What’s New for CSS in Beta 2 Preview?


We have heard all your feedback asking us for full CSS compliance. With the now public available Beta 2 Preview build on XP, we get a big step closer to this goal. Chris talked about our priorities for IE7 before, but I want to give more detail on our work with CSS.  We had 3 objectives, in priority:

  1. Fix some really nasty bugs posted on sites like positioniseverything.net
  2. Revise parts of our existing CSS implementation to be true to the spec
  3. Add the most-requested new CSS functionality to IE

Fix some really nasty bugs:

In IE7 we’ve been working hard to address the bugs in our engine:

  • We fixed the most of the bugs reported on http://www.positioniseverything.net/explorer.html  (click on the links for detailed explanation of the bugs)
  • We addressed several relative positioning issues (relative positioned elements do not have layout, an internal IE data structure, that sometimes caused calculation errors)
  • We made the HTML element truly independent of the Body (Scrollbars are now belonging to the Canvas and you can align absolute/fixed positioned elements next to them)
  • We addressed the auto alignment issue so you can more easily build 3 column layouts
  • Resolved issues with 1 px borders

Revise existing CSS implementation to be true to the spec:

Our second priority was to complete or adjust our implementation to be in accordance with the CSS spec.

  • We made major changes to our box model to support overflow correctly on non-replaced block level elements and inline-block elements.
  • We enabled :hover on all elements, not just on <a> tag
  • We make background-attachment:fixed work on all elements, not just body
  • We fixed a number of parser bugs like * html, _property and the /**/ comment bug

Add most-requested CSS functionality to IE

Finally, for IE7 we added:

  • Fixed positioning. This allows interesting layouts of sticky menus and sidebars.
  • Enhanced Selector support: first-child, adjacent, attribute, and child selectors
    • As we were implementing attribute selectors we also added support for CSS 3 attribute selectors: prefix, suffix and substring.

Compatibility and our updated CSS behavior

Obviously, we have heard the feedback asking us to be more standards-compliant in our rendering behavior.  We must balance this ask with the need of our customers (and end users) to have their pages not be broken. To find a balance we introduced a strict mode in IE6 that lets authors opt in into the more standards compliant rendering (and, if you’re putting in a modern DOCTYPE declaration, you’re being opted in automatically).  Pages authored under non-strict mode (or “quirks mode”) will not change behavior in IE7 – so the fixes we’ve done to be more CSS compliant won’t appear under quirks mode. However, if your content is under “strict mode”, our behavior is more standards compliant, and your pages may break – for example, if you use some of the CSS hacks or rely on our old incorrect overflow behavior. We understand the pain this might cause in the short term but we believe a move to a more standard compliant implementation benefits everyone in the long run.

Last but not least, big thanks to the community: here on the blog (yes, we are paying attention to the comments), the WASP team and those out on the web reporting CSS issues with IE. We know we have a long road ahead but I am very excited with the progress we are making.

 – Markus Mielke

Update: We now have a MSDN article that walks you through CSS compatibility issues: http://msdn.microsoft.com/library/default.asp?url=/library/en-us/IETechCol/cols/dnexpie/ie7_css_compat.asp

Comments (227)

  1. William J. Edney says:

    Markus –

    Thanks for the update! This information is what many of us having been waiting for since the beta2 announcement a couple of days ago. Here’s a couple of questions:

    1) Great news about :hover support on all elements. How about :focus support on focusable elements?

    2) Does IE7 properly parse the newer CSS3 ‘double-colon’ syntax for pseudo elements (i.e. ‘::first-line’)?

    3) Thanks for attribute selectors! Will they adjust to dynamic changes made in script (i.e. via setAttribute())?

    4) Are there any plans to add support for the CSS3 property ‘box-sizing’? This would *greatly* help those of us who want to use strict mode and all of these great fixes but still want to use ‘border-box’ sizing (Mozilla supports this via its vendor-specific ‘-moz-box-sizing’ property).

    5) Will IE7 support proper ‘dotted’ borders instead of the current scheme of making dotted look like dashed?

    Thanks again for all of these fixes!

    Cheers,

    – Bill

  2. Mitchel Tyrell says:

    There appears to be a bug someone has documented on css rollovers:

    http://locusoptimus.com/css-trickery/ie7-hover-bug.php

    Will this be addressed before the final version?

  3. Wraith Daquell says:

    I know this probably doesn’t belong here, but have any developers noticed that Visual Web Developer Express will open pages in IE6, even if IE7 is installed?

    Great for testing pages in both engines!

  4. Todd B. says:

    I noticed, while I was previewing IE7, that it doesn’t currently support the display: none attribute on the option tag of select/combo boxes.

    Are their any plans to add support for this?

  5. Eric says:

    Bravo! Sounds very good and this will be a huge treat for many web designers, especially in the future. Keep it up!

  6. John Coleman says:

    Ummmmmmm IE7 is the only browser in the world that doens’t render my sites. Ergo, you aren’t NEAR done.

  7. Emon says:

    The annolying min-height isn’t fixed yet.

  8. Xepol says:

    YOu still have the table column width bug, which is listed on positioniseverything.

    When the contents of a column expand beyond its size, IE just swells the column instead of clipping. This makes some things on my site render like total merde. Hope that will be fixed soon too. I hate the random resizing throwing everything out of whack.

    Say, have you noticed that when you scale tables with images, the cells suddenly get visible edges AND gaps? Try http://www.ctrlaltdel-online.com/comic.php at 200% or better to see what I mean (1000% is VERY obvious).

  9. Michael Sheakoski says:

    Dear IE7 Team,

    Great job so far on the new browser. It is a night and day difference from IE6 in terms of standards support and is truely going to make development so much easier!

    In this post you talked about "Add most-requested CSS functionality to IE" but I think you’re missing something HUGE! I’ve been reading every single post and comment in this blog so far and every time you make a post a few people ask about display:table / display:table-* yet we never hear an answer. The latest beta does not contain this support and I’m starting to get scared that the final will not support this feature either since it is never talked about by the IE team.

    I think this is one of the make or break issues for IE since it holds the majority of the marketshare and being the only major browser to not support this feature holds back EVERY web designer that wants to truely break away from table-based layouts. It will make the difference between being the browser that developers love, or continuing to be the browser that developers hate. PLEASE add support for display:table-* we are all begging you!

  10. 1½ half years of developing, and you’ve mostly only fixed bugs when it comes to CSS? Says quite a lot about IE6, or how high CSS-compliance was prioritized – your pick.

    I’m not surprised and I can understand why, but as a web developer I’m not satisfied.

    Considering how much IE7’s rendering was alike IE6’s on my site, before hacks to either were applied, I can’t help thinking that it might have been intended that the underscore hack still works, when used on styles in the style-attribute?! But – will it be fixed?

  11. Great stuff!

    What’s the status of the following? I couldn’t find anything specifically, and given the current state of bugs being reported I don’t have a machine I can spare to test these on:

    1. Left + right with scaling width on block elements. I.e. so that you can do borders with position: absolute and position: relative and have it scale to the page. Right now IE ignores the Right property entirely if the left property is set (ie 6). By fixing this bug, you’ll allow developers to build sites with all of the power of frames without frames. Very important to fix.

    2. The resize no scrollbar bug on overflow: auto; block type elements. Currently if you resize an IE browser window that has a block that will resize with the window, and you shrink it down enough that the content inside is larger than the block that contains it, the scrollbar will showup, but any elements that are relative to the size of the containing block will not get resized, thus causing a horizontal scrollbar to show up instead of maintaining the relative sizing between the parent and child blocks without scrollbars. This causes all kinds of issues for long navigation menus that are independant of the main content and are fixed relative to the size of the browser window (see #1)

    If these two and the transparent PNG issue were fixed, I could live with absolutely nothing else being done.

  12. Philippe says:

    Thanks for this info. Some progress is being made.

    Have a look at this: http://css-discuss.incutio.com/?page=IE7. Still loads of problems with Absolute and relative positioning.

    overflow:auto apparently still needs a width specified to work correctly (see http://emps.l-c-n.com/articles/97/notes-on-textpattern) the code blocks are way to wide.

  13. Al Sparber says:

    We’ve documented a few bugs with test cases here: http://www.projectseven.com/csslab/ie7/ and will be adding more as time permits.

  14. Markus [MSFT] says:

    @ Philippe, Al and the rest

    I love the links you guys sent out. Seeing the community chime in on bugs and wish lists is a great help. Keep them coming.

    — Markus

  15. ZeromusMog says:

    I’m a big OSS/Firefox fan but I recognize that for the time being if Microsoft doesn’t support something it’s going to hold all of us back. That said, as much as I’d like to see everyone on Firefox or Opera or Safari or pretty much anything non-Microsoft, the vast majority of people out there Just Don’t Care and will use whatever is there by default, or cheaper.

    Because of that, I am still concerned by this "SP2 Only" business, and even the requirement to pass Genuine Advantage. While I suppose asking for a version for people still on Windows 98 might be a bit much (though it really, really should be considered), IE7 should at the very least be available to all XP users, and forego this Genuine Advantage thing.

    We’re all eagerly excited to go forward, and move the www to new heights, but this cannot be done until most people are on board. If you make a site that sells something, you must program the site to appeal to the lowest common denominator or you will lose money. Internet Explorer is, and for the forseeable future continues to be, the lowest common denominator. We all would like to see that raised, but putting restrictions on who can download it and for what platform will keep us in limbo even longer since often downloading alone is a huge deterrant to the masses to upgrade stuff.

    If Microsoft truly cared about the future of computing and the Internet, they would spend a few of those billions of dollars received annually putting free SP2 install CDs in highly accessable retail locations that come with some kind of gimmick (free games or whatever) to garnish people’s interest that would also include an already-done Internet Explorer web browser that brings everyone up to the newest w3c www standards. (Of course if they REALLY wanted to care they’d support the Firefox project and nudge Windows users to use it instead of IE, but I’m realistic enough to know that this will never happen.)

