming tool? You don’t even offer an option to uninstall it.

    Now obviously I’m going to have to get an additional license for Windows and run it on another machine or on a virtual machine, but this shouldn’t be necessary.

    You guys also need to stop breaking your own UI standards. I can’t see any normal user wanting to use IE7. It’d confuse the hell out of them. It even confuses the hell out of me!

  7. PatriotB says:

    Small typo on this page:

    4th item, part d: "Do not mix HTTP content into an HTTP page" … should be HTTPS.

  8. ieblog says:


    Yes, this replaces IE6 on your XP SP2 system.


    We will have a beta coming up which will support all of our shipped platforms. The preview is only for x86 XP SP2.

    – Al Billings [MSFT]

  9. Anonymous says:

    I found two examples of bugs in IE6 that still don’t work in this beta of IE7 (oddly, one of these works in IE5.5)

    Both of these can be found at Nathan Smith’s Sonspring Design blog.

    Multi-Class Bug:

    Absolute Positioning Bug: (Screenshot of correct rendering:

  10. Anonymous says:

    What kind of example are you setting by putting up a tour design that is broken on all other browsers but IE 6? Disheartening, but good for the competition!

  11. Anonymous says:

    Monumental improvement, but still experiencing [some] problems with margins.

  12. Anonymous says:

    > So are you guys ever going to offer a standalone version of IE7?

    Quoted for Emphasis.

    Firefox runs from a 32 MB USB-Stick.

  13. Anonymous says:

    seems like such a simple bug to fix. hopefully beta 3.

  14. Anonymous says:

    This sites fails to render.

    Another question, is possible disable the Clear Type fonts in IE7 Beta2 ?. Thanks

  15. Anonymous says:

    If I try to visit the windows update site, the new browser crashes.

  16. ieblog says:

    If ClearType on by default bothers you, you should try 1) tuning it with the tool at, and if that doesn’t help, 2) disable it in IE (Internet Tools…Advanced…under Multimedia, uncheck "Use ClearType".

    -Chris Wilson [MS]

  17. Anonymous says:

    I can’t get my RSS subsscriptions to import.

    Otherwise, it runs well, albeit the UI is a bit confusing at first.

  18. Anonymous says:

    Too bad it replaces IE6 – I don’t have spare Windows installation to break, so I won’t be able to test it. However:

    If overflow is fixed (grrreat!) does IE7 have support for min-width/min-height then?

    Is hasLayout bug still triggered in same ways?

    Do hasLayout objects still clip their absolutely positioned children?

  19. Anonymous says:

    There seem to be some rendering issues in Community Server admin pages.

    Love the RSS support, as well as the tabbing. My first impressions are good.

  20. Anonymous says:

    SP2 won’t work on my computer (makes turquoise lines all over the screen) so I won’t be able to test my site. I’d be happe if someone here could report how it looks:

    It’s made for Firefox with transparent pings and advanced CSS galore, so if IE7 knows its stuff it should look fine.

  21. Anonymous says:

    Too bad most of us institutional developers won’t be able to do ANY testing whatsoever on our work computers–given that we run 2k. That sucks. That’s my first thought.

    My second thought, where does it say the requirements for IE7?

  22. Anonymous says:

    Windows Update, work for me.

  23. Anonymous says:

    Please disable clear-type by default on CRTs, what a nightmare that will be if you ship it turned on!

    Makes me feel like I’m going blind, I can only imagine what the more senior staff would say!

  24. Anonymous says:

    Fantastic! Installation ocurred flawlessly.

    I already notified many realtors in Costa Rica and the U.S. about the availability of this beta version and encouraged them to test their Web sites.

    Congratulations to the IE team.

    Pura vida!

  25. Anonymous says:

    Is there a way to keep the text from appearing huge? I tried disabling cleartype but everything is huge.

    Other than that it looks very impressive. Reminds me of AOL explorer.

  26. Anonymous says:

    I am wondering if the development of the tabs feature is completed to a point where there is no way to implement any other features that has to deal with tabs. The reason I’m asking is I would and I believe many others, would love to see a feature which allows you to group a bunch of favorites together into a folder. From there, have that folder located on your "Links" toolbar and that we would be able to right click or middle click on the "folder" and just have it open up in a group of tabs, very similar to FireFox.

    Is this something that can be implemented or brought up to the developers? I’d truly appreciate it.

  27. Anonymous says:

    The installation worked without a hitch. I like and appreciate many of the new features. That said, I don’t care for the new interface. I am seeing some major useability problems here. The new interface is confusing at best. Do you guys even follow your own GUI standards?

  28. Anonymous says:

    I think there may be an issue with OWA for exchange 2003 and the S/MIME control. I can not open a new message to send and when I try to reply to a message or forward a message I get the same issue. What happens is the page opens like it should, but it never completely loads and finally comes back and says: !An unknown error happened.

  29. Anonymous says:

    One thing I noticed in the search engine additions…A9 is missing. Can we get A9 integration?

  30. Anonymous says:

    Is there any way to turn off the address bar on a window? Our current application opens up a pop-up with no navigation controls, etc. to save screen real estate. The new "always on" address bar is eating into that. I’m not asking for a way to turn it off remotely, I’m asking if there’s an option in the advanced features to turn it off.

  31. Anonymous says:

    So is there a keyboard shortcut that puts the focus into the new search box the way Alt+D puts focus into the Address Box?

  32. Anonymous says:

    I’ve installed the IE7 Beta 2 Preview with no problems on my Windows XP Home Edition SP2 machine. I’m curious as to how you’ll handle updating the browser for IE7 Beta 2’s actual release and then final release. Will it simply be a case of installing the package over the previous version?

  33. Anonymous says:

    I haven’t downloaded the beta yet (will be doing so in the short future), but I do want to mention one BIG usability flaw in the design that I can see simply from the screenshots…

    The New Tab button should really be on the left. I guarantee someone will accidentally hit Close Tab when they mean to hit New Tab (which is the usability problem with putting New Tab on the far right of the bar; the last tab’s Close Button is always right up against it).

  34. ieblog says:

    Dark Phoenix,

    The button to close a tab is on the tab, itself. You shouldn’t have the problem that one has in, say, Firefox, where the close tab and close window button are next to each other.

    – Al Billings [MSFT]

  35. Anonymous says:

    the menu doesnt render horizontally.. IE7 bug considering this is a microsoft website 😉

  36. Anonymous says:

    "I am wondering if the development of the tabs feature is completed to a point where there is no way to implement any other features that has to deal with tabs. The reason I’m asking is I would and I believe many others, would love to see a feature which allows you to group a bunch of favorites together into a folder. From there, have that folder located on your "Links" toolbar and that we would be able to right click or middle click on the "folder" and just have it open up in a group of tabs, very similar to FireFox."

    It’s already there. Open the array of tabs you wish to save together. Press the + icon to the left of the tabs. Choose "Add Tab Group to Favorites". To open that array of tabs, locate the folder you just created in your favorites and press the green -> icon that will appear to its right. Viola!

  37. Anonymous says:

    (Let me know if I should report this problem elsewhere! Thanks!)

    IE7 beta 2 doesn’t seem to correctly render transparent (alpha-channeled) PNG images which are used in layers (floating DIV’s) on a page.

    It renders the alpha channel portions incorrectly. For example, a soft alpha-channel drop shadow becomes completely solid black.

    To see this in action: 1) Visit 2) Click "Previous Entries" on the top right corner of the page.

    The pop-up window will have a solid black shadow instead of a soft one.

    Thanks so much for investigating!

  38. Anonymous says:

    what’s the deal with the ftp client? is it just missing because this is a beta or will it be missing from ie7? its the only reason i ever use ie, except for the occassional ie only site. and yes, i have the option in advanced options checked to do explorer style folder view for ftp.

  39. Anonymous says:

    I understand that is already there.

    I guess I was not clear enough. You know how you have a "Links" toolbar? I’m not sure if you utilize it, but I know a lot of people who do. I have many links and instead of limiting myself to just those links, I added folders to the Links toolbar with links inside the folders. In FireFox, these are treated as links, you can middle click on the "folder" located on the "Links" toolbar and it will open the group in tabs. Instead with IE7, I have to open up or dock the Favorites section and then search for the folder in the "Links" folder and then click on the green button.

