Version: 3.5.9 and 4.0.3 from kubuntu ppa (using KDE 4.0.3) Installed from: Unlisted Binary Package OS: Linux As the title says, a frame is rendered smaller than it should. This happens to some sites and a recent example is this: https://admin.acrobat.com/meetqtjambi/event/registration.html Here a frame containing the registration form is only 3-4 lines high while it should be big enough to display the whole form. There are no scrollbars but you can scroll through the hidden elements by selecting with the mouse some of the text that is there and without leaving the left mouse button, scrolling bellow the displayed text. The page is ok under firefox. I will post other sites as test cases, when I stumble upon them.
Using KDE 4.1.00 (KDE 4.1.0) (KDEmod) in ArchLinux i686: I can reproduce this bug. Using WebKit (svn rev 838395) inside Konqueror I can't reproduce the bug. So I suppose that is a KHTML bug.
Created attachment 26519 [details] Frame in Konqueror
Created attachment 26520 [details] Frame in Firefox3
I can also reproduce this using: Version 4.1.62 (KDE 4.1.62 (KDE 4.2 >= 20080814)) kdelibs4.2 svn rev.849324 kdebase4.2 svn rev.849337 ArchLinux 2.6.25-zen2-20080814 i686
Created attachment 27990 [details] 1st part of a test case (iframe contents)
Created attachment 27991 [details] Test case The page contains an <iframe> that manipulates its height using Javascript. It seems that the height calculation goes wrong for some reason. The lack of scrollbars is intended by the page.
Apparently, my test case does not really work for some reason. The bug is 100% reproducible for me in 3.5.10, 4.1.2 (Kubuntu) and trunk rev. 871890 as long as the files are on my hard drive. The uploaded test case is still buggy in 3.5.10 for me, but to get the bug in 4.x, I need to reload the page a few times. Strange. Anyway, if you download both files and adjust the src="..." in the <iframe> tag, the bug should be reproducible.
The below helps on the testcase but not the website.. ... but I can reproduce the going-back crash with this testcase, yey, would be really good to kill it. --- ecma/kjs_html.cpp (revision 873020) +++ ecma/kjs_html.cpp (working copy) @@ -405,8 +405,10 @@ case LastModified: return jsString(doc.lastModified()); case Height: // NS-only, not available in IE + doc.updateLayout(); return jsNumber(view ? view->contentsHeight() : 0); case Width: // NS-only, not available in IE + doc.updateLayout(); return jsNumber(view ? view->contentsWidth() : 0); case Dir: return body ? jsString(body->getAttribute(ATTR_DIR)) : jsUndefined();
Reproducible on 4.2.90.
Thank you for the bug report. As this report hasn't seen any changes in 10 years or more, we ask if you can please confirm that the issue still persists. If this bug is no longer persisting or relevant please change the status to resolved.