    Also sorry if this is off-topic, but it does apply at least to IE: the Genuine Advantage thing is a joke. It should really be removed from IE7, although the whole program is a nightmare for everyone involved. Most people pay for Windows, and understand that software piracy is a bad thing. Thanks to the BSA, no legitimate business would even think twice about piracy anymore since all it takes is one disgruntled employee and lawsuits and fines will rain down from the heavens. Those out there with a non-genuine copy of Windows visit our websites and use our programs, and it costs US money when their computers are compromised or out of date. If you want to alert people when they install MS software that their copy may not be legit, fine. Make it nice and scary so they think they will be in trouble if they don’t get it fixed soon if you want. However, for every user that doesn’t have SP2 or the newest version of IE, that is more work and expense for web developers who have to support old and buggy code, network administrators who have to deal with the worms and viruses that get in through the back doors of unpatched systems, and a bigger hassle for application developers who are suddenly staring at a more diverse (in a bad way) and antiquated userbase they have to code for.

    I’m sorry not every copy of Windows out there is paid for, Microsoft, but stop your crusade to make the honest and hardworking independant developers and networks pay for it.

  16. James says:

    I love the new look, although it looks like it might be a little confusing for older folks (my grandmother’s probably going to wonder what happened to her "favorites" button). My favorite feature is hands down anti-alias – thanks for giving us a sample of what Vista’s going to be like 🙂

    Is there any chance the search capabilities will be touched on at some point? I use the Firefox search extensively, and it’s a joy to use because it’s so integrated into the browser. The IE search dialog box, in comparison, almost feels like it was coded in javascript rather then part of the browser.

    Oh, there are also quite a few rendering bugs lying around. Say I want to buy a painting on ebay (http://art.listings.ebay.com/Paintings_W0QQfromZR4QQsacatZ20125QQsocmdZListingItemList)… it’s a little more difficult then I’d like 😉 You’ve probably heard of these countless times, but I just though I’d bring it up

  17. In today’s entry: I take a look at the new Internet Explorer beta.

  18. PatriotB says:

    James – What’s wrong with the rendering of the ebay painting list?

  19. Andy says:

    PLEASE REMEMBER COLOR BLIND PEOPLE 🙂

    Really, the tabs have some kind of grey shading in them and what looks like blue on white. Problem is, you cant change the colors – so for color blind people – they cant read whats on em.

    Those two ugly icons left of the tab suck (in a nice way). I have had to turn the old menus back on because I use so many interfaces I cant remember what all those stupid icons do and am forever holding the mouse over them to find out what they are for. I do not want to have to think to use a product. If I did I would use Linux.

    YOU BUSTED GOOGLE TOOLBAR.

    Putting that silly white box thing in the top right hand corner where I never look – I guess you want me to go chasing over a large screen looking for things. So I use google toolbar which sits right next to my favorites menu – so I only have to look in one place.

    How do you relocate those icons to the right of the tab. On your screen pic they are on a floating tool bar – on my screen they are fixed in place and go off the side of the screen. You have to click on the double chevron >> to get to help.

    Menus

    Why are all the options upside down. Close/Exit is supposed to be at the bottom of menus – it is on every other windows products – so why is the close tab where the open tab menu should be.

    Where is the "Always open link in new tabs" option that was in Beta 1. I click on a link and get a new browser window – whats the point of having tabs if you cant use them.

    Did you guys actually Userbility TEST your product. Not being rude about your usability folks – but I suggest you may want to go find some new ones 🙂

    I mean that in a polite way of course.

    On a more positive note – Thank you for putting the stop and refresh buttons back. If only I could get at them with the mouse. Currently this field is down in the bottom left corner of the screen – they are in the top right corner. Nadgers if you think I am going to move the mouse that far when I have my fingers on the keyboard, or am loading a favorite (of which the menu is on the LEFT side of the screen).

    I wouldnt mind the IE6 icons back as well (so I can put words under em).

    CSS works great and so does the feed stuff – use it all heaps 🙂

    Cheers.

  20. Xepol says:

    If everything was designed for color blind people it would all be black and white. That said, configurability is desirable. I’m betting we’re getting the vista interface without the vista configuration gui behind it. Once that is there, I’ll be you see it more configurable.

    I like the 2 buttons, would prefer the 3rd button to be static as suggested by others. My google bar is just fine, but issues more likely relate to the rebar control itself rather than IE. I continue to encourage the IE team to finally put that stupid control out its misery.

    TAB MENUS. Why are the 2 close options at the top? BECAUSE THEY DID USABILITY TESTS. Those are the 2 most like menu options you will use in those 2 menus. How do I know this? I’ve been using IE 7b1 since it was relased last year. Frankly, the other menu options might as well not even exist to me. I have found the right click to open in new tabs to be intuitive (as in, I was doing it automatically without thinking about it before I releasized that I wasn’t sure how to actually tell someone to do it!). I find the close tabs menu to be equally inituitive (as in, it was the first place I looked to close the tabs and it was RIGHT there).

    As for the search tool, I have already tied it into my favorite site search engines such as IMDB.com and TV.com. I find it can be quite useful when combined with the google toolbar.

    Personally, I never had ANY problem with the go/refresh/stop buttons being unified, since the concept is state based and mutually exclusive at all times (try the esc key if you don’t know what I mean), besides which I use the f5 key to refresh anyways -> its just faster to reach.

    Check your internet options dialog for additional tabs options (hey, did you see the big blog entry about all the new UI stuff? I admit to missing a few things myself the first 12 times. You might wanna revist that one)

  21. Dark Phoenix says:

    >There appears to be a bug someone has documented on css > rollovers:

    > http://locusoptimus.com/css-trickery/ie7-hover-bug.php

    > Will this be addressed before the final version?

    I believe after a period of experimentation, it was found by the CSS-Discuss this that this bug has something to do with :hover and the @import statement. Speaking of the @import statement, was the Flash Of Unstyled Content ever dealt with?

  22. Dark Phoenix says:

    Oh, and sorry about double-posting, but I wanted to bring this up:

    http://www.meyerweb.com/eric/css/edge/slantastic/demo.html

    I believe the odd gaps that IE7 (and IE6 and IE5…) produce in the float code may be a result of this bug:

    http://www.howtocreate.co.uk/wrongWithIE/?chapter=Empty+Elements

  23. JS says:

    I would like the ability to shift-click the bottom right corner of the window and do a scaled shrink of the window. Double clicking the corner of the scale-shrunk or expanded window should spring it back to normal. Reason for this feature is so i can arrange the windows easier and also it helps when cutting and pasting. Not sure if this feature should be an IE thing or an OS thing. Would definitely use for IM windows as well.

  24. Robert Nyman says:

    I have no problem with the strict doctype vs. a non-strict one: a fair approach to me. But….

    > Add the most-requested new CSS functionality to IE

    The min-height/min-width and max-height/max-width has to be one of these requests. You can’t fix the broken box model/overflow problems without adding support for this.

  25. me says:

    Congrats again… Keep up the good work!

  26. Cascader says:

    Nice to see that MSIE team finally made steps towards improving CSS support. But still, in terms of CSS support, it stands behind Opera 5 and Mozilla 1.0 released several years ago. To be more constructive I am posting my wish list for future release:

    1. Generated Content (:before, :after preudo-elements, counters)

    2. Tables (display:table, table-row, table-cell, table-caption, inline-table)

    3. Paged media properties

  27. Mirek says:

    This really sounds great. I’m only still wondering if there will be box-sizing property. Hope so. Another important properties are min/max width/height.

  28. Gunther Schmidl says:

    I beg of you, fix the problem where typing "server.com:8080" without "http://&quot; gives you "Invalid URL". I really, really hate having to type http:// in front of such URLS by hand every time.

  29. David says:

    Thanks you for the good work.

    "a move to a more standard compliant implementation benefits everyone in the long run." I fully agree even if it’ll break my pages.

    Please don’t stop there and fix the min-height.

  30. CSS Warrior says:

    Recently released MSIE 7 beta 2 is &lt;a href=&quot;http://blogs.msdn.com/ie/archive/2006/02/02/523679.aspx&quot; target=&quot;_blank&quot;&gt;still very far&lt;/a&gt; from having decent CSS support. Good news is that most of CSS2 selectors were finally

  31. David Naylor says:

    max- and min-width would be nice… But I presume that’s a pretty code-heavy change to be coming in this late in the process.

  32. If you ask me, fixing the box model problem is going to be the number one reason that your customers sites break! After that, there’s no real reason for not including basic CSS2 properties such as :after, :before, content, max/min-height/width, display: table, outline, etc. None of those should affect a page currently coded for IE5/6, as there’s no existing support!

    Sorry, but I think you’ve fudged this release. This was a great opportunity to at least meet CSS2 standards; I understand that CSS3 isn’t an ‘official’ recommendation yet, but CSS2 has been for almost eight years!

    Instead you’ve put together a browser which will appeal to the average user, perhaps dissuading them from downloading Firefox or Opera, but without any real attempt to help the web progress.

    I’m pretty angry about this; after you’ve been so open with the community (using blogs, releasing betas, etc), the fact that you’re then going to betray us by not releasing a standards-compliant browser has really annoyed me.

  33. frankf says:

    So far, so good. I was pleasantly surprised when I found that background-attachment: fixed is fixed (as seen with the complex spiral here: http://www.meyerweb.com/eric/css/edge/complexspiral/demo.html), as well as fixing a lot of old CSS bugs. However, I still have reservations on a couple things:

    1. Floats still act strange as shown here: http://www.positioniseverything.net/explorer/floatmodel.html. And if you use the page zoom on this page for example: http://www.positioniseverything.net/explorer/dup-characters.html the text gets cut off by the the floating object.

    2. Will :before and :after be supported, particularly with content? I hope this gets fixed before IE7 is finally released.

  34. I have to add my weight to the requests for min-height. No :after and content: we might be able to live with, as I see IE7 follows the old IE Mac float-enclosing model (display: inline-block) and this doesn’t require hiding from other browsers, AFAIK. But many, MANY fluid layouts will be unfixable in IE7 without min-height. In the words of Frank N Furter, you’ve removed the cure but not the symptom.