    What I am requesting is that the way the folders located inside or on the "Links" toolbar, are to be treated just like links themselves in that if you click on the folder which is located on your "Links" toolbar it will open the group in tabs.

  40. Anonymous says:


    It dosent really work with credential manager like the old one did. Hope you fix it soon.


  41. Anonymous says:

    When selecting a group of tabs to open from the favorites pane, they are added to the already open tab. This is the intended use, I’m sure, but a bit annoying if one wants _just_ the group to open. If it were possible to close ALL tabs (right now you have to have one open at all times), one could have the browser open like this by default and then open an array of tabs from the favorites pane without having to close the already open tab afterwards.

  42. Anonymous says:

    Ok, I installed it … it’s nice, you’ve fixed some bugs … great.

    However, don’t you think it’s a little dumb that your BETA overwrites the latest stable release?

    I’m a developer, and it’s necessary for me to code for compatibility for IE6.x as well as look forward to the future release of IE 7.

    Why, oh why haven’t you provided the capability of running these two browsers on the same machine?

  43. Anonymous says:

    as has been said, looking real slick 🙂

    couple quick ones…. noticed right away that alot of links weren’t picking up the style/color changes on the a:hover via my stylesheet… only thing that comes to mind is that they’re preceeded by div/id, something like:

    #topbar a:hover {

    background: transparent;

    color: #f00;


    noticed some scripts not working too, such as my hoversmack(), and others while hop’n around online…. example:

    the lightbox script:

    worked find under IE6 and IE7 (beta 1), and Firefox/Opera… but not IE7 (beta 2 preview)

    pages definitely loading faster (which is awesome), though there seems to be some issues with margins… quick test page and toggle’n between FF v1.5 and IE7B2 saw things shifting for me back-n-forth.

    (okay, i shut up now)

  44. Anonymous says:

    Ahh, you essentially want the group functionality to be accessable from outside the favorites pane (in this case the Links toolbar). That _would_ be a great addition. I’d like to see link groups as shortcuts that could be placed on the dsktop, as a matter of fact. One click from the desktop would open the browser with the proper array of tabs.

  45. Anonymous says:


    Did I explain it well enough lol? I’m hoping I did as this is something I’m waiting patiently for. What you said is very great too.

  46. Anonymous says:

    forget foo…

    Lightbox worked fine for me…


  47. Anonymous says:

    ABJ: To uninstall, you can look under add/remove programs, and make sure the checkbox for "show updates" is selected. Then, under Windows, you’ll see IE7. Uninstall should restore to previous state.

  48. Anonymous says:

    just realized the mouseover popups over at Netflix aren’t working for me either 🙁

  49. Anonymous says:

    Benji: IE7 team worked with A9 to make sure IE7 works with A9 OpenSearch. You should be able to add just about any provider from simply by clicking on the search description file.

  50. Anonymous says:

    really? couple people emailed me to say that it was no longer working… tried it out myself and it simply redirects me to the image i’m linking to in the "href" on the two i added to my lastest post about IE7… even though there’s a "return false" on the onclick trigger for the links.

    hmmmmmmmmmmm…. build 5296 here, right?

  51. Anonymous says:

    Yes, build 5296 .. just downloaded it an hour ago.

    I’m actually implementing Lightbox in a site I’m working on … was one of the first things I checked. 😉

    This is purley a "proof of concept" that I threw up for the client to decide if they like it … the colors I’ve chosen were demonstrations to show what could easily be changed, so don’t tease me about how it works… if it even works for you.

  52. Anonymous says:

    I work all day long applications of my own design. They support websites I inherited. I’ve updated them over time, but not been able to convert to CSS, though some of the applications I’ve built do use CSS.

    I am excited to learn that my applications aren’t slow, but that IE6 was slow. (Due to some glaring omissions, like the inability to find text inside form elements, I’ve never been able to use other browsers.) On the other hand, I’m dismayed at how much time I’ve collectively wasted, considering how much faster IE7 is than IE6 ever was.

  53. Anonymous says:

    ahhhhh… first thing i see is the baby 🙂

    damn, works fine for me under FF…. fired up a new IE after a reboot, hard refresh, and yours works fine. *scratches_head*

    head back over to my site, repeat and rinse… hoversmack() still doesnt work, no js errors that i can see.

    oh well.

  54. Jeff says:

    Forgetfoo … I found your blog by googling for hoversmack().

    It appears to be working fine in in IE7 for my installation. (Yes, I saw the nude Brittany)

  55. Anonymous Coward says:

    I found a couple of bugs; how do I report them?

  56. frankf says:

    It seems to work better than IE6, but there are still some CSS2 properties that need to be implemented. One in my mind is <a href="">content</a&gt;, along with <a href=""&gt;:before</a>, which would be useful for people marking up scripts, or adding images next to selected links or the like.

  57. ieblog says:

    Anonymous Coward,

    See the blog post on Feedback from earlier today:

    – Al Billings [MSFT]

  58. Ryan says:

    MSFT devs,

    I’m wondering if you’ve read my comments at all and if it is something you guys will take into consideration OR if it is too late in the development cycle to implement such features.

  59. TheThinker says:

    Some thoughts:

    1. I don’t like the refresh button to the right of the address bar. It’s much better where it always was next to the fwd and back buttons.

    2. The "Stop" button now looks like a "Close something" button … you’re kindof breaking some UI conventions there.

    3. If I create a new tab, then select another tab it opens to about:blank. Then, if I click on another tab, then back to the new tab, and position my cursor in the address field, it doesn’t auto-select the text in there. Also, CTRL+A doesn’t select all in the address bar either, I need to manually paint it, or use my backspace or delete to clear the field.

    4. I do like how you’ve liberated some space from the toolbar/menubar area and freed up screen real estate.

  60. Dao G. says:

    Well, I’m disappointed. It’s a mess how IE 7 renders my sites:

    If you try to interact with the site, it becomes painly slow. And of course there are plenty of small visual quirks.

    But especially this site, which makes heavy use of CSS, is totally screwed up:

    Both of these sites used to look better even with IE6. (Try Firefox, Opera or Konqueror/Safari to see how they *should* look like.) I know this is due to the special fixes I serve for IE6 but not for IE7, but I really thought the hacks would not be necessary anymore.

    I hope IE7 will never ship in a shape like this. Please tell me this is early state. Otherwise I don’t know how I’m supposed to create websites in the future. All back to table design? Damn …

  61. PEINAO says:

    Washington Mutual does not work with Windows Internet Explorer 7

  62. Acid Dropper says:

    Hmmm … I don’t think you should consider releasing ANYTHING until you render the Acid2 test a little better than this.

    That’s just sad.

  63. the dragon says:

    No, I won’t test with it, and

    I’m currently uninstalling your preview. If you don’t care enough about web developers to make a stand-alone version, I won’t test it beyond the point I’ve already tested it in a few hours until people start raising a stink, at which point I’ll tell them the reason we couldn’t test earlier is because testing IE7 would require us to buy extra PCs, which we aren’t willing to do until it reaches a high enough market share. IE6 isn’t supported unless you release a final standalone version of IE6.

    All of the unsupported techniques that worked with Beta 1 to make a standalone no longer work with Beta 2 even with the reg hack. I tried installing the preview and creating an IE6 standalone, and that didn’t work either.

    Thanks for nothing.

  64. Ian says:

    Getting there. Some UI changes needed. I think the RSS is a little confusing, I would rather it list my feeds like FF bookmarks then list the articles when mouse hovered.

    Also, CSS is still a problem, alot of work still to be done. Basically, i would like to build a site using CSS and it will work in IE7 as good as FF or Opera.

    Just please hurry, ie6 is a nightmare!!!

  65. My site renders fine, which is good.

    But I need to be able to run IE6 for general use. Please provide a way to switch back and forth, or at least a way to uninstall the IE7 beta.

  66. Rendering…

    1.) Pure CSS drop down menus work with :hover support now:awesome;

    2.) Some sort of box model issue that will have to be reported.


    The "classic menu" appears below the address. In a top to bottom left to right language I must presume the website I am visiting has precedence over my browser?

