IE7: Old Bugs For New

Some good features but flawed rendering.

30th August 2006 · Last updated: 10th December 2011
 

My IE7 Bugs List · Some Bugs Other People Have Found · Conclusion · Comments


IE7IE6 was reknowned for its bugs. It looks like IE7 is going to be almost as buggy. New features, sure, but new bugs too! It's almost like they've traded old bugs for new. All the CSS fixes and security improvements Microsoft have made in IE7, which I definitely applaud, aren't worth much if it results in a browser that's less stable than IE6. I'm talking display problems, serious stuff like missing content, scrolling problems..... not to mention the hideous new interface. I won't dwell on that here. Instead, I'd like to document the bugs I've found myself, using IE7 Beta 2, 3 and RC1.

On the official IEBlog, Microsoft General Manager Dean Hachamovitch says "With the exception of a very short list of issues we're aware of and working on, we think the product is done." Well they may think so, but judging by the number of complaints in the IEBlog comments on many posts, users are still finding a range of bugs. I would say IE7 is far from done. The problem is, they hope to ship this browser in the 4th quarter this year, 2006, via a Windows Update. That's fine, and you can choose whether to install it or not, but there are just way too many outstanding bugs to consider this "done". I can only hope they will fix the bugs in future updates.

I was about to report the bugs I've found using the official feedback database, until I came across this post: Internet Explorer feedback system a disaster. I quote: "Microsoft has announced that within the next couple of weeks they plan to close every report that has been in the database for over a month, asking users to manually reopen all reports that still apply... It seems the system as it is currently designed is unable to scale to reasonable levels and they are letting it collapse in on itself." There are other perceived problems as well - read the full post.

So I decided to stay clear of the feedback system altogther, and mention my bugs on the IEBlog pages instead. What's more, I am still concerned that you have to have a Microsoft Passport ID to access the bug database. This is despite the following written by Al Billings of Microsoft on the IEBlog on 24th March 2006 (emphasis mine): "The current version of the site will require visitors to have a Passport account in order to view or enter bugs. In an update in a couple of months, visitors will be able to view existing bugs without logging into the system." As far as I can tell, this update still hasn't occured.

My IE7 Bugs List

Now onto my list of bugs, spotted after barely using the browser. I fear there may be many more that would come to light were I to use this browser more!

NOTE: I'm using the standalone version of IE7. Why? Like many others, I don't feel I can trust IE7 to overwrite IE6 just yet. Plus I keep hearing of horror stories installing it the official way (two lengthy reboots anyone?). I have read that the standalone version can cause problems, which is true of the broken menus, but I'm testing the rendering, which should be identical. However, if any of these bugs don't appear in the normal version, I'll retract them.

  • Content disappears! Take a look at this page and scroll down to the bottom. You should see a box at the bottom of the centre column that starts "If you're new to UDM 4". (Screenshot.) Now scroll slowly up or down. Switch to another program. Switching back to IE7 can make the whole box disappear! (Screenshot.) Scrolling can sometimes toggle the missing box on and off too. Even worse, I once saw the entire centre column of text go missing! (Screenshot.) There's a serious rendering issue here.
  • A background using a gradient image can disappear when hovering over links in a list. You can repaint the background white! I made a demo of this bug. It turns out to be down to a line of CSS that references a missing element. But that shouldn't affect anything!
  • There's a layout bug on this page. The last block on the page is too far left. (Screenshot.)
  • Scrolling with the mousewheel is choppy - it can jump back up or down. IE6 never had this problem.
  • Scrolling can carry on moving by itself after stopping the mousewheel. Very disconcerting.
  • Scrolling is much slower when zoomed in (almost treacle-like at higher zoom levels). Try this site. Note: the site is unreadable for me at 400% due to the top and bottom bars covering the text! This is down to the next bug:
  • Zooming wrecks some images (PNGs?) by turning part of them black. See this site for an example. (Screenshot.) Note how the text fade at the top and bottom is no longer seen after zooming - the gradient images used have become solid bars!
  • The Stretched Button Bug is still present. To see it, look at the bottom button on Tools -> Internet Options -> Security tab.
  • The zoom icon can overlap the text on a trusted site. (Screenshot.)
  • The zoom tooltip can obscure the control. (Screenshot.)
  • Clicking the zoom icon goes from 100% to 125% to 150% then back to 100%. Huh?
  • Sometimes clicking the zoom button or arrow for the menu does nothing. You can still zoom in with CTRL and + but the zoom button doesn't update. (It still says 100%.)
  • Background images are not zoomed! This can mess up some pages when zooming in, and spoils the effect.
  • Pages seem to always reload on going back. I wonder if this is deliberate, as a way to get more hits? (Cynical, me?)
  • Still no favicons in the address bar! At least not in the standalone version.
  • The html {border:0} code no longer works. In IE6 it removed the thin line around the page.
  • The graphic of the man below the side menu is completely missing from this site.
  • There's no default underlining for abbreviations (so you can't tell there's a tooltip available if you hover over the text).

