User-Agent: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.6) Gecko/20040206 Firefox/0.8 Build Identifier: Mozilla/5.0 (Windows; U; Win98; en-US; rv:1.6) Gecko/20040206 Firefox/0.8 open http://mrskin.nudesonline.com in a new tab, view then close that new tab. when closing the tab an undesired popup window appears. the site is somehow able to evade firefox popup block! Reproducible: Always Steps to Reproduce: 1. open tab with options set to block all popups 2. visit http://mrskin.nudesonline.com 3. close tab of http://mrskin.nudesonline.com Actual Results: a popup appears! Expected Results: no popup should appear.
Confirmed with Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.7b) Gecko/20040327 Firefox/0.8.0+ Also happens in a month-old build of Seamonkey. -> XP Apps
The popup is allowed when closing the tab by (almost) any means: middle-click, context menu, tab close widget. Though Ctrl-F4 on Windows is an exception. Also note the popup is suppressed if the testcase is adjusted so the function is called on the frame itself. That is, onunload=window.open instead of onunload=parent.exit I'm adjusting the summary to reflect this. That's the longest summary I've ever written in a bug report. The popup should be caught by the general unload handler suppression. It should be independent of the method used to close the window(tab). In Jesse's test case the unload handler of the frame window is calling a function from the frameset window. At the time this happens, the frameset window doesn't yet realize that it's about to be unloaded. That's rather unfortunate. Somewhat happily, this seems to happen only under the restricted circumstances of this bug.
*** Bug 197895 has been marked as a duplicate of this bug. ***
I'm seeing this on Firefox 1.0PR on Win XP. Mac doesn't seem to be affected. Requesting blocking-aviary 1.0
currently testing at www.aol.com using the [X] close tab button blocks the popup onunload. using context Close Tab does not block the popup.
I can no longer reproduce this with builds from 0917
Yes this one has been fixed, no doubt by the general rewrite in bug 252326. (The subsequent adjustment in bug 259117 was probably also applicable.)