    Buttons and toolbars are a mess, revert back to IE6 GUI where we can customize everything as we please.

    I’m happy where the (X)HTML rendering for the most part but disappointed in the GUI. Either way good job so far.

  67. ovidiu says:

    Still bad/buggy/bizzare CSS support. Why not standard compliant? When are you going to learn?

  68. ColdFusion650 says:

    does anybody know what the deal is with the ftp client missing? is it just because this is a beta, so they decided to compile it without it, or is the ftp client gone for good?

  69. Hey there my site works great love the fact that you added RSS viewer to it :-).

  70. Aaron says:

    Will the phishing filter feedback/report change?. putting it inside the IE window and using website kinda worry me. I know that you have to click the yellow warning to get the phishing report and check yellow lock thing but are there any posibility that malicous people can create a look a like site and trick people into believing that it’s a IE phishing report.hmm possible inexperienc usere or early adopter user of IE 7.0 will make a mistake.

  71. malyfred says:

    and try to add comment on any post. Do you see? Width of textarea is changed when it gets focus.

  72. Ken says:

    Does IE Beta2 support this?

  73. Ryan says:

    MSFT Devs,

    ONE more thing. Would it be possible for users to re-arrange favorites that are in folders which are placed on the "Links" toolbar in any order they wanted to instead of having it go by alpha order?

  74. Craig says:

    When typing URLs in address bar, if I jump to another tab it forgets what I was typing in the address bar of the first tab. Anyone else get this? Tis a bit annoying.

  75. Rosyna says:

    IDN support seems broken. If I go to I get an error "This web address contains letters or symbols that cannot be displayed with the current language settings." However, when I look at the language list, both Japanese and English are in the list.

  76. Mutt says:

    Tools->Phishing filter, and popup blocker does nothing, I would assume it should open an options box or something? Tools-FeedDiscovery same does nothing

    View-text size, encoding does nothing…

    Should at least all ‘basic’ user functions be working in a beta 2 build?

    Back to Firefox 1.5 for me…you guys have a lot of catch up to do..

  77. drew says:

    Well, still no support for addEventListener(). Can’t say I’m surprised, as I’ve seen the question asked more than a dozen times on this blog, and it’s been completely ignored every time.

    Can somebody please say whether support for this part of the spec is planned for inclusion and just hasn’t been implemented yet, or if this is just going to be ignored as it has been so far?

  78. EricLaw [MSFT] says:

    @Rosyna– This is behaving as expected, although I can see that the message is potentially ambiguous. For security purposes, IE7 does not permit mixing Latin and Japanese characters within a single label. We describe this in detail here:

  79. Rosyna says:

    Eric, that doesn’t really make much sense (for Japanese) as roman and Japanese text is mixed all the time in Japan. It’s one of those "marketing" things. Also, the roman script and Japanese are not ambiguous. You can’t use one to spoof the other.

  80. Tiago Magalhães says:

    This was looking fine under IE6 and Firefox.

    Tested under IE7 beta and the left menu is…

    …interesting O.o

  81. Kelly Ford [MSFT] says:


    You asked about a keyboard shortcut to put focus in the new search box. The key for this is ctrl+e.


  82. EricLaw [MSFT] says:

    @Rosyna: "the roman script and Japanese are not ambiguous. You can’t use one to spoof the other."

    Unfortunately, there are actually a fair number of Roman/Japanese confusables.

    0x30A8, 0xe05, 0x4e85, and a list of others are all potentially ambiguous.

    While it would technically be easier to permit potentially confusing combinations, it is important to us that the IDN experience in IE is secure and not a new treasure trove of spoofing attacks for the bad guys.

    As you noted, IE7 will still navigate to the site (no harm in that) but simply does not render the potentially confusable string (to prevent spoofing).

    Thanks for the feedback.

  83. mh says:



    0x30A8, 0xe05, 0x4e85, and a list of others are all potentially ambiguous.


    In which encoding?

  84. There is a bug in IE6 still in IE7 with regards sizing boxes.

    Take this example:

    A div, ‘div1’, with the style "left: 200px; right: 200px;" has another div, ‘div2’, in it with "width: 100%". This stretches ‘div1’ to 100% the window size, when it should be the width of the browser -200px on the left and -200px on the right.

    This is close to how my website-in-development is set up, and it is really a nuisance, as I can’t find any way whatsoever to fix it.

  85. Ben Buchanan says:

    <q>1) the Holly hack won’t work properly anymore, and 2) it probably won’t be applied, because as we said before, the Star-HTML hack is gone.</q>

    What do you mean "probably"? 🙂 Surely you guys should know if the selector/hack will work or not!

    Anyway, would love to test but since it blasts IE6 (and I have no way of putting it back) I won’t be able to do it. If you guys could ever find a way for us to have multiple versions of IE it’d really help us (yeah I know it’s not easy, I’m just saying…).

  86. Brian says:

    I see the whiners are out in force. Awesome job on IE7. Much new stuff to get used to, all of it looks good.

  87. 米MicrosoftがIE 7ベータ2の一般提供を開始,「CSS対応機能の実装を…

  88. Bill says:

    I’m also having trouble with Outlook Web Access. When I try to compose a new message, the window pops up properly and begins to load but eventually fails.

  89. Mystere says:

    Dao G

    You might start by using valid CSS first before you complain about CSS rendering problems.

  90. says:

    <style type="text/css" media="all">

    input[type=text] {



    input[name=funny] {



    input:focus {




    <input type="text" name="funny" />

    Strangely, the input will show up with the 15px padding but NOT with the pink background. For some reason the attribute of "type" is not properly recognized in a selector. Aditionally, the pseudo-class of :focus does not do anything.

  91. Dao G. says:


    I use valid CSS 2.1 (except for opacity and filter).

  92. Dao G says:

    … oh, and except for expression (also MS proprietary.)

    -moz- isn’t invalid according to the CSS 2.1 grammar. Nontheless, I don’t think this struggles IE7.

  93. says:


    *:after {

    content:’Hello to the after-life!’;



    It also seems that the :after psuedo-class is not implemented. This means that the has been rendered useless in IE7.

    The clearfix was actually useful for ALL browsers and used to cleverly handle float layouts without the use of a <br clear="all" /> all over the place.

    Since IE 7 seems to have finally implemented a decent box-model, but left off most of the psuedo-classes, this leaves IE 7 is in a strange "bug free but impossible to code for" gray area (when it comes to strict XHTML/CSS layouts).

    Check out Opera 9 beta’s acid 2 test and be amazed.

  94. krsaborio says:

    In celebration of the release of Internet Explorer 7 Beta 2 Preview, I have created a Web page with screenshots of most popular browsers since 1993.

    From NCSA Mosaic (pura vida Chris Wilson!), to Netscape, to IE, to AOL, it’s pretty cool to see the evolution on the interfaces.

    The page is at

  95. Kevin Burton says:


    I usually give MS a lot of crap (and you guys sometimes deserve it) but I’m glad you got this out over the firewall.

    Good job!

    (Now if only it was based on FireFox 😉


  96. David Leary says:

    Great program. IEROCKS!!!

    Except it doesnt work for my site. 🙁

    My menu goes all corrupt on the left, it is all done with css. The funny thing is that it cuts out the second half of all menu items, even the text.

    It works fine on IE6,FF,Opera!!

    Any thoughts?

  97. Majk says:

    Does IE7 support "!important"? I tested a couple of sites and it seems to me that it doesn’t.

  98. A comment before the doctype still triggers quirks mode. and this is crap, cause i want quirks in IE 6 and standards mode in IE 7. please do something about this.

  99. Rosyna says:

    How are ⼅and エ spoofable? Especially when the roman characters 0x0049 (I) and 0x006C (l) are even more spoofable? 0xe05 (ฅ) isn’t a Japanese Character.

    Also, the info bar suggests it can be fixed by adding Japanese to the language list, when it cannot (as far as I can see). So yes, it is ambiguous and wrong. The problem then becomes when people go to legitimate sites and get that error. 月 itself is not easily confused with any roman character that I know of.

    Safari, for example, has a list of Languages that have no characters that can easily be confused with each other.



