Some Bugs Other People Have Found

A full list is probably quite long, but here are a few bug demos I've noticed on the IEBlog:

Conclusion

I personally feel that IE7 is worse than IE6 for surfing. There's no way I could use this as my main browser, with its weird scrolling habits, missing content, wrongly rendered pages and so on. Opera and Firefox feel much smoother, rarely failing to render a site properly. Plus they offer a huge range of useful features not found in IE7 (though IE7 does allow extension-like plug-ins). Also the interfaces can be tweaked as much as you like. With IE7, you're stuck with the unpleasant position of the navigation buttons and the horrible fake-Vista style. I'm all for the new Vista look, but with IE7, Microsoft should have maintained the style of IE6, which fits nicely with Windows XP. Oh wait, I said I wouldn't mention the interface.

Even if they cure all the bugs in IE7, it still feels like a dinosaur with a new coat. The rendering engine is ancient (from the days of competition with Netscape - remember them?) and really should be laid to rest. While the Microsoft IE team have done a wonderful job in a short space of time in repairing IE, the core of the browser seems to always cause layout problems. There's no way that content should ever go missing! Or weird things happen with backgrounds. The fact is that modern browsers are based on newer rendering engines, deliberately created for today's web. (Opera 7 for instance was a complete rewrite of the renderer, while Netscape abandoned their broken browser engine for the newer Mozilla code, also used in Firefox.) I hope that IE8 and 9 don't use the current renderer, Trident, and build instead on the newer engine that I think is used in Vista. Until then, we're still stuck with a buggy version of IE. It just won't stay bug free.

Comments (10)

Comments are locked on this topic. Thanks to everyone who posted a comment.

  1. Dave Lane:

    Hi Chris,

    I'm confuse by your comment which quotes mine on the IEBlog site - specifically on this entry:
    http://blogs.msdn.com/ie/archive/2006/08/24/715752.aspx After I suggested that the IE7 development team could base their browser on a working rendering engine like Gecko or KHTML as they are both "free to use" you retorted that IE7 is free to use. You missed my point. By "free to use" I meant "free to incorporate into their code" due to its licensing. I think you'll find that the IE7 Trident rendering engine cannot be freely integrated into another product without MS's say-so.

    The rest of your comment is somewhat incoherent, so maybe I've misinterpreted it, but you seem to be defending IE7 RC1, and yet you pan it here on your blog... what's up with that? Anyway, what I do know is that MS's IE5-6 and now 7 as well, have increased the cost of developing a standards compliant website hugely.

    Posted 7 September 2006 @ 10:10 pm.


  2. Dave Lane:

    If MS was to bear the real cost of their incompetence in developing a properly standards compliant browser (on par with those I mentioned above), then they'd be out of business.

    The web developing world is justifiably livid with Microsoft's abuse of its monopoly position and many of us wonder how a company that has stolen so much and has built such wealth for itself could simultaneously be so demonstrably stupid. I assert that they're not just stupid, I believe their utterly calculated in breaking things to support their hegemony, and they simply use their PR department to patch things up with the grass munching herds of normal Windows users.

    Those of us who know how much better it can be, however, are impatient for the world to catch up.

    Cheers,

    Dave

    Posted 7 September 2006 @ 10:12 pm.