  35. Dao says:

    > We made major changes to our box model to support overflow correctly on non-replaced block level elements and inline-block elements.

    Still work to do, this displays incorrectly:

    http://design-noir.de/bugzilla/ie7-overflow.html

  36. Mike says:

    I knew the day would come when sites built with browser specific hacks would break. That’s why I’ve avoided using hacks. It’s been very easy to do and still have no need to compromise design. I’ve been checking sites I’ve built over the past year (about 25 and not a table in one of them, all XHTML 1.1) and found for the most part rendering to be fine. I think you’ve done a fine job on 7 so far.

  37. Dao says:

    Properbly, there will be support for min/max-height/width.

    Maybe this should be added to the blog entry.

    http://blogs.msdn.com/ie/archive/2006/01/31/520883.aspx#522483

  38. David Naylor says:

    "I’m pretty angry about this; after you’ve been so open with the community (using blogs, releasing betas, etc), the fact that you’re then going to betray us by not releasing a standards-compliant browser has really annoyed me."

    We’ll just have to hope for a swift 7.5 and 8.0.

  39. Trackback from dotnetkicks.com

  40. Dao says:

    I also wanna add once more that I am still waiting for a comment on :before/:after. An answer like "yes", "no" or "we don’t know yet" can’t be too much to ask for.

  41. Aleksey says:

    Did you ever try test IE with Acid2 Test? http://webstandards.org/act/acid2/

    Opera 0.9 beta and Safari pass this test. Firefox and IE6 or IE7 – don’t.

  42. M. Schopman says:

    I am dissapointed, to see changing a class name still fires requests (not modified/304) to the server unless you have specified expiration headers on the webserver. In the past this caused enormous amounts of problems.

    It caused wininet.dll crashes when adding images dynamically with JavaScript, it made the browser hang/crash.

    The solution should be, in my opinion, so simple, and with an enormous rendement. Not only will page loading increase in speed, but this also prevents JavaScript generated components with images of becoming sluggish to load.

    I will also send this to the bug reporting mail address, but I just want to bring this to your attention, with the kind request of: please address this for IE7, please.

    My suggestion:

    When cache has been set to automatically, the initial request on a file should be automatically. Once that is done, every new request for that file in that page should use the cached version.

    When cache has been set to every visit to the page, the initial request on a file should cause a load of a new version of that file. Once that is done, every new request for that file in that page should use the cached version.

    I know in the past this has been answered by "This is caused by your cache settings", but that is a totally unlogical and unhelpfull answer. This is clearly a bug and these are the type of "simple" bugs which you earn points with very quick.

    I hope this gets adressed before the final/release candidate.

  43. seb says:

    Has border-spacing been addressed, at the moment i’m forced to use the cellspacing attribute?

  44. M. Schopman says:

    border:outline is not supported either, this would have been very helpfull in creating wysiwyg editors where you do not want borders to change the layout (like floats).

  45. Dao says:

    border:outline does not exist anywhere, but the outline property is really missing:

    http://www.w3.org/TR/REC-CSS2/ui.html#propdef-outline

  46. Bogtha says:

    @M. Schopman,

    The caching behaviour you describe would play havoc with some forms of Ajax.

    If Internet Explorer doesn’t support the Expires header properly, then *that* is a bug. But it seems you want Internet Explorer to ignore the Expires header and cache things for as long as the page is open. Given that every situation in which that is useful is fixed properly by setting a valid Expires header, I can’t see how you justify hobbling Internet Explorer in this way.

  47. Thomas Tallyce says:

    Dear IE team,

    What would be really useful now for a new posting is if you could list the problems with beta2 that you’re aware of and are planning to fix for beta3.

    Then we can all stop commenting on those issues you know are a problem while you are busy fixing them. I imagine the regression with #id a:hover is one of those.

    And I echo the call for implementation of the desperately-needed min-/max- width/height.

  48. Daniel says:

    What about :before and :after?

  49. Jason says:

    There is a graphical bug regarding the tooltip when you hover over text and images.

    It does not happen every time but if you hover your mouse over the a link or an image with a title or alt tag, the tooltip will appear. If you look closely a some rogue "pixels" obstruct the text. If you move the mouse a little, they disappear. It seems to be a rendering problem with the tooltips.

    P.S I would ALSO like to see the implementation of ‘display:table’. This is a key feature if we want to move away from table based designed websites.

  50. Hi,

    I don’t know if you cought my comment in a previous blog-post, but as this one is more on-topic I’ll repost it here. These are my findings of a couple of still-existing bugs in IE7’s CSS implementation (but first two suggestions):

    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.

    Bugs

    ====

    Furthermore I’m seeing several problems on my website: http://www.grauw.nl/ :

    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.

    [1] http://www.grauw.nl/articles/web/bottomfooter.html

    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;).

    ~Grauw

  51. Dao says:

    > 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.

    http://blogs.msdn.com/ie/archive/2006/01/31/520883.aspx#522483

    > 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.

    Would make sense. By the way, adding border[-bottom]-style:dotted or dashed results in a "nice" behaviour, as there is no default width in any direction. 🙂

  52. Seb Frost says:

    MS is working on min/max height/width, for those of you who insist on mentioning it in every post!

    "…min/max width/height properties, which are currently not supported. The change of the overflow behavior is the precursor to be able to support these properties. We are currently working on implementing these…"

    http://msdn.microsoft.com/library/default.asp?url=/library/en-us/ietechcol/cols/dnexpie/ie7_css_compat.asp

  53. Bogtha says:

    I too would like an answer for display: table-cell etc. I’m less concerned with whether it will be implemented or not, and more concerned with *knowing*. If you tell me that you aren’t going to implement it, then I can get on with "fixing" the sites that use it. If you tell me that you are planning on implementing it, I can sit back and wait for the next beta. But not knowing puts me in the position of having to do the work and then finding out I didn’t need to.

  54. Daniel Riggs says:

    Display:table, row, cell etc. That is one of the things I was looking for IE7 to support but this is a nice start.

  55. Jack Chapple says:

    I agree that :before,:after, and min width/height MUST simply work in IE7…!

  56. CecilWard says:

    In view of the fact that a CSS bug has been found which depends on whether a stylesheet is @import-ed or linked, ("a:hover not rendered when declared in an imported style sheet" – http://www.projectseven.com/csslab/ie7/import.htm), this means that people reporting test cases may need to duplicate test cases to get full coverage; even perhaps testing linked, @imported and embedded.

    Unfortunately, this adds a whole extra dimension of complexity to things when trying to replicating others’ results.

    So just be aware.

  57. RaulGonzalez says:

    Using CSS positioning to create a rectangular area and place a text element inside, in previous versions of IE the text element was rendered in the vertical center, no matter if it is a oneline or two line element. I have found that IE7 is rendering the same element below the middle of the element, using it as a baseline, this is my code:

    <style type="text/css">

    #hButton {

    PADDING-RIGHT: 0px; PADDING-LEFT: 0px; PADDING-BOTTOM: 0px; WIDTH: 100%; PADDING-TOP: 0px; BORDER-BOTTOM: #999 1px solid; HEIGHT: 49px

    }

    .bLink {position: relative; top: 50%; height: 1px;}

    .bText {position: relative; top: -50%;}

    </style>

    ….

    <div id="hButton" style="background-color:#CCCCCC">

    <div class="bLink">

    <div class="bText">Text<br>text</div>

    </div>

    </div>

    Any help will be very appreciated. Thanks

  58. I Hate It says:

    Sigh.

    And so my points from prior posts are proven.

    No Central Bug Tracking = Bug reports/fixes spread across hundreds of sites /w no real organization (though perhaps its only 10s of sites for the moment)

    Please do not leave things this way for the release version.

  59. I Hate It says:

    Oh. And looking back at some of the earlier posts, I think the "most" requested CSS upgrade was full CSS2/2.1 support. Not, please fix just a few bugs, but full support.

    Halfway measures are just asking for further disgust with IE. Though maybe we’ll get the CSS 2.1 support we desire in IE8 sometime in 2012, of course by then we’ll all want CSS4 support.

  60. MacK says:

    I can no longer FTP from IE. Is this a known issue?

  61. Alex Carvalho says:

    PNG-24 support is excellent but unfortunatly the CSS3 opacity attribute is not yet supported and the fallback filter:progid..Alpha does work properly when PNG-24 are involved…

  62. Josef Dunne says:

    Hi,

    Seems that IE7b2 is an improvement on IE6, however my most annoying issues with IE7b2 is the CSS :hover bug when using @import, this must be fixed for the final release, i saw the workaround, but who wants to put a redundent a:hover {} under their @import statement for the benefit of IE7? Nobody.

    The other thing that needs to be addressed is the min-width, max-width, min-height and max-height CSS properties, these will make alot of web developers lives easier, im sure you want that. Please implement these in the next release.

    Cheers.

    Joey

  63. Richard says:

    Add my vote for min/max width/height, display table (cell, row, etc) and :before and :after.

  64. ryan says:

    – still no min-width, max-width

    – floats still do the wrong thing when lined up against eachother.

    – no real changes to any of my STRICT doctype pages

    i am still going to have to double my development time to make a version of the site that works for EVERYBODY ELSE and a version that works for ie.

    thanks

  65. min-width, max-width, min-height and max-height says:

    min-width, max-width, min-height and max-height all needed.

    Also, for any generated element, regardless of nesting, and positioning (fixed/absolute/relative)

    DOM access to the *ACTUAL*

    obj.style.offset[Top|Left|Width|Height] is MANDATORY!!!!!!!!!

    Doing any serious DHTML without this is IMPOSSIBLE.

  66. cooperpx says:

    IE7 Beta 2 Scrollbar / Overflow bug test case

    http://cooperx.netfirms.com/Ie7b2_vscrollbar.htm

    This test case displays a scenario with no equivelent workaround: The dreaded horizonal scroll bar shows up whenever the vertical bar appears.

