Skip to: site navigation/presentation
Skip to: Thoughts From Eric

Freezer Case

Since a few people asked for it, I’ve created a test file that reproduces the Internet Explorer freeze reported yesterday.  You can find it with the title “Internet Explorer Freezes — BEWARE!“.  If you follow that link with a non-IE browser, you should see a static copy of the “Ten Things To Do…” post, with two differences:

  1. I stripped off the theme stylesheet so the masthead graphic isn’t shown.
  2. I inserted a warning/explanatory comment near the top of the document.

Under the hood, the main difference is that the style sheets are all embedded in the document, so if you’re so inclined you can download that single page to your hard drive and fiddle with it to your heart’s content.  If anyone can narrow down this problem to a very minimal test case, I’d love to see it.  Refer back to “When Browsers Attack!” for notes on what I discovered.  There may well be more to the story, but if so, I didn’t find it.

I’ll reiterate the warning, in case it somehow wasn’t clear enough: iif you load this document in IE/Win, the odds are very, very, very high that it will freeze Internet Explorer, necessitating a force-quit of the application.  It may also crash Windows.  If you don’t want those sorts of things to happen to you, then don’t load the document.  Clear?  Good.

23 Responses»

    • #1
    • Pingback
    • Tue 21 Sep 2004
    • 1205
    Received from Very True Things - Conspiracy theory

    […] ight in recent weeks when the CSS in question has been in use on his site for some months. Comments on a later post point to the culprit possibly being a recent pat […]

    • #2
    • Comment
    • Fri 17 Sep 2004
    • 1231
    Matt Wilcox wrote in to say...

    Oddly enough that page works fine in IE 5 and 5.5, it only kills IE6 on my system (WinXP, SP2)

    • #3
    • Comment
    • Fri 17 Sep 2004
    • 1234
    Dave Child wrote in to say...

    I had a similar problem, by the sounds of it, on a recent site. I tracked it down, in the end, to a couple of form elements and their positioning. I’d added relative positioning to one element to allow absolute positioning of another within it. However, in IE, that resulted in two form elements overlapping (as best I can tell). That resulted in a crash every time. When I changed the form element positioing, all was well once again.

    • #4
    • Comment
    • Fri 17 Sep 2004
    • 1236
    Steve wrote in to say...

    I opened this URL in IE 5.5, Win98, repeatedly with no problems. Is this an IE6 issue possibly?

    • #5
    • Comment
    • Fri 17 Sep 2004
    • 1253
    Ean wrote in to say...

    It seems to work without the DOCTYPE…

    • #6
    • Comment
    • Fri 17 Sep 2004
    • 1508
    seth wrote in to say...

    I’ve made a reduced version of the problematic page (warning: this page will crash IE).

    I was trying to cut out anything that wasn’t causing the problem, so as to focus efforts in the correct spot. Unfortunately, this only brought about more confusion.

    Using the reduced page… Making ANY ONE of these changes produces a page which works in IE6:

    1) Remove ANY single declaration from the CSS

    2) Remove top padding from div#main. This includes making the declaration padding: 0em 4em 3em 4em

    3) Remove a single x from either of the skip nav links. (You’ll see what I mean when you see the source)

    Along the way, I had this configuration (works in IE6). Notice that the last <li> is repeated (in IE6). Whoa.

    I’m gonna keep stabbing at it, but thought I would see if anyone else found the new page useful in the hunt.

    • #7
    • Comment
    • Fri 17 Sep 2004
    • 1647
    Joshua wrote in to say...

    In my daily reading of meyerweb, I had no freezing issues with the page in question, (IE6/NT). Yesterdays reading at home also warranted no forced shut down either, but that’s Mozilla for you.

    • #8
    • Comment
    • Fri 17 Sep 2004
    • 1747
    Eddie Sowden wrote in to say...

    I also made a reduced version of the bad page here. Made this earlier but didn’t reply. Like the one Seth made nothing can be reduced (or added) from the file or it seems to appear in IE. One strange thing is some of the classes Seth removed cant be removed from my example.
    This one has stumped me and I think it is just a selection of, too many errors in such a strange way, that cause IE to just fall over and cry.

    • #9
    • Comment
    • Fri 17 Sep 2004
    • 1919
    Greg Campbell wrote in to say...

    I also have a minimal test case. I added some comments based on my findings. Very strange indeed.

    • #10
    • Pingback
    • Fri 17 Sep 2004
    • 2046
    Received from bla2kay » Internet Explorer mit Style Sheets killen

    […] zornigen Post �ber den Internet Explorer, hat Eric Meyer nun in einem weiteren Post einen Testcase ver�ffentlicht, wie man den Internet Explorer 6 mit reinen, standardkonfo […]

    • #11
    • Comment
    • Fri 17 Sep 2004
    • 2159
    Lachlan Hunt wrote in to say...

    This is the strangest bug I’ve ever encountered. Even with stylesheets disabled in IE, it still crashes. I’ll do some more experementing later when I have time, but for now, this is just another reason for users to drop IE.