  3. Chris Hester:

    Hi Dave, thanks for dropping by! Sorry I wasn't more friendly to you on the IEBlog.

    "you seem to be defending IE7 RC1, and yet you pan it here on your blog... what's up with that?"

    It depends how I feel! Really, though, if you check other IEBlog posts I've commented on, you'll see me attacking IE7 just as much as here, if not more so. I guess I just misinterpreted your comment that IE7 was not free. Thanks for explaining it clearly for me.

    Personally I wish they would dump IE's engine for Gecko or Presto. But I guess too many intranet and business programs rely on IE-only code for that to happen.

    Posted 7 September 2006 @ 11:06 pm.


  4. Ross Bruniges:

    You get the crazy scrolling when using your mouse wheel when using the full version too - though of cause a beta - pissed me off so I uninstalled it!!!!

    Posted 13 September 2006 @ 9:41 am.


  5. Daniel Lewis:

    For some reason when I read this blog post I had a bit of deja-vu. Not sure why.

    Anyways, you have made some very valid points. I use Flock (and Safari and Mozilla Suite), and Mac OS X. So I will keep in mind these bugs when I am making sites.

    I have recently made a blog post about this all (and promoting the use of other browsers), hope you don't mind me linking to your post.
    Here is my post: http://vanirsystems.co.uk/danielsblog/?p=37

    Thanks,

    Daniel.

    Posted 15 September 2006 @ 2:46 pm.


  6. Klaus Hartl:

    Hi there, just wanted to add another important one to the list: in IE7 RC1 the universal selector selects HTML comments as well and you can even apply styles to them. It will become absolutely useless...

    Test page

    Posted 28 September 2006 @ 11:09 pm.


  7. Steve Ganz:

    Here's another for your list:
    min-width is applied as width on input buttons.

    Posted 22nd October 2006 @ 12:03 am.


  8. Chris Hester:

    Gérard Talbot has a list of IE7 bugs.

    Posted 23rd October 2006 @ 08:19 pm.


  9. Gérard Talbot:

    Hello Chris,

    * "It's almost like they've traded old bugs for new."
    IE dev. team fixed more bugs than they created new ones. Any regression bug is nevertheless difficult to accept. IE 7 now fails at least 3 specific CSS1 testsuite tests.

    * "I'm talking ... scrolling problems"
    Between RC1 and final release, the IE dev. team fixed at least one scrolling problem. Try http://www.gtalbot.org/DHTMLSection/ScrollingTBody.html with RC1 and with final rel.

    * "I decided to stay clear of the feedback system altogther and mention my bugs on the IEBlog pages instead" Excellent decision! Demonstrate IE7 bugs in bright daylight... not into a place where bugs are closed for no justifiable/good reason.

    * "In an update in a couple of months, visitors will be able to view existing bugs without logging into the system." MS creates expectations that later get frustrated, get ignored or get defeated: so, eventually, users understandably get upset and then stop cooperating, etc.

    * "Take a look at this page": For unknown reasons, I can not load http://www.udm4.com A reduced testcase would have worked around such possibility.

    * "There's a layout bug on this page." Quirks mode website has been entirely redesigned. Again, a reduced testcase is always better.

    * "zoom icon goes from 100% to 125% to 150% then back to 100%. Huh?" That's by design, I suspect.

    * "The html {border:0} code no longer works." That's not a bug. By design, IE6 had a default thin border around the html element. They decided (my assumption here) to harmonize some of their default property values with what other browsers were doing to reduce cross-browser efforts imposed on web designers.

    I agree with pretty much everything you say. IE7 has an awkward UI design (toolbars, navigation buttons). The IE dev. team will have lots of work to do on float model, inline box model, "hasLayout", DOM 2.
    Expect the next release of IE 7 to be in 18 months: you can mark my words.

    Gérard

    Posted 24th October 2006 @ 09:00 am.


  10. Jacquie:

    Can you explain this one? In IE6, little cameras and a "view ad" appear on the page. In IE7, they're gone. Any ideas how I can get around this one? Thanks!

    Posted 4th November 2006 @ 12:21 am.