  67. 1. We currently parse :focus and then invalidate the selector since we don’t have support for matching this pseudo-class at current.

    2. No we don’t currently support the double colon syntax.

    3. Changing of attributes does a lot of work inside of the DOM. In general there are a class of attributes that trigger format recalculations and some that don’t. If they trigger the proper recalculations we should reapply the style sheets no matter how the attribute is set. Use the blog feedback if you have specific scenarios you are interested in.

    I’ll leave the rest for our CSS PM Markus!

  68. Markus [MSFT] says:

    Answering a couple common questions:

    – :before and :after will sadly not make it into IE7.

    – people reported that :hover would not work from an @import. That’s a bug and already fixed in our builds.

    – min/max we heard the request and working hard on it. Hopefully we make it for IE7.

    — Markus

  69. Auxiron says:

    check out my rollover buttons on my website… they dont’ work correctly in IE7.. they disappear on rollover… had no problems in IE6 though… they are at

    http://www.aerocet.com/mambo

  70. Dutin Wilson says:

    I’m not impressed with IE7’s rendering of webpages. It does do a good job of rendering what it does support. That is a big improvement over previous versions of IE.

    Support for display:table,display:table-cell, and display:table-row are mandatory. This is needed to create websites without table layouts.

    Other CSS things like min-height, max-height, etc. need to be supported. These are getting used more and more because most alternate browsers to IE support them.

    IE7 still doesn’t support PNG files correctly. It does transparency, yes, but color gamma is incorrect. Using a program like pngcrush to remove the gAMA chunk from PNG files are still required to view the colors correctly on IE 7.

  71. The usage of underscores in the style="" attribute was an oversight that has been caught and remedied. Thanks for taking the time to test this out though as it helps us prioritize additional work.

  72. IEBlog : What’s New for CSS in Beta 2 Preview? Compatibility and our updated CSS behavior Obviously, we have heard the feedback asking us to be more standards-compliant in our rendering behavior. We must balance this ask with the need of our customers

  73. Dave says:

    To add to what Dutin said above this post, at http://entropymine.com/jason/testbed/alphagamma/ there is a good example of the incorrect gamma display in IE7.

  74. Mike Romero says:

    Great stuff.. I’m about to embark on a site redesign and its good to know that I can focus on creating valid CSS and not worry about workarounds 🙂

  75. Innocent Bystander says:

    If you are only going to activate the bug fixes for strict mode, you are doing nothing for the vast majority of sites out there. We want the fixes in QUIRKS MODE!

    In fact, because IE 6 uses a different default style sheet for strict mode than quirks mode, the majority of websites INTENTIONALLY INTRODUCE ERRORS to FORCE QUIRKS MODE to get the stylesheet most people are used to seeing.

    So, this design decision you have made is VERY POOR. Please fix it and enable the more compliant behavior for quirks mode.

  76. Jack T says:

    http://en.wikipedia.org/wiki/Comparison_of_layout_engines_(CSS)

    Please fully implement CSS 2.1 . pleassssssssssssssssssssssssssse

  77. Marek Materzok says:

    I’m quite happy with the changes in IE7. My sites render a lot better now – with hacks disabled. I think, that the next big things to implement are:

    * min-/max- width/height

    * display:table

    Of course, the target should be full CSS2 compliance.

  78. ieblog says:

    Auxiron, your page worked in IE6 because it was getting thrown in to Quirks mode because of the XML prolog (<?xml…>).

    -Chris Wilson [MS]

  79. Wendal (Stepping up to the plate) says:

    Anyone here, have the resources (server, bandwidth, good-heart) to open up a public bug-tracking system for IE7?

    (and yes, I am hoping MS would, but I’ve been a Web Dev long enough to know I shouldn’t hold my breath)

    So, if anyone has the resources (I for one, would be willing to contribute (financially via paypal), and (technically, via bug reports/test cases/fix updates)) I would be very interested in getting this going.

    The industry is in obvious need for this, but I’m really losing faith in this coming to fruition. (e.g. We don’t even have a firm commitment to the cause from MS yet.)

    I have about 1-200 test cases of HTML, JS, and CSS bugs for IE6 (25% or so are now fixed) to contribute for the record, I just need to be able to host them for all to share/search etc.

    Thanks.

    Wendal

  80. Suyi says:

    There should be an IE plugin(activeX) that would allow people to run FF extensions for IE with same functionality (or atleast close). Is there?? is there?? i know theres no such plugin… but that would have been so cool!

    <wink>hint</wink>

  81. Suyi says:

    oh hey. srry. wrong section.

  82. ThomThom says:

    Will pages we thrown in quick mode with an XML prolog in IE7?

  83. Rob says:

    Just wanted to say thank you to the IE team for the work you’re doing. There’s a lot of people on here who just moan, a lot more people on here who give constructive criticism, and those people know more about the various bugs than I do but I just wanted to say thanks. As a web developer I really appreciate all you’re doing to improve things.

  84. Shining Arcanine says:

    Please consider adding support for:

    display: inline-table;

    display: table;

    display: table-row;

    display: table-header-group;

    display: table-row-group;

    display: table-footer-group;

    display: table-column;

    display: table-column-group;

    display: table-cell;

    display: table-caption;

    The first one would let me pull an IE specific hack I am using right now and the others would make it easy for me to use table-style positioning in places that do not contain tabular data.

  85. Ingo Chao says:

    Markus,

    You have fixed the

    /**/ comment-bug

    but not the

    !important-bug

    by doing this, you are breaking many old pages that rely on the alternate box model hack.

    This abmh safely assumes that modern browsers would properly implement !important.

    This hack would not break if !important would be correctly implemented in IE7. The abmh addresses old browsers with the old box model solely.

    test

    http://www.satzansatz.de/cssd/ie7b2_abmh.html

    link to the abmh:

    http://www.info.com.ph/~etan/w3pantheon/style/abmh.html#tech3

    The "Call to action" goes in both directions: Once you have started fixing bugs, you can’t stop half way.

  86. Paul Bellows says:

    OK. Wow. Whatever the hell you’re doing with font rendering is brilliant. I love it. Not just pretty, but practical-pretty because it is actually easier to read.

    But I can’t tell how text-sizing is being calculated now. I’ve got a mix of sites behaving pretty erratically, and some showing text at about 80% of what I’d expect.

    I’ll spend a little time trying to boil it down so I can ask a reasonable and well-researched question (wouldn’t that be a novelty in this list of posts) but I’m wondering if this is something that you’ve documented somewhere? I read through everything I could find in the docs (again, a novelty in this list apparently) but I can’t seem to see anything about the rather striking new rendering of text in IE7.

    Really nice work, thank you. Some of my concerns have been very thoroughly addressed, and the overall UI and feel of the browser is so well thought out that I hardly recognize it. I love the RSS discovery and can’t wait to explore some of the underpinings of it.

    Thanks, and please let me know about type/font sizing.

  87. Markus [MSFT] says:

    @Ingo

    We have fixed the !important bug in our recent builds (past B2P)

    — Markus

  88. Ingo Chao says:

    @Markus

    Great!

    Updated the now obsolete demo.

    Thanks.

  89. M. Schopman says:

    To boghta, the xmlHttpRequest is a new request, and in the current situation such requests are always cached for 60 seconds, unless you add a new Date().getTime() function in the URL.

    This is different from re-requesting images while loading a page. Gecko, Opera, Khtml, all do it correctly. Why wouldn’t this be possible for IE 7?

    There are better ways to handle such issues, than just reloading files from the server on each get request with such cache settings.

  90. cwilso says:

    Wendal, we’re actually already working on this (a public bug-tracking system for IE bugs) – we’d hoped to have it set up before we shipped the Beta 2 Preview, but it didn’t quite clear in time. Stay tuned for more from Al Billings.

    -Chris Wilson [MS]

  91. Christian Augustin says:

    Hi there,

    I hadn’t the chance to test yet, but I read the posts and thought about some points. First: No support for :before and :after isn’t funny, because even if IE7 tries to behave like a CSS 2.1 compliant browser, it isn’t. So many developers have to support 3 (!) different stylesheets in the future: One for IE 5-6 in quirks mode, one for IE 7 and one for the truly conformant browsers …

    One solution would be to force IE 7 into quirks mode (yes, I would like to do so so that I can spare an additional style sheet).

    But the problem is: The one trick to force quirks mode is no longer available — using the XML declaration with strict DOCTYPE.

    What about a meta tag (like the one to hide the image toolbar) to force IE 7 into quirks mode if needed? This one could help developers to have modern, XML style XHTML and at the same time have IE7 behave like its predecessors (until IE 8 will arive and support such oldies like :before and :after …).

    And the horizontal scrollbar problem mentioned … I hope it will be gone with the final release. This one was one of the main reasons I decided to force IE 6 into quirks mode by using the XML declaration … (see above 🙂

    That’s all for now. I would appreciate support for :before and :after …

    Kind regards

    Christian Augustin (from Berlin, Germany)

  92. Mystere says:

    I’d also like to add my request for min/max-height/width support, also the fixing of the current width/height functionality so that it doesn’t behave like the min-* functionality.

  93. I have to admit i am a little disappointed. I was hoping for :after and :before. How hard can this be to implement?

    Although, i am over the moon to hear that min/max height/width will be implemented in IE7

    I know i am probably being a moany old git, but would like to see :after and :before, also would like to see table-*

  94. Markus [MSFT] says:

    @Mystere

    Our overflow fix mentioned in the blog exactly addresses that widht/height doesn’t behave like the min-* functionality.

    — Markus

  95. Sean McBride says:

    Hey IE Team,

    Thank you so much for focusing on CSS compatibility (at least more than before) in this next release of IE. I hope that you can have full CSS2.1 compliance before the first real release.