  100. xfinx says:

    You can download multiple IE versions here (they are standalone 🙂 )

  101. zeph says:

    It would have been FAR more useful if you’d released a standalone beta, Not very forward thinking!

    There are standalone hacked versions of older IE out there for the finding – google for em and they’ll turn up on the first page of results. These aren’t a perfect solution at all, but they’re better than nothing…

  102. Matt Abbott says:

    IE7 is already shaping up to be a tour de force in the internet world, I like all the work at has be done.

    @all nay-sayers: This is a BETA, its not the final release, I’m sure all most of the glitches and bugs are on a list of things to be ironed about in the future. They still have a lot of time left before the planned release date so don’t start slating the team just because they haven’t solved ‘your’ paticular issue.

    @zeph: As previously stated, IE is infact a collection of components which are distributed and used by different areas of the OS. IE7 brings updates to not only the rendering engine but I believe changes to the underlying protocols to, so any changes will affect the whole OS, thats why it replaces the current generation IE on your system and cannot be run side-by-side. If this was done, certain issues would no doubt arise between applications dependant on the older version of the components and applications using the new versions.

    That any clearer?

  103. Thomas Tallyce says:

    a:hover stuff seems broken when preceeded by div/id as per the above poster’s findings.

  104. Dao says:

    > This is a BETA, its not the final release,

    Yes, this is a beta, but as it is especially addressed to web developers, it should somehow show what can be expected in the final version.

    > I’m sure all most of the glitches and bugs are on a list of things to be ironed about in the future.

    But it’s also said "Please send us your feedback on the IE7 Beta 2 Preview". So this is what the "nay-sayers" are doing.

    > They still have a lot of time left before the planned release date

    What do you mean by "a lot"? As I see it, it’s dang few time to teach IE7 proper CSS (or more proper, as there’s progress at least).

    > so don’t start slating the team just because they haven’t solved ‘your’ paticular issue.

    Again, don’t mistake slating for reporting.

    And I’m glad you quoted "your". Of course these are issues of us, our visitors and the IE team.

  105. Peter says:

    Thanks for the release.

    Since you fixed the box model overflow problem, I think you should support the min-height property as well. I use the overflow:visible property for contentEditable areas and I now have to remove the height for the box to expand with the content. But a min height would be nice.

  106. Bob says:

    Wierd stuf with my news comments in IE7 b2. In IE6 it works fine (except there is the "gilliotine bug"). In every other browser it works fine. In IE7 the comments flows under the page wich ends long before the comments do.


  107. Cecil Ward says:

    The CSS :first-line/:first-letter whitespace bug described in is still not properly fixed in IE7 build 5296.


    p:first-line, .morestuff { } – which FAILS,


    p:first-line , .morestuff { } – which works.


    Cecil Ward.

  108. Richard says:

    The page is rendered on my Dell 2405FPW as if the scroll bar is not there. It should be rendered as if the scroll bar was there.

  109. Richard says:

    Here are my complaints:

    The placement of the menustrip breaks Microsoft’s recommendations for Windows XP and makes Internet Explorer 7 look ugly. Please place the menu strip at the very top of the window like it is in Internet Explorer 6 and in Firefox.

    I absolutely hate the design of the context menus. The tab context menu is designed in such a way that I find myself closing important tabs as I am used to Firefox’s "new tab" option being at the top. I rarely use the context menu to close a tab (I middle click) so that option would be best at the bottom of a tab.

    When I highlight, right click on a word and do a Google search it is not opened in a new tab, and I lose all existing work (such as what I am typing now which I have typed 4 times already) in any text boxes. Internet Explorer should open a new tab and remember what was in the text boxes.

    Internet Explorer 7 at the very least fails to produce a discernable face in the Acid 2 test:

    Please fix at least some of the bugs so that a face appears and the whole thing is not red.

    Tabs are not drag and droppable and it does not seem as if much usability testing was done in the interface prior to Beta 2.

    At the very least, could you integrate something along the lines of Firefox’s Adblock Plus and Tabbrowser Preferences extensions. Adblock Plus functionality (and hopefully compatibility) would make pages load faster and make things less annoying. Tabbrowser Preferences functionality would make Internet Explorer 7 easier to use. If implemented, well thought out default preferences would be nice. And if you are not willing to integrate something like AdBlock Plus, at least make sure that it is possible for a third party to write it and integrate the settings in Firefox’s Tabbrowser Preferences extension. It would make Internet Explorer 7 much more usable.

    If the bare minimum of what I have asked for is not in the final release, do not expect me to switch back from Firefox. I doubt many others will too. I must complement you on your implementation of Firefox’s Foxpose extension through.

  110. hasLayout is still an issue, but can still be triggered with the usual height:1%;.

    I must admit I’m very disappointed about the rendering, on my website it’s almost exactly the same as it is in IE6 – which is not a good thing, as I’ve been using the underscore hack (which still works when you use it in the element’s style-attribute though!) to get IE6 to display it more like Firefox and Opera. The problems are mainly about measurements; letter-spacing, margin and padding. So now I’m implementing more sniffing in order to have IE7 display my site better…

    But the UI is nice though! I do have several requests for it though, but mainly I miss an option to place the tab-closing at a static place instead on directly on the tabs, like it is in FF, and it would be really nice if the favicons were displayed on the tabs. And the placing of toolbars is very buggy btw.; using Google’s and your’s Developer Toolbar as ‘extras’.

  111. Rik Hepworth says:

    Great news that the box model changes prevent IE auto-growing the box if content overflows, but what about support for min-height, min-width, max-height and max-width?

    For example, on there is a banner graphic running down the left side. To ensure that the banner isn’t cropped there is a min-height for the container div of 650px. Works fine in Firefox, Opera, Safari etc. Breaks in IE, but is currently fixed by using the * html hack to provide a height=650px rule for ie only, _relying_ on the expanding box behaviour of IE.

    In IE7, however, the layout breaks. The fixed behaviour over * html means that IE7 doesn’t see the height=650px rule, but it fails to implement the min-height rule that it ought to.

    This is less trivial than it sounds – if you’re going to fix your rendering behaviour to prevent the box expanding to fit content, you _must_ give us support for the css rules to manage that behaviour. I’d echo the general call for standards compliance, overall.

    On a more positive note, I’m impressed with the improvements to functionality, security and UI – good job!

  112. Dan Whiting says:

    Excellent to see this available ahead of time. I have been looking forward to the RSS functions, and am impressed by how they appear in the interface. The one suggestion I have is to have the new posts for a feed appear below the feed title in the list of feeds. It seems that would be helpful to quickly scan what updates are available rather than going to each site.

    Keep up the great work!

  113. Kev says:

    Conditional comments takes the sting out of the worst of the bugs but there are some very strange :hover issues. These seem to be rendering very slowly on a:hover for me.

    All in all, this is a step in the right direction for which you deserve kudos but its not exactly the MS saviour I think a lot of us were hoping for after the last blog post detailing the CSS bugs that would be fixed in Beta 2. However, I’m bearing in mind that this is only a Preview :o)

  114. Demis says:

    Personally, I agree with implementing CSS rules. But we still would like to be able to target IE browsers in our CSS stylesheets.

    We are currently able to target with

    <!–[if IE 5.0]>


    The CC’s mean that we have to have the rule on all our HTML pages. Which will add development/maintenance time, plus unnecessarily increase bandwith.

    Why don’t you provide a way to target IE with CC’s in CSS like:

    /* [if gte IE 7.0]

    .onlyForIE7 {fix-outstanding: css-bugs;}


    If thats not your style, you could always have:

    @media IE7 {

    .onlyForIE7 {fix-outstanding: css-bugs;}


    Not doing so, will mean that everyone will go on a CSS hack hunt after IE 7 is deployed. If one is found, everyone will prefer to use that as it is much easier to maintain and code for as we are able to put it next to the offending rule.

  115. A.H says: is not working!


  116. There’s a problem with padding and absolute positioning.

    I briefly documented the problem at

  117. Mystereman says:

    Dao G,

    No, your site does not validate against the CSS validator. If even a single invalid CSS element is used, the page is non-standard, and rendering cannot be predicted. Firefox does weird things with invalid markup, for instance. So, any complaint of "It is standard, except for…" means it’s not standard.