    • #12
    • Comment
    • Sat 18 Sep 2004
    • 0542
    Robbert Broersma wrote in to say...

    Both Sowdens’ and Campbells’ testcase do not crash IE6@XP on my PC.

    • #13
    • Comment
    • Sat 18 Sep 2004
    • 0725
    Mike Rainey wrote in to say...

    I have tested the above test cases in all IE browsers since IE4. IE4, IE5, and IE5.5 do not crash. IE6 crashed, but only one of them.

    What I mean by one of them is that I have two computers with IE6 on them. One that has all the available security patches installed and one that does not have any. The one that has all the security patches chashes horribly, the unpatched one opens all the test cases fine without crashing.

    I then began applying patches one by one, starting with the earliest available patch. After applying each patch, I went to the test cases again and all worked fine. When I got to the latest patch, KB867801, the browser frose on the test cases.

    As I cannot test this theory anymore, I am not certain this is the case. If anyone else would like to test this, please do so.

    • #14
    • Comment
    • Sat 18 Sep 2004
    • 0845
    seth wrote in to say...

    I just checked, and I DO NOT have that KB867801 installed on my IE6/XP machine.

    Neither Snowden’s or Cambell’s test cases freeze my browser. My original test does.

    • #15
    • Comment
    • Sat 18 Sep 2004
    • 1122
    Keith Burgin wrote in to say...

    I’m sorry. Did you say it would fre

    • #16
    • Comment
    • Sun 19 Sep 2004
    • 2159
    Bern wrote in to say...

    Not sure if this helps – change the line before the FREEZE line to padding: 0.25em 1% 0 0; and it no longer freezes on my system (Win2000)

    • #17
    • Comment
    • Mon 20 Sep 2004
    • 1656
    Eric Hodel wrote in to say...

    I have a test case that is window-width sensitive. I can load the testcase just fine, but when I expand the window enough it crashes.

    I don’t need to use a ul, a div with the proper margins in place of the ul will trigger the crash (along with a width and float for the wrapper).

    In the situations where IE does not crash, the content of the ‘list items’ is displayed twice, the second time below the wrapper div.

    In a related win32 bug, while IE is hung, you can open up lots of Windows Task Managers.

    • #18
    • Comment
    • Mon 20 Sep 2004
    • 2025
    Brett Merkey wrote in to say...

    IE6 and strict mode don’t go together. It’s the DOCTYPE here. The list of bizarre behaviors that IE6 displays in strict mode is criminally long.

    Every time a co-worker has ignored my strong recommendation against strict mode in our IE shop, they have seriously regretted it.

    • #19
    • Comment
    • Tue 21 Sep 2004
    • 1723
    M. Ford wrote in to say...

    Both Sowdens’ and Campbells’ testcase do not crash IE6@XP on my PC.
    Seth’s loaded too, and then reloaded and stalled IE 6.
    I have XP SP2 and all updates installed on Win XP Pro.
    Now I’ll try the original test case…. can’t resist

    • #20
    • Comment
    • Tue 21 Sep 2004
    • 1727
    M. Ford wrote in to say...

    The original test case stalls IE and it becomes unresponsive. but my system doesn’t go down. I can shut down IE 6. so it’s not really a ‘crash’ of XP… at least for me.

    • #21
    • Comment
    • Wed 22 Sep 2004
    • 1838
    John wrote in to say...

    I sent you an e-mail about 4pm today (9/22) that covers the fix for this issue.

    I will leave it up to you, to decide, how to cover the changes that were/are required – for your “regular visitors” who are interested.

    Thanks for the challenge.

    • #22
    • Comment
    • Thu 23 Sep 2004
    • 1705
    John wrote in to say...

    Ok – it was the use of “Float: left;” for both the UL & LI. Apparently, the use of it for the entire UL and then the LI would cause an infinite loop to start as IE tried to align the elements. Removing it from the UL fixes it – but not vice-versa…

    The reason it just showed up is it had something to do with the length of the “values” (link text) shown- hence when the values reached a certain length – it looped.

    I don’t have a clue why – this is just my “quasi-educated guess” – so MS or W3C would likely know more… If I can’t control it – I find the workaround and live with it – everything else is a waste of brain chemicals…

    • #23
    • Comment
    • Sat 15 Jul 2006
    • 0803
    nigel wrote in to say...

    just to say – thanks for posting this bug
    i had the same problem with a bunch of checkboxes within divs
    and only found the problem by randomly cutting bits from the css ;)
    Does anyone at microsoft know about this I wonder?

Leave a Comment

Line and paragraph breaks automatic, e-mail address required but never displayed, HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Remember to encode character entities if you're posting markup examples! Management reserves the right to edit or remove any comment—especially those that are abusive, irrelevant to the topic at hand, or made by anonymous posters—although honestly, most edits are a matter of fixing mangled markup. Thus the note about encoding your entities. If you're satisfied with what you've written, then go ahead...

September 2004
August October