    I have one question, as this has consistently confused me ever since I first started hearing about it. I keep hearing that some hacks used to aim CSS fixes at IE6 will now be fixed, so developers will need to fix their pages. Why would this require fixing? If the bugs used in the hacks have been fixed, then the IE6 specific rules shouldn’t be added and the page should render just like in other browsers (i.e. correctly.)

    The only reason I can think of why developers would need to "fix" anything is that the bugs they are using these hacks to fix will still be present, so we’ll have to find OTHER ways to fix them. This isn’t really "fixing" at all. It’s actually just spending more time trying to "rehack" IE7 CSS, a potentially long and arduous process.

    It would be much better if CSS compliance for IE7 could be brought up to around the level of Firefox or comparable OS browsers. None of the CSS hacks aim at a specific browser, so if all the bugs are truly fixed, pages should render just the same as in any other compliant browser right? Problems will only arise for developers if ONLY SOME of the bugs are fixed…

    This more than anything is the reason why you should continue to push for full CSS2 compliance. I’ll be cheering you on.

  96. Avi Lapidus says:

    Why have you slowed down the progression of the web by not fully implementing CSS2?

    You double my workload as a web developer because I refuse to use table layouts, and am forced to work out the many bugs and incapabilities that are caused by IE. My website has conditional comments that contain table tags, which generates extra bandwidth and slower rendering speeds for even my site’s visitors that use Firefox and Opera, browsers that have supported CSS2 for years.

    If IE7 will not fully support CSS2, I beg you to please, please, please support display: table/table-row/table-cell by IE7’s release. And on behalf of the entire standards-supporting community of web developers, I demand you to fully implement CSS2 as soon as possible!

    PS. Please respond whether or not you plan to implement the CSS display properties by IE7’s official release. Thanks.

  97. Markus [MSFT] says:

    @Avis

    We will not support display:table properties in IE7

    Just to reiterate:

    We never said we would have full CSS2.1 compliance by the time IE7 ships. We are committed to the standard and working hard to make progress.

    For more info check out Chris blog post: http://blogs.msdn.com/ie/archive/2005/07/29/445242.aspx

    — Markus

  98. Dark Phoenix says:

    Well, I’ll mention one that IE partially supports but needs a bit of an update.

    What about inline-block according to CSS2.1? IE6 only supports it on elements that are inline by default.

  99. Helpful Student says:

    Hello,

    I’m sorry that this is off topic, but I recently downloaded and installed Internet Explorer 7 beta 2 from the official Microsoft Internet Explorer website and noticed when I went to visit the Website of Carthage College (www.carthage.edu) Internet Explorer 7 was unable to compose many of the pages on the website properly where as it’s preticessor could. Is there any information that could be made available to the webmaster of Carthage College in order to rectify this situation? Again, my most sincere apologies for this obviously miss-posted message.

  100. Christian Augustin says:

    A follow-up to my previous post. After thinking about :before and :after I stumbled about this one:

    If IE 7 in strict mode will handle floating boxes like standards compliant browsers, there will arise a big problem for many, many sites which rely on semantic column layouts using only divs floating left/right. Why? Because they use code like this one to force the enclosing div to the height of its enclosed boxes:

    :after {

    content: "";

    border-top: 1px solid transparent;

    clear: both;

    }

    (or similar, this is very often needed with Firefox etc.). And if IE 7 handles floats like Firefox and Safari, but does not support :after (and :before for similar purposes), those sites will break even if they use standard CSS 2.1 in a rather defensive way …

    Once again, the solution would be to force IE 7 to quirks mode (in IE 6, this was no problem because the enclosing box did not collapse) — but the XML declaration no longer forces IE 7 into it. So with modern XHTML code many sites will break. I think most developers thought — just like me — that a browser which supports CSS 2 rendering would also implement :before and :after (what most other browsers do a couple of years now). Okay, there could be a possibility to trick IE 7 with dynamic properties or an HTC, but this would be a really ugly solution (and would fully rely on JavaScript enabled by the user).

    So I repeat my suggestion to give us a possibility to intentionally and directly force IE 7 to quirks mode with strictly coded pages and without the need to play with DOCTYPEs and XML declarations. The suggested meta tag could be such a solution, or something like the conditional comments (which are really usefull and I do like) or whatever you would like to implement. But please don’t continue to misuse DOCTYPE and XML declaration for this purpose. Or support CSS 2.1 with all its necessary ingredients …

    (One more reason to force IE 6 to quirks mode was to get things done the same way in 5.0, 5.5 and 6.0 so that all three could be handled with just 1 additional stylesheet included by a conditional comment 🙂

  101. Dao says:

    I extended my testcase and added screenshots:

    http://design-noir.de/bugzilla/ie7-overflow.html

    http://design-noir.de/bugzilla/ie7-overflow.png

    http://design-noir.de/bugzilla/ie7-overflow-zoom.html

    http://design-noir.de/bugzilla/ie7-overflow-zoom.png

    As you can see, applying the proprietary zoom property somehow fixes the behaviour partially. I was told this would be because of the "hasLayout flag" — maybe you know what this means, I don’t.

  102. Mystere says:

    Sean Mcbride,

    The way I see it is this. IE7 will not fix all bugs, nor be fully CSS compliant when it’s released. However, it will fix a lot of them. Sites that rely on hacks for bugs that IE7 will not fix will break, but sites that rely on hacks for bugs that IE7 fixes will not.

    This isn’t nearly the panic situation that many developers think. IE supports conditional comments, which would allow you to include a specific IE7 style sheet. In my opinion, this is vastly superior to hacks anways, as it’s a supported feature, and doesn’t rely on breaking HTML standards.

    In my opinon, hacks are obsolete, though still necessary for legacy browsers.

  103. Very upset web developer says:

    Markus,

    I read in horror that you do not plan on supporting display:table, display:table-cell, display:table-row in IE7. This is completely unacceptable because it is a basic functionality required for web developers to stop designing pages the same way we’ve been doing since 1997!!!

    PLEASE STOP HOLDING BACK THE WEB! You won’t believe how angry this makes all of the developers who need to dumb down not only their HTML but the underlying Javascript, CSS, ASP / PHP just because IE *is the only browser* that can’t break away from the ancient table-based layout. Do us all a favor and release a browser that is actually functional this time. We’ve only been waiting for the past 10 years…

  104. Mystere says:

    Upset web developer,

    As a web developer with several years of experience, I have to disagree. While display:table would be nice, it’s certainly not a requirement to design table-less designs.

    Your problem is that you can only envision your design as a table-design, and rather than rethink your design patterns, you simply want to try and make div’s into tables using CSS.

    While that will work (on any browser that supports display: table), even if IE7 supported it, you wouldn’t be able to use it for years because legacy browsers that don’t support it will take that long to go away.

    Break outside the boundaries of table thinking. You might be surprised by what you can do.

  105. Very upset web developer says:

    Mystere,

    Without display:table you are very limited in multi-column designs, and most of all you need to jump through hoops just to do something extremely simple.

    1. With display:table you can make 2 or more columns, set a background image for each, and either set a static width or they will automatically adjust their width to fit the content they contain as well as being equal height. This is the ideal multi-column layout that cannot be accomplished with IE4/5/6/7’s current CSS support.

    2. All of the "faux columns" designs we are forced to do require an extra wrapper div as well as a custom made background-image in order to look like there are equal height columns. Now if you want to make changes to the colum widths you not only have to change the CSS but have to load up an image editor and change the image (hack) too. In addition one of the columns is required to be a fixed width so it matches the background-image of the wrapper. It is just ridiculous all of the trickery you need to apply just to get a simple layout right now!

    3. So yes you can get somewhat close to the same effect of a table-based layout with IE’s current CSS support but it is not very straightforward for developers, it is limited in flexibility because it cannot accomidate changing widths of the content inside, and it goes against the whole point of CSS being able to easily control the style of the document because you need to change the width of the background-image of the column!

    >>While that will work (on any browser that

    >>supports display: table), even if IE7

    >>supported it, you wouldn’t be able to use it

    >>for years because legacy browsers that don’t

    >>support it will take that long to go away.

    With Microsoft’s new stance on security, IE7 will most likely be heavily marketed and pushed by Windows Update so I expect uptake to be very quick. People still running 95/98 are likely in the market for a new PC by the time Vista comes out which will include IE7.

    >>Your problem is that you can only envision

    >>your design as a table-design, and rather

    >>than rethink your design patterns, you

    >>simply want to try and make div’s into

    >>tables using CSS.

    >>Break outside the boundaries of table

    >>thinking. You might be surprised by what you

    >>can do.

    Actually that is not my problem. Designing a page layout shouldn’t require floats, negative margins, background-image hacks just to accomplish a simple 2 column layout. I’ve been developing for over 5 years and have tried every CSS column layout published as well as spending hours on my own approaches and you kow what? Sometimes the best thing to do is wrap your layout in a table if you want a true multi-column solution because there is no other way to do it in CSS without display:table support. Think about all you have to put yourself through right now to do such a simple layout and its hard not to be upset about it.

  106. @Christian Augustin:

    If you really, really wish to use get IE7 into quirks mode, you can still do it with a comment before the DOCTYPE – at least in IE7B2.

    A ‘visual’ testcase to confirm it: http://dev.skalske.dk/css/comment-before-doctype.html

    @IE team:

    I think it would be really nice if you could add support for activating focus on a form element, placed inside a label element, when clicking on the label. Right now this doesn’t work, unless you’ve linked them together using the for-attribute and an id:

    http://dev.skalske.dk/css/inputInsideLabel.html

  107. Shining Arcanine says:

    Microsoft, when Internet Explorer 7 Beta 2 is done, will it be publically released? I am anxious to see what has been done in response to the feedback from Internet Explorer 7 Beta 2 Preview.

  108. Hmm, I wonder why on earth I ever placed the last test-case is in the css-folder… Then this was the wrong thread to mention it in, sorry.

    Nonetheless, it would still be really nice to get support for it though, and somehow I imagine it’s not that hard to implement?! //End of spam.