  118. Cecil points out above that in our preview we aren’t properly checking pseudo-element error conditions when used in a group selector. This was a condition we happened to miss and have fixed it so this won’t be a problem moving forward.

    When we first started talking about CSS pseudo-elements, I also made the mistake of using a poor example where-in the pseudo-element could appear anywhere within the selector as long as there was only one. This was a mistake on my part and isn’t how we actually handle this.

  119. Taylor says:

    Would you sue me if I took all the features of IE7 and made my own browser? Of course I qould probably make an inferior rendering engine, but would that be wrong?

    That being said, y’all are lucky the Mozilla Foundation is not in the least litigious.

  120. Steve says:

    Whew – got around the msfeeds.dll install issue. Had to change permissions using regedit Hkey_Classes_Root/Htafile

  121. I’ve noticed that IE7 does not render padding and margin correctly. All of my sites work perfectly in IE6 and all the Gecko based browsers ie. Firefox, Safari etc…

    Is it really all that hard to go the extra mile and support web standards like Firefox. I mean come on you’ve already incorporated the use of tabs!

    The problem I’m facing is that the 98% of people that are using IE6 will be forced to upgrade to this inferior browser once released and included in the automatic windows updates. Unless the css rendering engine is corrected or at least take it back to the way IE6 renders css.

  122. Seb Frost says:

    Andrew Ingram:

    The only code you need to add to make the positioning in that example work in IE6/IE7 (without adversely affecting Firefox) is:


    And IIRC IE is doing nothing wrong in not assuming this automatically – I don’t think the HTML specs say that the HTML & BODY elements should stretch to fill the viewport even if they have no flowed content – Firefox merely assumes this. Just different interpretations of something that isn’t laid down in the spec.

  123. Shining Arcanine says:

    Example 11 is not rendered like Opera and Firefox render it. The bottom of the containing box should not wrap around the inline floated box. Instead its bottom should ignore the float and render as if the float was not there. Load the page in Firefox to see how it should be done.

  124. Shining Arcanine says:

    In the example, the text fails to wrap correctly in the red bordered box. It should not be crossing into and over the border.

  125. Shining Arcanine says:

    The escaping floats bug has not been fixed:

  126. Shining Arcanine says:

    Another bug that was not fixed:

  127. John Wrana says:

    This is IE 7 Beta. It’s still work to do until a first final release will be out. So it should be clear that IE 6 will still be the main browser for a long time, so webdevelopers need to test their sites in this one.

    If you don’t release a standalone version the whole process, the switching over from 6 to 7, will unnecessarly last longer than it must.

    Should all webdevelopers buy a new PC with XP2 just for installing and testing IE 7? Sorry, I’m very interested in testing the new IE, but without a standalone version, this is really crap.

    Happy Coding.

  128. Shining Arcanine says:

    The box model is not being followed in strict mode:

    Inline images are not being aligned with the baseline of the text:

    By the way, on those pages, the max-width is 700 pixels, which Internet Explorer 7 refuses to honor.

  129. Grah3am says:

    Min-height, min-width, max-height, and max-width are an absolute must for table-less css layouts. I have been using the * html hack in order to specify "height" for ie6, but now that doesn’t work. I have 60 client websites that will break in ie7 if this is not honored.

  130. SurrealLogic says:

    I noticed that ctrl + scrolling my mouse wheel changes the zoom level rather than the text size (the behavior in IE6). This is proving a bit problematic as a lot of my layout is based on the ’em’ unit of measure, which breaks down when zooming in and out (but still works fine when the text size changes). So my question is: what is the proper way to design a layout that adjusts well for both changes in zoom as well as changes in text size?

  131. Brian West says:

    I like *most* of the fixes with IE7. But it CANNOT be released the way it is. Right now the Box Model Overflow problem has been corrected, but the min-height support has not been added. So now there is no way to hack a min-height into IE. If you arent going to add support for min-height, then please dont fix the problem with Box Model Overflow, unless you can tell me another way i can simulate the min-height property without drastic changes.

    For me, min-height was one of the biggest problems with IE and it baffles the mind as to why support for it hasent been added yet. I wish at least you guys could explain why you arent supporting it yet. Did you forget about it, or is there some reason you don’t want to add support for it?

  132. Dao says:


    We are not talking about Firefox. We are talking about IE7. And as filter and expression are MS inventions (that should not be applied anyway, as I use * html) I don’t see how this has anything to do with how IE7 renders my sites like crap.

    The CSS standard says how user agents have to handle selectors, properties and values they don’t know. So rendering *can* be predicted. Even if my stylesheet is invalid, the valid parts have to be applied. That’s actually a point that every browser gets. Firefox acts correct, as does IE. So I don’t know what you are actually trying to say.

    By the way, I know damn well what is part of the standard and what is not. If I include non-standards code, that’s also thanks to MS. Thus I’ll keep saying my CSS is valid except for ~. Please stop bothering me because of this.

    That said, I don’t think the CSS validator is useful anyway. Most oft the color warnings are stupid and throwing parsing errors because of -moz- or -khtml- is just wrong:

  133. Shining Arcanine says:

    Internet Explorer 7 Beta 2 Preview does not render PNGs correctly:

    Apparently no browser does (although the ones that do, render things better than the public preview). Why not make Internet Explorer the first?

  134. Cecil Ward says:

    A redraw issue in this build of IE7.

    Go to

    and hit Page Down to scroll down, then Page Up to get back to the top. No problem.

    However, if I hit Page Down, then scroll back up to the top again by dragging the scroll bar back up with the mouse, parts of the black thick border are not redrawn (surrounding the set of "float" boxes).

    tested under WinXP at 1024×768.

    Cecil Ward.

  135. Mario Donick says:

    It’s okay, the only thing I recognized while looking my site: sometimes the pictures (simple <img align="left" /> tags) overlap the text. I used hspace on it, so this seems not work.

    Other sites, like seem to have gone mad 😉

  136. Michael says:

    Webpage scaling seems to work fine now, in beta 2. IIRC it didn’t work correctly in beta 1.

  137. Codrus says:

    If you are worried about IE7 messing up your installs, you should consider installing a copy of Windows inside Virtual PC and running IE7 inside a sandbox there. Sandboxing different OSes and browsers is something VPC has always been very useful for. It may not be your ideal solution, but certainly better than buying a second PC.

    VPC is also part of the the MSDN Subscription.

  138. Cecil Ward says:

    A suggestion: It would indeed be very helpful if, at some point, someone from MS would consider posting an updated and expanded follow-up to Chris Wilson’s "fix list" which featured in

    If certain still-outstanding CSS bugs are definitely not going to be fixed, then we can at least make a plan.


    Cecil Ward.

  139. I just tested it and PageSpy works with IE7 just fine:”>

    I use IE inside the program too and it works ok.

    BTW today the new version released:

  140. Markus [MSFT] says:


    Your changed behavior most likely will be due to our overflow fix. this also is the reason for the rendering problems. is informed and they are updating their site.

    Here is a good doc that walks you through our CSS changes that most likely can cause rendering issues:

    Remember this is a preview release to give sites exactly a chance to evaluate and prepare for IE7.

    — Markus

  141. Markus [MSFT] says:

    @Brian West

    We are workign on min/max-height/width. The overflow work was the precursor for us being able to introduce anything but min-height like you see with IE6.

    — Markus

  142. Markus [MSFT] says:

    @Shining arcane

    Beside escape-floats bug (which we will not fix for IE7 since it requires big architectural changes) all other PIE issues should be fixed. Can you please provide some more details where you see things not working?


    – creeping text, Example 11 of the floatIndent


    — Markus

  143. red says:

    say IE7 is only 50% compliant with CSS2 when it is released.. do we have to wait for IE8, or worse yet, upgrade to the latest Windows OS to get new updates to CSS compatibility?

    IE7 is nice and all, but people do have a valid point when they bring up Firefox. Firefox works with NT4 and 95. IE7 will only work with NT 5.1 (XP) and greater, in other words, MS decided a long time ago to stop releasing every new IE to all their OS’s.

