| Summary: | java applets in htlm pages dont work in konq | ||
|---|---|---|---|
| Product: | [Applications] konqueror | Reporter: | Markus <markus.gapp> |
| Component: | kjava | Assignee: | Konqueror Bugs <konqueror-bugs-null> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | ||
| Priority: | NOR | ||
| Version First Reported In: | unspecified | ||
| Target Milestone: | --- | ||
| Platform: | Compiled Sources | ||
| OS: | Linux | ||
| Latest Commit: | Version Fixed/Implemented In: | ||
| Sentry Crash Report: | |||
|
Description
Markus
2003-11-11 16:46:40 UTC
For me some applets show as gray rectangles, and often reloading the page helps. Not sure what causes it, but it started during or after the N7Y meeting. I get also only gray rectangeles and my cpu load goes to 100%. This makes some websites unusable e.g. http://www.oeamtc.at/index.php?type=menu&id=103 , https://www.netbanking.at/netbanking/index.html , http://www.orf.at/031112-67476/index.html . I think this is really a grave bug as it makes these websites unusable (I have to kill java) and I have no problem when using konqueror from kde 3.1.4 (regression) These applets seems to work for me. For http://www.oeamtc.at/index.php?type=menu&id=103 you means applets on the pages linked on the main frame, no? Strange combo box on these, the list popups appears outside konqueror. Somehow khtml becomes crashy on these pages, not sure which one. I got a 'ASSERT: "block_stacking_updates == 0" in layers.cpp (528)' assertion. Hmm, running two different instances of konqueror with a jvm running, makes one of these not working at all. What is using 100% CPU, java or konqueror? I would expect konqueror, the embedding of the applet, but killing java would help then of course. Those pages are a problem here as well. > Hmm, running two different instances of konqueror with a jvm running, makes
> one of these not working at all.
Ok, this is because I run jvm with
-Xdebug -Xnoagent -Djava.compiler=NONE -Xrunjdwp:transport=dt_socket,server=y,address=8888,suspend=n
second jvm exits with address already in use error. So that's my fault.
Still I would be interested in who is using 100% CPU. If it's konqueror, a backtrace of a manual crash would be nice. (get it with eg. 'kill -SEGV <pid>')
Also the console output might be interesting.
This one is fixed by making java window undecorated. Reopen if it still fails in > KDE-3.2 |