  109. draco says:

    when will the button element work? currently it doesnt work according to http://www.w3.org/TR/html4/interact/forms.html#edef-BUTTON

    IE7b2 is sending the wrong value to the server — the enclosed text that’s supposed to be displayed instead of the value attribute.

  110. Christian Augustin says:

    @Allan Rasmussen

    Thanks for the "Comment before DOCTYPE" tip (a hack one again?). I will check this (if it works with XHMTL 1.0 strict and XML declaration). And I have to check if it is fully syntactically correct (my BBEdit Syntax Checker does not complain).

    But this again would be a misuse and it is not guaranteed that this behavior will last to the final (perhaps it is seen as a bug?). Better to have a simple, direct and ease to recognise way to get IE 7 behave like IE 6.

    And to table-*: Yes, this would be nice. But not necessary to keep on working without troubles.

    My concerns are mainly for compatibility with those CSS 2.1 features which are widely in use now (because they did not interfere with MSIE 5-6) and which will cause websites to break if IE 7 does not implement them.

    It is one thing to avoid hacks (and it is easy to incorporate conditional comments instead), but a total redesign because IE 7 leaves out essential features of CSS 2.1 isn’t funny …

    (Oh, it would be soooo easy to do M$ bashing here … but I try to keep calm 🙂

    By the way: Dear Microsoft Developers — keep on your work, you are on the right track, I think. But please don’t stop halfways because you have tight deadlines (better release 3 months later) …

  111. @Christian Augustin:

    It works, and it’s valid. But surely, it relies on a bug that should be fixed – and I would never use this hack. But then, nor have I ever had any desire to put IE into quirks mode…

  112. Bob Easton says:

    min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height min-width min-height

  113. Mystere says:

    Upset web developer,

    I still disagree with you. Your anger comes from trying to implement multi-column designs, but display:table isn’t really the appropriate semantic use this. You’re just turning divs into tables.

    You might want to check out the most recent alistapart article:

    http://www.alistapart.com/articles/holygrail

    Yes, it includes nasty hacks like negative margins, but it seems to work pretty effectively to solve the problems you claim can’t be done.

    In any event, my point is that using display:table to create multi-column designs is just as hacky as everything else. It’s just an easier hack. The real solution is the CSS3 ‘column’ attribute, of course this isn’t a final standard yet.

    display:table creates a semantic state in which you create tables that aren’t really tables. There’s no tabulation going on. Yes, this helps with screen readers and such that look at the semantic markup of the HTML, but such devices are becoming more and more intelligent and are starting to look at CSS as well.

  114. I’m delighted to see so many people here begging for "min-height" (et al.) to be supported. I’d like to add the weight of my (quite loud, as all my friends tell me) voice to this plea!

    It’s surely one of the most useful and essential properties in the entire CSS recommendation!

    And regardless of that, my sites break in IE7, and as far as I can see the only thing that will fix them is min-height… I’d been getting round that in IE6 by taking advantage of properties that didn’t render correctly, but now that they DO, we *need* min-height!

    Please, please see sense, and make youselves a nice little spot in history by implementing this (and the rest of CSS2.1 for that matter, it can’t be *THAT* hard;) )

    I’m loving IE7b2 btw, impressed with lots of the new features and behaviour etc., and I’d like to see this improvement continue right up to (and beyond) the full release.

  115. Dao says:

    > In any event, my point is that using display:table to create multi-column designs is just as hacky as everything else.

    Nonsense.

    > It’s just an easier hack. The real solution is the CSS3 ‘column’ attribute, of course this isn’t a final standard yet.

    No. "column" addresses continuous text, not the page’s main layout.

  116. Mystere,

    So, you use position: absolute for your site layout?

    Because obviously using floats to create layouts is even more so of a hack than using the table- properties for that.

    ~Grauw

  117. I too would be very disapointed if min-height and min-width weren’t implemented. I use those a lot and keep stumbling on the IE-implementation.

  118. Very pleased to see the support for fixed positioning. Unfortunately it appears to still have quite a few bugs. I caught one bug in my application, and while trying to make a simplified testcase, discovered two more as well!

    The testcase is available at http://sab39.dev.netreach.com/fixed.html

    The three bugs in short: the horizontal scrollbar isn’t taken into account by fixed positioning; trying to position to the far right leaves a one-pixel gap instead; and the text in my fixed-position div overflows its borders.

    The first one in particular is almost impossible to workaround (especially since it disappears when you resize the browser window) and I’d really like to see it fixed for IE7 final.

  119. CecilWard says:

    1) IE7b2p seems not to understand that whitespace in a CSS selector means "descendant of"; see test case 3b below.

    That is, the selector,

    [xmlns][lang]

    is not the same (at all!) as

    [xmlns] [lang]

    2) The CSS parser/lexer seems to be in a weird state _after_ encountering the new attribute selectors. See test cases where the *.class and .class are not equivalent below.

    This needs hundreds and hundreds of test cases, programmatically generated. I just put together this set quickly by hand, and my poor effort is woefully inadequate. Others will, I’m sure, find many more bugs in this area.

    Regards,

    Cecil Ward.

    ==

    Test cases:

    <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"&gt;

    <html xmlns="http://www.w3.org/1999/xhtml&quot; xml:lang="en" lang="en">

    <head>

    <title></title>

    <style type="text/css">

    p {

    display:inline; margin:.25em;

    }

    [lang][xmlns] .c1a, [lang][xmlns] *.c1b, [lang][xmlns] p.c1c, [lang][xmlns] * p.c1d, [lang][xmlns] body p.c1e,

    [lang][xmlns] #i2a, [lang][xmlns] *#i2b, [lang][xmlns] p#i2c, [lang][xmlns] * #i2d,

    [lang] [xmlns] .c3a, [lang] [xmlns] *.c3b, [lang] [xmlns] * .c3c,

    [lang] [xmlns] p.c4a, [lang] [xmlns] #c4b

    { font-style:italic; color:black; }

    </style>

    </head>

    <body><hr />

    <p>All following should be italic: </p>

    <p class="c1a">1a.</p>

    <p class="c1b">1b.</p>

    <p class="c1c">1c.</p>

    <p class="c1d">1d.</p>

    <p class="c1e">1e.</p>

    <p id="i2a">2a.</p>

    <p id="i2b">2b.</p>

    <p id="i2c">2c.</p>

    <p id="i2d">2d.</p>

    <p><br /></p>

    <p>All following should not be italic: </p>

    <p class="c3a">3a</p>

    <p class="c3b">3b</p>

    <p class="c3c">3c</p>

    <p class="c4a">4a</p>

    <p id="c4b">4b</p>

    </body>

    </html>

  120. CecilWard says:

    I suspect that the CSS selector

    html[xmlns]body #myId { }

    is illegal. (Lack of whitespace before ‘body’). Well the W3C seems to think so. (A mess.)

    Whereas what we probably should have written is

    html[xmlns] body #myId { }

    IE7beta2P seems to accept the first example, when I suspect it shouldn’t. It also does the right thing with the second form.

    Anyway, Firefox 1.5 and Opera 9 beta junk the entire ruleset when shown the questionable first "no-whitespace" example, and do the right thing with the second form.

    Regards,

    Cecil Ward.

  121. Erland Flaten says:

    agh, so irritating that min-height min-widt is not there. Snatch code from webkit.opendarwin.org and stop acting up, please. I am so tired using dev time fixing up incompatible stuff like bad CSS-support etc. I want to use my time on design, UI, content consept etc.

    But the tendence is right but this goes soooo slow.

  122. I haven’t read everyone’s comments yet, but I just wanted to know if the IE devs realize how much their work on IE7 matters to the community. The decisions they make, with regards as to what is and isn’t supported in CSS affect every web devloper on the planet in a very large way.

    People here are begging for min-height and such. It’s in the spec, but if the IE guys decide not to put it in, then it effectively is not a usable attribute for web developers. I plead with the IE team to take several extra months and create a browser that can handle CSS as well as the Gecko based browsers do.

    Nobody is going to remember it came out in June instead of September, but for years to come web developers will be stuck with whatever choices the IE devs made. I somehow doubt we’re going to see new CSS support after IE7 is out until HTML 2 and CSS3 are ready — and even then it will be years.

    It’s a tough position to be in, but please take the time to do it right. And please, let’s see some min-height and content CSS support!

  123. //IE7 Beta 2 Scrollbar / Overflow bug test case

    http://cooperx.netfirms.com/Ie7b2_vscrollbar.htm

    This test case displays a scenario with no equivelent workaround: The dreaded horizonal scroll bar shows up whenever the vertical bar appears.//

    This is exactly what I was talking about in my post. Basically what’s happening is that the overflow isn’t causing a recalc of the layout of the contents of the DIV.

    Please please please fix this. It happens all of the time and makes for really crappy layouts.

  124. Tom says:

    This is annoying. There’s the problems with min-height and -width still?

    That means we have to do this:

    <div id="content">

    <p>stuff</p>

    <!–[if IE]><p id="stupid-ie"></p><![endif]–>

    </div>

    #content {

    min-height: 10em;

    }

    #stupid-ie {

    min-height: 10em;

    }

    It’s just ridiculous.

    @Mystere

    >In any event, my point is that using display:table to create multi-column designs is just as hacky as everything else. It’s just an easier hack.

    No, it’s not a hack. It describes presentation, which is CSS’s job, whereas using actual tables describes presentation also, but that’s not HTML’s job. displau:table is the right way.

    >The real solution is the CSS3 ‘column’ attribute, of course this isn’t a final standard yet.

    That one is not going to help for page layout. Even for long blocks of text it would be extremely annoying.

  125. Nick Presta says:

    I would be willing to host a IE7 bug/list repository.

    If someone can get in contact with me (via IM or email), I can give a subdomain with FTP access + email account and we can start grouping these bugs instead of having them spread out all over the web.