    I don’t suppose you’ll be backporting CSS et al support to IE6 SP1, will you? I don’t mean just XP SP1, I mean Windows 2000 as well. No one said you had to backport all the security stuff (although that would be nice).

    This would be better if this headline came out: "Microsoft reverses decision on standalone IE"

    I suppose they are also waiting in 2007 to release SP3 just to wait for XP to enter extended support, and slap us with a crappy "Update Rollup", right?

  144. Dao says:

    Markus, can we expect support for :before and :after?

    And since you implemented the attribute selector, these would be nice, too:




    This may be CSS3, I don’t know, but it’s handy anyway.

  145. I would LOVE to test out IE 7, except I really can’t give up having IE 6 to test with.

    Please, please, please distribute a stand-alone version. Or, at least, instructions on how to get IE 7 to run alongside IE 6. Otherwise, you will be losing a large share of potential beta testers and valuable feedback.

  146. Suggestions


    1. Please please please please please please please please please please please please implement min-width, max-width, max-height and ESPECIALLY min-height since that last one could previously be simulated by using height and * html, but not anymore now.

    2. Wouldn’t it be trivial to add a default styling for <abbr> which makes it have a (dotted, dashed, whatever) underline? Maybe that’ll cause it to be used a little more often.



    Furthermore I’m seeing several problems on my website: :

    1. IE7 doesn’t ‘get’ the absolutely positioned comments links. They are now all below the page. Please also note the spacing between them which stayed intact. This seems to have to do with the bottom: absolute positioning (and maybe with the next problem as well?)

    2. The border continues in the margin area between the blog posts.

    3. On the ‘articles’ page, it doesn’t display the – character (EN dash, U+2013), even though I specified multiple fallback fonts which contain it (Verdana, Arial MS Unicode, Lucida Sans Unicode, etc).

    4. The width is not limited to 50em max anymore, it used * html { width: 50em; } especially because IE doesn’t understand max-width. See my suggestion above.

    5. The ‘Grauw’s Web Spot’ heading is too wide. I specified width: 75%;, which should be relative to its container (which is absolutely positioned), but instead it looks like it takes 75% of the entire document’s width.

    6. There is a single line of white between a post heading and the dotted line around it. This is caused by the post heading having a position: relative; on it.

    7. the ‘bottom’ style property seems to have some more trouble: the method to create a footer at the bottom which is described here [1] doesn’t seem to work anymore on IE7, and this is not because I used a * html hack. By the way, because ‘height’ has been fixed and ‘min-height’ isn’t supported, it could not work anymore anyway. Again, see my strong suggestion above.


  147. Lauren says:

    quick request (i havent installed the beta as i have dev. work to do):

    support for pseudo-classes on non-anchor tags, for example

    div {background-color:#fff;}

    div:hover {background-color:#ffc;}

    could you squeeze that in if its not there yet?

  148. Additionally,

    8. The ‘dotted’ style only seems to work when all borders are either dotted or not showing. E.g. the level-2 heading (‘Web log’) doesn’t have a dotted underline, even though it should. It’s using border-bottom: 1px dotted #888; in combination with border-left: 4px solid #AAA;, and if I remove the latter it becomes dotted.

    9. General performance of my website is a bit on the low side, it hogs while scrolling.

    10. The date that is shown in the heading of each post has no spacing on the right, even though it has a padding-right: .3em; style on it (it’s absolutely positioned using right: 0;).


    p.s. Lauren: that’s working already.

  149. beast says:


    Are you happy with firefox? What do you doing here? :))

    I tried your firefox and is NOT the better browser. An alternative would be Opera and NOT firefox.

  150. Todd says:

    So far, IE 7 looks like an improvement to IE 6, I’m happy to see the position: fixed is working. Also, glad the memory leakage seems to not be as killer as it was in 6. However, I’m still seeing some memory leakage. Sure, hope this all gets fixed, IE and Firefox both. Opera so far seems to be the best browser for memory consumption. Haven’t profiled Safari yet.

  151. Dao says:

    > Are you happy with firefox? What do you doing here? :))

    Could you please take a close look at the topic?

    > I tried your firefox and is NOT the better browser. An alternative would be Opera and NOT firefox.

    What a valuable reasoning.

  152. James Edwards says:

    It remains to be seen how good, how stable, and how standards-compliant IE7 will turn out to be, but inevitably it will have unique issues, and inevitably we will need a way to identify it for CSS.

    Conditional comments are a poor solution, because they require changes to HTML – a site which uses a conditional comment to include a stylesheet for IE6 (only/or later) is now going to have to edit that comment, on every page in which it appears, to accomodate IE7.

    CSS hacks are better, because they provide a centralized location from which to do this kind of thing – a few edits to a single stylesheet, and bob’s your father’s brother. But IE7 provides *no unique CSS hack* – from what I can discover there is no way, in pure CSS, to identify IE7 [I’ve compiled a spreadhseet of the tests at –]

    But if you put back * html, then we could use that in combination with an attribute selector, to identify IE7. Please – put back support for this syntax – without it you’re making our jobs a whole lot harder than they need to be.

  153. earther says:

    I don’t want to overwrite IE6 so tested on the BrowserCam installation.

    Most sites fared OK but the CSS horizontal rollover/dropdown navigation on is totally fubared. It took hours of hair-pulling to get it to validate and work cross-browser in the first place. I can’t bear the thought of having to take this back to the drawing board.

    For those who haven’t yet installed (or uninstalled), here’s what the screen capture looks like – Now that’s a real mess!

    I’m not going to spend time trying to fix it at the moment. I’m hoping that changes will be made on your end so the glitch goes away on its own.

    Right now, I’m not a happy camper. 🙁

  154. Cindy says:

    Even if IE7 is perfect with respect to css, you know that all the css hacks, etc will still have to remain because the entire user base is not going to stampede over to IE7. I’m still supporting 5.5 for crying out loud. I can already see that I will be putting conditional output if I detect IE7. It’s just more coding. (Thank god I generate my html on the fly, I don’t even wanna think about embedding hacks within static html like holly hack, etc :-O )

    That said, PLEASE put out a standalone version for testers and developers! I’m sitting here actually thinking about partitioning my disk into multiple windows xp installs (not even sure I can do that with one legal copy of xp to begin with) before I come to my senses.

    It seems a standalone can be hacked, but it’s very fragile with respect to the December security patch, so I’m really rather cautious. I’ve also seen a large number of people complain that they *can’t* uninstall back to IE6 after doing (legitimate) IE7 upgrades. Hoo boy. It’s not like we have a spare XP I can hack around with either. Several 2K installations, one other XP installation (where said coworker is jealously guarding his working IE6 copy as well), and two macs (pre OSX, and OSX).

    It really is annoying; all other browsers I can pretty much safely assume random users will have the latest versions, but not IE. I basically support FF, Opera, Safari, MacIE (shudder), MSIE 5.5 and 6 and it will soon be MSIE 7 separately. I’m doing stuff with Unicode and plane 1 characters which would probably be excellent feedback for IE7, so how ’bout it? Standalone version not affecting my IE6. Thanks.

    Ah hell, while I’m already asking for the moon, make IE7 versions for everything from Win98 onwards…

  155. Dark Phoenix says:

    > a[rel~=external]

    > a[href^="http:"]

    > a[href$=".jpg"]

    > This may be CSS3, I don’t know, but it’s handy anyway.

    When I ran these through the tests, they appear to work perfectly. Also, IE7 gets ~ (CSS3 indirect adjacent sibling selector).

    As mentioned though, :focus, :active and :lang() don’t work right, and I noticed some odd issues with absolute positioning using the left property (right works perfectly). I thought it might have been because of my incorrect conditional commens ([if IE], which I change to [if lte IE 6]), but the problem didn’t go away. Unless IE7 is getting conditional comments it shouldn’t, I think it might be a positioning bug.

  156. Dao says:

    > But if you put back * html, then we could use that in combination with an attribute selector, to identify IE7.

    That would be very annoying. Everytime you need to fix something for IE6 but not for IE7, you would have to "refix" it.

