User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:184.108.40.206) Gecko/20071127 Firefox/220.127.116.11 Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:18.104.22.168) Gecko/20071127 Firefox/22.214.171.124 This issue happens all over this site and many others that I know of. When a page is loaded that contains two or more iframe elements such as rich media ads they distort. An ad size of 728x90 displays the left corner in a 3oox250 ad slot is one example. There are many variation to how this can occur. It has been known to happen with quick polls as well. Anything using iframes seems to pull the error. This error is not caused by incorrect ad tagging. The ads have been placed in their spot accordingly. If they were tagged incorrectly then they would display in full in the incorrect position. Reproducible: Sometimes Steps to Reproduce: 1.Go to page and refresh until problem displays 2.It should happen when two sprint ads are on the page 3.There must be two iframe ads on the page for this to work Actual Results: The ads on the page swapped and distorted in two slots. The one on the left displayed the box and and the box ad displayed the ad on the left. Expected Results: The ads should display correctly in the appropriate area without distorting the page. This seems to happen with every other release of FF. It appears to be fixed one week and back the next. I have spoken with the ad serving agency and they spoke to a long standing issue with Firefox.
This is still happening to our site in 126.96.36.199
Michael, is this still happening with Firefox 3?
I just tried testing this with FF 3.0.3 and was unable to reproduce it. So it seems to be fixed with the current version.
Michael, thanks for retesting. Matthew, is it no longer a problem for you either?
Well, I spoke too soon. I am seeing this problem again under FF 3.0.4
OK. Is there any DOM mutation going on here? Or just a straight parse of the HTML?
Not sure what that means, can you clarify?
Is the page moving the <iframe> elements around in the tree after the parse? Also, is this a situation where there are two ads on the page and they get swapped, or a situation where there is just one ad frame and it just doesn't load the right ad?
How can I check if its moving elements around the tree after parsing? Is there a tool in FF that would show this? In all cases thare are atleast two ads on the page, and they get swapped. If you would like to contact me offlist I can provide screenshots.
This is still happening for me as well. Although now with an array of sites. I can also provide screen shots. As Michael stated above, in all cases there are at least two iframe ads on a page. I will look into parsing shortly, but the problem seems to be intermittent. It also exists with widgets and other items created using iframes. I have even seen it on espn, but only in FF.
> Is there a tool in FF that would show this? Not easily. Basically, the question is whether there is page script which sets innerHTML including the iframes in question, or uses removeChild/appendChild/insertBefore to move the iframes or their ancestors in the DOM. Or creates the iframes using createElement, for that matter. See also bug 462076.
We're have the trouble with dynamically loaded iframes constantly and repeatedly. The trouble was proved to be not dependent on DOM nodes relocation. Which is more, we can help you to reproduce the bug.
This has been happening to us with 3rd party ads as well. This also causes the Facebook "Like" button to show up in ad slots. It also loads a hidden iframe used by the "Add This" widget (a.k.a. "Bookmark" or "Share", www.addthis.com) into an ad slot, causing a blank ad slot. It only happens when the visitor refreshes the page, not on initial load. It does not happen all the time, but it's not a rare event either.
Component: General → Layout: HTML Frames
Product: Firefox → Core
QA Contact: general → layout.html-frames
Component: Layout: HTML Frames → Layout: Images
Product: Core Graveyard → Core
You need to log in before you can comment on or make changes to this bug.