  126. Michael Sheakoski says:

    I think it is really important for everyone in here to really push for IE7 supporting the much needed CSS2/2.1 standards which were established over 8 years ago!  I am mainly speaking about adding support for:

    1. min-width / min-height

    2. display: table / table-cell / table-row

    3. table-* if you are feeling generous

    The entire internet is based on standards which evolve throughout the years and if you are going to release a browser used by the majority you have a responsibility to follow them otherwise you end up holding back the web. I’m sure you’ve heard the saying that one satisfied customer tells 5 of their friends about their good experience, but one dissatisfied customer tells 20 strangers about their bad experience. Why do you think there is an exodus to Firefox and Opera. We are your customers, we buy your software, we develop for your software, we are begging you to support at the very least an 8 year old standard (CSS2/2.1). We don’t care if it delays things for a few months to produce a browser which meets the minimum requirements for current standards, we care about having to write our pages once and not having to do hacks or workarounds to support Internet Explorer!

  127. Nick Presta says:

    Yes. (in agreement with the above poster)

    The last thing we (developers) need are hacks for IE < 7, hacks for IE7 and then the standards compliant CSS for all other browsers.

  128. Gábor Hidvégi says:

    @Mystere

    I must agree with you, it seems a little bit weird idea to implement a table with divs and css just because of the trendy tableless designs…

  129. Ben Cooke says:

    The point of the CSS table properties is to replicate the *visual* characteristics of a table. There is no implication that the element being rendered as a table is any kind of tabular data because CSS does not deal in semantics, only appearance.

    Applying the CSS table rendering model as specced is no more odd than using the CSS box model; it’s just a defined set of rules for layout which lend themselves to creating grid-based designs rather than the flowing box designs the standard box model is good at.

    You pick the layout model that best fits the layout you want to create. It has nothing to do with tabular data.

  130. Claire says:

    Bug Report:

    not sure if this has been mentioned/reported yet, I tried weeding through the examples

    A variation on an IE6 bug but with a new twist for IE7B2

    hasLayout and lists, if the list is in a table cell and has text alignment set to anything other than left the text disappears.

    Demo: http://www.tanfa.co.uk/css/examples/haslayout/lists-haslayout-ie7-bugtest.asp

  131. Glad to see a lot of effort in improving CSS selector support, and to see that you’re going to add the !important keyword to adhere the specs.

    I’d like to report a quite strange behaviour related to attribute selector. The following css rule:

    div[title="test"] em{background: #FF0}

    will make the whole <code>div</code> in yellow instead of the single <code>em</code> element.

    Keep up the good work.

  132. haka says:

    Your site "Cascading Style Sheet Compatibility in Internet Explorer 7" is nice, but not enough. What about the redering differences between IE 7 and Opera, Firefox. I found some, but wont pick them all by try + error 🙁 To fix them there are only conditional comments, no "css conditional comments".

    Please give us more information.

  133. Nick Presta says:

    Seriously, IE dev people, look at this chart and make IE up to standard.

    http://www.webdevout.net/browser_support_css.php

    http://www.webdevout.net/browser_support.php

    http://en.wikipedia.org/wiki/Comparison_of_layout_engines_%28CSS%29

    That’s all we want. If IE7 was as up to standards as the current FF, Opera, Safari and Konqueror, we would be happy.

  134. Al says:

    Good job (with the standards)!  Now keep it up!  The more standards-compliant the better.

  135. Mystere says:

    To those mentioning that column is only for multi-column text, I disagree.  Appropriate use of column-break attributes creates mulitple columns with specific formats.  According to the latest specification, they can contain block level elements and act like table cells, block level elements, and inline-block.

    there is no difference between multi-column text and multi-column layout in this regard.

  136. Dao says:

    Mystere, column is only meant and only useful for multi-column text. This text may contain block level elements, yes (of course!). But you are not able to address one particular column in order to make it narrower or wider: "All columns of a content box have the same column width and column length." <http://www.w3.org/TR/2005/WD-css3-multicol-20051215/&gt;

    I cannot imagine how anyone could create a site layout with this.

  137. konga says:

    It’s simply critical, that you add support for min/max width/height otherwise one won’t be able to make proper use of the fixed float-property.

    "height: auto" would also be deeply appriciated.

    Can’t bear the thought of having to wait for IE8 to get this fixed… It wouldn’t be fair…

  138. Tom says:

    >That’s all we want. If IE7 was as up to standards as the current FF, Opera, Safari and Konqueror, we would be happy.

    Or even just close to those standards. People won’t care if the release is delayed as long as you make it worth the wait.

    >To those mentioning that column is only for multi-column text, I disagree.  Appropriate use of column-break attributes creates mulitple columns with specific formats.  According to the latest specification, they can contain block level elements and act like table cells, block level elements, and inline-block.

    Multi-column text is used via the column-count and column-gap properties. There’s nothing for acting on a specific column. They are for making things like newspaper articles. It is not possible to layout a page with them.

  139. Tom says:

    Whoops, I forgot column-width.

  140. mystere says:

    Tom,  Yes, you can select a specific column, by using column breaks.  You can force column breaks to send content to the next column.  No, you can’t randomly position code in columns from anywhere, but i’m not really sure if that’s a good idea anyways.

  141. Shining Arcanine says:

    While this might be evidence for the French being strange, the French police are switching all of their 70,000+ PCs to Firefox because IE does not properly support the W3C Specifications:

    http://news.yahoo.com/s/afp/20060205/tc_afp/franceinternetsoftwarepolice

    Microsoft, you might want to reconsider not fully supporting Acid 2 and CSS 2 in IE7.

  142. mocax says:

    How do I disable scrollbar styling (eg. scrollbar-face-color)?

    It’s annoying as hell…..

    I want all my scrollbars to remain consistent across the entire OS…..

  143. CecilWard says:

    The layout in the attractive site http://www.thewatchmakerproject.com/journal/300/ie7-beta-2 appears very badly mangled.

    This doesn’t appear to be simply due to CSS filters not working. Lack of :after support means that easy clearing won’t work, but there’s more to it than that. Needs looking into further at a later date.

    IMHO it simply isn’t an option for Micrsoft to fix bugs in or add support for ‘half’ of a certain ‘logically related’ group of features. A perfect example of the kind of thing that will, or ‘ought to’, prevent IE7 from shipping, this site might serve as a worthwhile ‘use case’ to help work out whether the proposed fix list for IE7 makes sense.

    It might be an idea to build a list of such ‘challenging but not incorrect or evil’ real-world sites, and use it to drive the process, which I’m sure that the MS team have already done. But more probably couldn’t hurt, and would give an idea of what the community expects. Removal/rework of any hacks and filters should be a requirement for a site to make it onto such a list.

    Cecil Ward.

  144. Gamer says:

    This may or may not belong here, but I couldn’t find a better place to ask…

    Was it just me, or did IE7 take a huge toll on system resources?  I specifically noticed this, not in IE itself, but in gaming (specifically World of Warcraft, among others).

    I could play several games just fine when I had IE6 installed.  After installing IE7 and rebooting, there was enough drag on the system that several games froze for about a half second every 4 or 5 seconds.  Even without any IE windows open- merely having 7 installed seemed to drag me down significantly.  I tried several things to optimize my system, but the only thing that worked was going back to IE6 (and I’m running an Athlon64 3400 with a gig of ram, so it’s not like I’m playing on an antique or anything).

    I’m afraid I don’t have hard numbers to back anything up, as I was in a hurry when I decided to try removing 7, but when it was installed, several games were virtually unplayable, due to the slow frame rates and general system drag that IE apparently caused.

    I realize that this was a beta, but any idea if the final release will exhibit the same behavior?  If so, I’d imagine a lot of the gaming crowd won’t be installing…

  145. Gamer says:

    In retrospect, the above comment is indeed misplaced- somehow I missed that this particular thread deals exclusively with CSS.  My apologies for taking up the unnecessary space (Mods feel free to delete both posts if possible)

  146. You really need to get these in as well:

    box-sizing

    :active

    :hover

  147. Ben says:

    I would rather that Microsoft just keep working on this until the product is finished. I don’t care as much about the release date of ie7 as I do that it play nice.

  148. Ryan B says:

    I prefer the CSS interpretation of current Internet Explorer browsers.  The W3C CSS standards are extremely difficult to work with.  When working with CSS layouts, the IE CSS just seems to translate better between what you code and what appears on screen.  

    It is no suprise that when you search the web for a solution to a CSS compatibility problem, and you stumble onto a forum thread where a designer is having a similar problem, more often than not the designer has created something that works well in IE but has a problem with Firefox, for example.  The reason? The W3C CSS standards are un-intuitive.

    One thing that really disturbs me is the W3C CSS box model.  If you specify a width, padding and borders should be factored into that width (like IE 6 does without a DOCTYPE), not added to it.  A 300px wide box should be 300px wide, not 302px because you put a 1 px border around it.

  149. Sean Foushee says:

    Ryan B, then you’re lazy.  Sorry if thats harsh, but if you would rather the standards, that have been in place for YEARS, be changed to suit IE’s failed interpretation of those standards then you really aren’t a web designer or care about developing a product that the majority of web users can utilize effectively.  You know what I think when I see a site on the modern web that says "Made for IE"?  The developer is lazy.  

    Again, I’m sorry if this is too harsh for the comment section, but the stark reality is that IE has become the 800 lbs. Gorilla that is holding back progress on the web due to Microsoft’s belief that it must be a part of the OS for enterprise apps and only run CSS standards they like and only in a way they think it should be interpreted.  IE is the problem, not the standards, and based on my limited testing with IE7 I don’t see a pot of gold at the end of this Beta rainbow, but another several years of more IE specific CSS-hacks while web progress is sacrificed at the Microsoft alter of "standards compliance."

    Microsoft IE Team, I’m glad you’re taking steps to move the browser forward, and the neat eye candy like the new print feature is nice, but the haphazard implementation of CSS in IE7 will not win over many developers who must work with your app on a daily basis.  You must fix CSS to at least the level of today’s modern browsers before this browser is released to the public, or it will do more harm than good.  Is Microsoft ready to answer all of the support calls and emails from new IE7 clients wondering why a website(s) they visited for years in IE6 looked better than in your NEW browser?