    I think conditional compilation would be useful:

    foo { bar: thew3way; }

    * html foo { bar: ie6only; }


    foo { bar: ie7only; }


    > When I ran these through the tests, they appear to work perfectly. Also, IE7 gets ~ (CSS3 indirect adjacent sibling selector).

    Ok, then it didn’t work just because of :after (not implemented) and :hover (broken) …

  157. Dao says:


    foo {

    bar: thew3way;

    /*@cc_on bar: ie7only; @*/


    * html foo {

    bar: ie6only;


    * html foo

    /*@cc_on , foo @*/ {

    bar: ie6and7only;


    I really like this. 🙂

    It’s clear that IE7 will need some workarounds … wouldn’t be nice, if server-side scripts or conditional comments (html) were the only way to apply them.

  158. Seb Frost says:

    I hate to say it but it sounds like unless MS really pull their fingers out there are still going to be a lot of issues with the CSS support in IE7.

    Short of them ditching the whole thing and going with Gecko it seems stupid to think for a second that this is going to be hassle-free coding for us.

    In an ideal world we’d write one version for all browsers, but we’re realists here, we know that isn’t going to happen… conditional code is a TERRIBLE idea, it leads to more problems in the future with new versions of browsers, but do we have a choice?

    I think I too would like to see the deliberate inclusion of some sort of IE7-aimed CSS hack.

    Definately NOT * html as someone suggested further up, because that will break all our current sites using that for IE6 problems.

    how about "html:ie7" or something like that – it’s inobtrusive and other browsers will simply ignore a psuedo-class that they don’t recognise, wont they?

    Personally I’m waiting for the day that someone writes a decent virus that will corrupt everyone’s installs of IE6 and IE7 and get the majority to switch over to Firefox…. 😉

  159. Seb says:

    "unobtrusive", doh!

  160. Tod Liebeck says:

    I’ve got a demo of a CSS box model issue with IE7b2. Basically nesting two absolutely positioned DIVs with top+bottom+left+right positioning on each does not have the desired effect. I wrote up a quick demo of this here (see second post with screenshot):

    Glad to see IE7 will (hopefully) be supporting CSS positioning on both sides (top+bottom, left+right). I very much disliked having to use "CSS Expressions" when IE is detected to dynamically change width/height properties.

  161. aarfing says:

    Thanks for your efforts. I have a request. Since you’ve now fixed the float-issues, it is essential that you implement the :after pseudo-class. Otherwise it makes it really problematic to make proper div-based column-design, because you no longer can clear the float. It wasn’t a problem in IE6 because of the bugs with floats – now it is a HUGE problem.

    I’ve attached an example. You can check it in Safari, Camino, Firefox, Opera etc. to see how the :after class makes things work.

    Please, consider this issue.

  162. despite claims ie is still not fully supporting xslt! please view

    in both ie and firefox! i wish to help solve this problem as it appears that mozilla is the only browser on the market which does support full functionality and thus restrictions are being plased on the full potential of xslt!

    please contact me to discuss further if you wish.

  163. Cecil Ward says:

    Bug report – CSS and rendering issue,


    The vertical gaps between the "buttons" shouldn’t be there, if judged according to the rendering shown in IE6, and Opera 9.0beta and Firefox 1.5.

    (Tested under WinXP.)


    Cecil Ward.

  164. Dao says:

    > Since you’ve now fixed the float-issues, it is essential that you implement the :after pseudo-class. Otherwise it makes it really problematic to make proper div-based column-design, because you no longer can clear the float.

    Actually #container { height: auto; overflow: hidden; } is the proper way to solve this. But that neither works in IE7, I think.

    Nontheless, I think :after and :before are essential, too. I’m still waiting for a statement — will they be implemented or not?

  165. Cecil Ward says:

    Bug report: text rendering or layout issue


    Notice the text in the box "A call to action" is incorrectly wrapped so that in places some words stick out past the right hand border,(viewed at 1024×768, tested under WinXP) because the box is too narrow.

    However, there is more weirdness here. I downloaded both the HTML and the single .CSS file, saved both files, and then viewed the HTML again. When saved locally, the page then rendered as expected. Very strange.

    Can anyone else reproduce this?

    I notice that the page contains a child-selector hack aimed at directing styles away from IE5-6/Win. But that can’t be the issue here, and the local-versus-web thing remains very weird.

    I say "the box is too narrow" in the sense that in IE7 the text wraps at the same points as it does in other browsers, so the text doesn’t appear to be ‘too wide for the box’.

    Am I missing something?

    Best wishes,

    Cecil Ward.

  166. sekrit says:

    Rendering bugs: (Digitally Imported Radio)

    IE Crashes when any NetScaler 9000 configuration page is viewed, which is a Java applet. I can go to sun’s java site and confirm that I have java installed, but whenever I look at NS config applet, the browser dies with "abnormal termination."

  167. a public preview of IE7 beta2 that everyone can download [via; see also]…

  168. jason says:

    <b>A Huge Jerk:</b>

    <i>How do you seriously expect people to use this program if it overwrites their most important web programming tool?</i>

    What, it overwrites Firefox?

  169. laughing says:

    @ jason

    Haw, well played.

    I truly do think of FF as a deveoper’s tool. With context highlighting in the source-viewer, effective javascript debugging, and all the tools available with the web developer plugin, I wonder how the hell I ever did my job before I had Firefox.

  170. Transparent PNGs still don’t work – with the background appearing as green instead of… well… transparent. For an example, see my site (the small icons next to the Categories and Published Date are transparent PNGs)

  171. Kelson says:

    I tracked down a problem with padding and absolute positioning, worked out a testcase, and came over here… only to discover someone else had already reported it here.

    If you need another testcase, here’s

    Basically, if you apply padding to a block with absolute positioning, it shifts the content to the right, leaving its width unchanged, causing the content to overflow the right edge of the box. Oddly, vertical padding seems OK — only horizontal padding runs into this problem.

  172. Shining Arcanine says:

    Markus, in the Magik Creeping Text demo at, the gray boxes are the wrong size. In the 3 pixel gap demo, the box appears to be being rendered with a slightly smaller width than Firefox and Opera render it with and the text flows out into the border before wrapping.

    As for example 11, I think I made a slight mistake when I said that there was a problem. The following is a screen shot of how IE7 Beta 2 Preview renders example 11 and Opera renders example 11:

    Since the difference between example 1 and example 11 is "display: inline;" and that the specification specifically says that display is to be ignored with the exception of "display: none;," the problem is in Firefox and Opera and not in IE. Sorry about that.

  173. Shining Arcanine says:

    I would like to apologize for my comment on the box model not being followed in strict mode at the following page:

    Noticing that example 11 was a Firefox and Opera bug, I went back and checked everything. For the box model, I was wrong, IE7 is following it there perfectly.

    I am a student and for the past year I have been focusing on things other than web design so please excuse the mistaken report.

  174. James Edwards says:

    There definitely is a box-model problem, affecting the width of absolutely-positioned elements – it appears that an element which is absolutely positioned has its width rendered in quirks mode (or that’s the effect, anyway).

    Another test case for that –

    And a simple test for the float/overflow problem –

    I’ll withdraw my earlier statement about putting back *html – I agree, it would cause more trouble than it saves. But some kind of specific hack or conditional CSS would be very helpful – something nice and clean that other vendors could implement similarly, like @vendor-version { … }

  175. Zoltan says:


    I found this page:

    there is a serious problem with the scrolling.

    I think, it is something with the CSS rendering.

    It’s just fine in IE6, although the menu should stay at the top of the page (like in FF).

    And there is something wrong with refressing wysiwyg editors while editing the text, especiali after paste or backspace (like one at hotmail).

    Is there any hope, that favourites/history/feeds will have some filter/serch input? I also hope for some option that make IE remember the positon and status of the threes after restart. And a drag&drop like reordering tabs feature. And an option for automatize Restor Last Tab Group. And a minimize to systray feature 🙂

    These are the things that prevent me to make IE7 the default browser. All the other changes are great, and I like this new look very mutch.

    Maybe you should do some more propaganda for Add-ons (like mozilla done for the extensions) and my hopes may easily fulfill 🙂

  176. Zoltan says:

    oh, I almost forgot.

    I also hope for a feature (option) which bring the focus to the last viewed tab after a tabclose. And when there is only one tab left, there should be a close button which replace the tab with a blank one.

  177. Shining Arcanine says:

    Regarding my earlier statement concerning example 11, I reported it at Mozillazine and they told me that Firefox and Opera are rendering it differently than IE7 is rendering it because of "<div style="clear: both;"></div>" so on second thought this probably is a bug in IE7. You can see the latest on the conversation here:

  178. `legacy' software user says:

    I’m sorry; I’d love to test my site in IE7, but I can’t afford $200 to pay for a Windows license to try it on. If you’d like to release the source for a standalone version that runs on Gtk+2 I’d be happy to try it on <a href="">my OS</a>.

  179. Cecil Ward says:

    A correction to my earlier post. I wrote:

    > When saved locally, the page then rendered as expected. Very strange.

    I had used the tool TopStyle 3 to download analyse and re-save the page. TopStyle adds an HTML comment at the top of the document, which forces IE7 into quirksmode presumably. So mystery solved.

    But that rendering is still odd.

    Cecil Ward.

  180. Zoltan says:

    someone mentioned earlier, that MS break their own recommendation for the GUI, with the classic menu.

    as I can see from the vista, that recommendation will change, so nothing wrong with that.

    btw, I’m happy with that. I like the new GUIs (vista, IE7, O12).

  181. Shining Arcanine says:

    Zoltan, the recommendation for Windows XP is not going to change (as I doubt XP Service Pack 3 will give us Windows Vista style windows) so they are breaking their windows user experience design guidelines.

    It would be great if they would code things so that the menu bar would be compliant with their own guidelines on Windows XP and like it is now on Windows Vista.

  182. Nick says:

    After installing the IE7 beta 2 preview, when I open the McAfee security center stays in a blank screen the same thing with My McAfee VirusScan 10.0

  183. paweljq says: this page doesn`t work in ie7

  184. mailww says:

    US Post Office

    when trying to print out shipping lable, it blocks then allow but fail to print lable.

  185. mailww says:

    Doesn’t support IE7 for lots of pages


  186. Ron says:

    I just want to let you know that I can’t run Service Pack 2 on windows XP, period. It doesn’t work and there’s no apparent solution to it, except to buy new hardware, even though it’s already new.

    Do MS programs have to be so "conditional" that you can ONLY run certain things on certain versions?

    Without being able to test with IE7 I will just have to say I don’t support it when designing websites. But I’m not going to support IE6 when IE7 is released either, so I will have to trust IE7 to work as every other browser does.

    I’m sick of this stupid "built into windows functionality" IE has. MS has a monopoly on IE and the only reliable way for me to test IE7 would be to buy Vista, which I shouldn’t need in the first place.

    There shouldn’t be any reason why IE7 can’t work on Windowx XP SP1.


  187. Pawel says:

    I like IE7, with the following exceptions:

    1.The buttons for Favorites Center and Add / Subscribe should be removable.

    2.The annoying ClearType font should be turned off by Default.

    3.The search bar should be optional. Do not compete with Google.

    4.The File, Edit… menus should be on by Default.

  188. Marco De Luca says:

    Standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, standalone version, pleeeeeeeeease!!!

  189. earther says:

    I tried a standalone version yesterday but once again was getting multiple windows opening, error messages etc. as I did with Beta1. At least the version automatically deleted the offending registry key.

    I surrender. Have it your way. If MS will not provide a standalone version, I won’t be helping to test. Neither will I be rewriting code that works perfectly in FF, Opera, Netscape, Safari, Konqueror etc. to accommodate another buggy MS browser.

    If there are remaining or new CSS bugs in the final release of IE7, I am considering not supporting it. Those using IE7 would get a message explaining why. If enough designers followed suit, maybe things would change.

    Wow! That feels better now that I got that off my chest.

  190. Zoltan says:

    Shining Arcanine, all the upcoming ms softwares break that recommendation. o12 will also be available for xp users. even the wmp do not have menu bar, in msn messenger it can be turned off.

    I understand that for xp it should be on by default and it sould be appear at the top of the window, but IE7 is for vista, and in vista the navbar and the titlebar are combined with each other and it’s part of the "glass" frame. so both functional and design reason it can’t be on the top.

  191. Zoltan says:

    If the Favorites Center is docked the popup windows will also have that. it can be confusing

    and annoying too.

    oh, and I give a vote for the standalone version 🙂

  192. Ron says:

    The issues affecting input buttons in IE have still not been fixed even though the problem is so weird and has no reason to be there.

    This is a test case about what happens to buttons in Internet Explorer:

    Is something being done about this?

  193. Bill says:

    One of the nice things about IE that you could trust, because IE (unfortunately) is bolted into windows, no matter where you are in the world, you could always find a box to do a quick FTP upload to your sites. Great feature, very useful, so what have you done? Removed the FTP client.


  194. i see a lot of problems coming at us web developers.

    even things like suckerfish drop downs dont work properly now, (they did in IE (after ie hacks…)

  195. ravi says:

    I have found very interesting problems that do not occur with IE 6 and Firefox.

    We are rendering several database applications through the browser where the browser loads a page for about 20 minutes.

    There are loops in there like:

    while(Loop through MySQL Array){        

    Update Database

    print $key;


    1) IE 7 beta breaks up in between without proper warning when this happens. It just times out.

    2) Also there is no outputs written to the browser. Those should be written during the several minutes of excecution, dynamically.

    3) In a different application generate Images on the fly which takes about 4 seconds because these are rather complicated graphs of energy usage. The image display fails at the first time. this only works when reloading (faster because the image generation software does caching), or when right clicking and selecting ‘display image’.  

    Other browsers like IE 6.028 and Firefox wait properly and display the right content!!!!Apart good work but please have a look into this, or two!


  196. Cecil Ward says:

    IE7’s behaviour when you try to do FTP client things from within IE7 does seem more than a bit weird/buggy at present.

    But you can (well, I had no problem) still do FTP   tasks successfully by going to Windows Explorer, (see item in IE7’s "File" Menu).


  197. Nicolas says:

    There is no support for alternate CSS ????

  198. Jdowe says:

    When is the true go live date for this product?

  199. Bill says:

    I think this comment is a very important issue. I was testing a couple of my sites which appeared to be broken with over-lapping content. Have spent ages trying to fix my sites I realised I was looking at my sites with one notch (10%) up on the zoom feature. Setting it back to default fixed my sites! Ok, I should have realised but two thoughts come to mind. Warn testers of this issue. What sort of design would best fit zooming in and out.

  200. Denis Séror says:

    With IE6, modal and modeless windows have a minimum size of 100 x 100 px.I have measured with IE7

    that the minimum size is now 300 px wide and 160px high.

    As a consequence i cannot use my web application as it is, and i have to change the design of some control panels.

    Can you 1. confirm these minimum dialog window sizes, 2. tell me where I can find such specification or limitation changes.

    The "DHTML" part of the site "MSDN Home >  MSDN Library >  Web Development >  HTML and CSS" does not seem to be updated.



  201. Ron says:

    I tested my pages in IE 7 beta 2 preview as the IE team requested and discovered that conditional comments don’t work correctly in IE7 or IE6.

    Please open this page in IE to see what I mean:

    IE7 only displays half the data it should, while IE6 doesn’t return anything. I have used the <comment> tag, because I’ve used it on a live page combined with conditional comments to make something work in IE6. That particular page doesn’t render in IE7.

    So have a look at my example page and tell me I made a mistake. I used two separate windows machines with diffierent versions of IE to test this.

  202. Dao says:

    <label><input></label> without the ‘for’ attribute is not working as it should:

    I fixed this with some few lines of JS code:

    (Thus not working for labels added dynamically after the site was loaded.)

    See it in action here:

    I bet you can do something similar, hardcoded into IE7.

  203. atctech says:

    The smoothing of the fonts in ie7 makes it unusable for me. I switched back to ie6. I hope this is not a permanent feature as I will not use ie7 if the fonts remain like this.

  204. As a former Microsoft Developer Evangelist,&amp;nbsp;Rudi Larno was a well-known name in the Belgian development…

  205. Let”s start with card credit gas station application loan payday

Comments are closed.

Skip to main content