  150. Lou Rutker says:

    Is there any simple way to determine if a page (not one that I controll and therefore not one that I can add a script to) is being rendered in quirks mode?

  151. Jonathan says:

    I’d also like to see display: table and friends implemented. (min|max)-(height|width) would be nice too.

  152. Michael Ward says:

    I’m glad that the IE7 team has committed to implementing min/max-height.

    But let me re-iterate this: under no circumstances should IE7 be released with the overflow bug fixed and no min/max-height implemented.

    Doing so would result in a horrible, horrible mess for many standards compliant sites that rely on either IE6’s non-standard rendering or support for min/max-height in better browsers.

  153. Benjamin says:

    I know this isn’t related to IE, and if someone could tell me where to submit it to that would be fab…my DNS server is bad, forcing me to connect to a different DNS server via IP Config in connection properties…since I’m constantly utilizing my school’s and my house’s wireless network, is it possible to get those included in the profile instead of not. Cuz its a pain changing the DNS address between automatic assigning and the nearest public DNS.

    Thanx.

  154. libertas83 says:

    I have read through alot of comments and it seems people are forgetting the release of IE7 is tied to Vista. That is why time is very limited and they can’t add full support to CSS 2.1. They have to make sure IE7 is stable on both XP and Vista along with makins sure it is secure and fix CSS bugs. That is quite a difficult task and I’m impressed with the work done so far. Thanks IE team for doing the best with the time you have. I am disappointed about the :after and :before not being implemented, but I look forward to it in the future.

    What I really look forward to is the final version of this: http://dean.edwards.name/IE7/overview/

    This javascript is the best hack to add CSS support to IE5 and IE6.

  155. trigger says:

    Folder view for FTP sites doesn’t appear to work anymore, still looking for the history button, and the new icons for favorites are annoying(I know I can turn them off).

    otherwise looking good.

  156. Opera user says:

    Markus [MSFT] said:

    > Just to reiterate:

    > We never said we would have full CSS2.1

    > compliance by the time IE7 ships. We are

    > committed to the standard and working hard to

    > make progress.

    I don’t doubt you’re working hard, but if you can’t comply CSS2.1 *at*least* at the same level that already existing browsers, why bother?

    Please, consider delaying IE7 release until you can say "Maybe it will be just for a few days, but today IE7 is the browser with the best standards support".

    I don’t know your reasons, but leaving :after and :before out of IE7 is a BAD idea. It will make IE7 an old browser from day one.

  157. Arnold Visser says:

    I see a lot of really nice fixes and features! specially in the CSS field.

    But I also see that you drop the _height "hack".

    I know the reason you drop this is because it is more a bug than a feature, and since you’re browser share is the biggest on the market, you want to decide what is the best intrepertation of the CSS "rules".

    But if you ignore the _ hack, and you don’t want to make a complete min-height feature, isn’t it possible for you to make "min-height" be treated as "height" as "height" is treated in current IE6?

    Maybe not the best solution, but defenatly the solution that won’t break a huge amounth of websites…

    I think a huge amouth of web developers will advise users to keep using IE6 or move to firefox/opera, instead of IE7

    Did you see the opera 9 technology preview? and the way they handle CSS?

    I think you can learn a lot from them!

  158. lessetr says:

    PLEASE

    CSS2

    PLEASE

    CSS2

    PLEASE

    CSS2

  159. Dave says:

    I’d like to report a bug that I’ve come across.

    In IE 6 ,the "Find" dialog is present on a page, clicking a link on that page causes the dialog to close before the page unloads.

    In IE 7 Beta, the "find" dialog does not close, when I try to close it after the next page has loaded, I get an error: "An error has occurred in this dialog" 385, Permission Denied.

  160. Dao says:

    Please think about introducing an -ie-, -msie- or -whatever- prefix for CSS features you’re working on but that may not be ready for the final IE7 release.

    After all, an alpha state feature is better than nothing. Even if not workable throughout, it could be useful in some cases.

  161. mrpoopooplops says:

    How about a limit on the strength of CSS filters in malicious sites?

    I can’t count the amount of times the following code has maxed my memory and CPU usage, locked up my browser and forced me to reboot.

    Shorly no-one really neads an element to have glow/wave strength of ‘999999999’ or more!

    <html><body>

    <style>

    *{

           filter: progid:dximagetransform.microsoft.wave(strength=999999999);

    }

    </style>

    <!–

    Lots of content here to kill IE

    –>

    </body></html>

  162. Tom says:

    Mystere,

    It’s not a viable layout method. We are forced to use inferior methods just because M$ doesn’t want to update it. There have been very few CSS 2.1 support updates so far. It looks like you completely redid the UI to make it look like you’re innovating to cover up the fact that you’re not.

  163. Dao says:

    See how the :hover menus flicker:

    http://www.playable.de/?stil=modern

    See how the menu (:hover and non-:hover) gets screwed up:

    http://design-noir.de/webdev/

    Also note that I had to add #menu a { float: left; } because with #menu a { display: block; } IE stretched #menu li to the entire width of #menu, which is wrong.

    Finally, my comments form is partially unstyled:

    http://design-noir.de/Bilder/

    This used to work until I changed the CSS slightly. I suppose it’s a bug in the CSS (selectors) parser.

  164. cjordan says:

    I agree with Tom, and also with everyone who’s said it too little too late. It seems to me that if they have *that* much trouble updating the thing it because they tied it too closely to the OS.

    Dumb, dumb, dumb… I don’t know whose bright idea that was but they should be hung in the town square by their toenails or beaten with a wet noodle.

    It seems *extreemly* silly to me that other folks are able to produce a stand-alone browser that can do everything IE can do, and in most cases it can do them better (I’ll admit though, that I like the "short-cut", that is innerHTML).

    Isn’t this *exactly* what they got sued for in the first place a while back? Having their web browser be required for the OS? Ugh.

    When will they learn? You’re right. Probably never. The open source folks really seem to have the right model. Microsoft ought to take a page from their book.

    I’m not saying that Microsoft needs to start making open source software. They’ve got every right to make money doing what their doing, but with all the resources MS has at its disposal, you would think that they’d want to stay on the BLEEDING edge.

    I know a programmer with Microsoft who works on the Outlook Express team (or at least he did several years ago). From what I’ve heard, often times the programmers *know* that the way they’re coding things isn’t the "right" way or the most efficient way, but management ties their hands.

    Because the darned thing is tied so closely to the OS, they have to be careful that changes they make don’t adversely affect the core OS development team. Or it could be that the way a certain piece of code was written by the OS development team has forced the IE team into a corner.

    *sigh* what was that old saying about ‘too many cooks’ again? I think that can go double for programming teams. Microsoft also needs to stop releasing code before it’s *really* ready. This again is (*I* think) symptomatic of their stupid, nay, insane propensity to intertwine everything to the OS. Simply put, there’se just too much bloat in their OS and in most of their products.

    Good thing they spend lots on their marketing department so they can keep the sheep.. ahem, I mean the consumer public buying their products.

    Okay, I’m done now.

  165. Michael Sheakoski says:

    Target.com sued by blind student

    http://arstechnica.com/news.ars/post/20060213-6166.html

    I read this article and thought, you know what… if IE supported all of the CSS2.1 features that would allow us to do everything we need tables for right now in CSS, we could easily make the internet a more accessible place for those with impairments. I truely believe that IE’s current lack of support for standards is the main reason that the blind have a harder time navigating the web because developers right now need to jump through hoops and spend hours to achieve certain layouts without display:table that take only seconds if that support was there.

    We need these basic CSS 2.1 features added so we can better support our blind and visually impaired users:

    1.  display: table

    2.  display: table-cell

    3.  display: table-row

    4.  display: table-*

    5.  min-width

    6.  max-width

    7.  min-height

    8.  max-height

    9.  :before

    10. :after

    Try browsing the web using a screen reader with your eyes closed and you’ll quickly see how much of a mess IE’s lack of CSS support has created. For the sake of humanity, please support at least these BASIC CSS 2.1 FEATURES!!!

  166. Pascal says:

    After reading these minimal improvements and seeing the graphical nonsense further in the blog every developer should be angry.

    IE7 will be released in 2006 so we (the developers and internet users) are screwed til at least 2012! How can Microsoft even talk about inovation with this fact?

  167. Nice, hopefully CSS2 will be supported with some seriousness.  The standards are there to be ignored by the likes of microsoft i often think.  So at least some bugs are (theoretically) ironed out.  

    (YAY – Fixed Postioning)

  168. Dao says:

    http://www.hixie.ch/tests/adhoc/css/box/block/001.html , 002.html , 011.html , 013.html , 014.html, 015.html , 016.html

    I’ve only tested the box/block/* pages, those above fail. Some others are XML, so they are not working at all.

    Maybe you should check out some other tests yourself:

    http://hixie.ch/tests/MANIFEST

  169. Edward Kranz says:

    A designer’s review of Internet Explorer 7 Beta 2

  170. At the end of January, Microsoft released a public beta of the next version of Internet Explorer [download | readme | tour]. IE 7b2 offers more stringent security measures and other improvements including tabbed-browsing, updated CSS, RSS support and

  171. TechBlog says:

    Scott Graff writes at the IE7 Blog that the final release of Internet Explorer 7 is almost upon us — as in, this month. He also reveals a little more about how its distribution through Windows Update will be handled:…

  172. Nie, nie będę obiektywny. Nie napiszę, że to dobrze, że coś się zaczyna zmieniać (Polski, skrócony odpowiednik). Internet Explorer był, jest i będzie najgorszą przeglądarką na rynku- mam nadzieję, że określenie "najpopularniejsza" bardzo

  173. Bink.nu says:

    In Dean’s recent Internet Explorer 8 and Acid2: A Milestone post, he highlighted our responsibility to

  174. At the end of January, Microsoft released a public beta of the next version of Internet Explorer [download | readme | tour]. IE 7b2 offers more stringent security measures and other improvements including tabbed-browsing, updated CSS, RSS support and