From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.9) Gecko/20020412 Description of problem: When I attempt to delete a RHN system profile using Konqueror as my web browser, it fails to delete the profile at the "System Deletion Confirmation Page." When I click the red button "Confirm Deletion of This System Profile", the same page re-loads rather than deleting the profile. Mozilla subsequently successfully deleted this system profile. Version-Release number of selected component (if applicable): Skipjack beta 2 KDE 3.0.0-9 How reproducible: Always Steps to Reproduce: 1. Login to RHN with Konqueror. 2. Attempt to Delete a System Profile. 3. Click "Confirm Deletion of this System Profile" Actual Results: Page reloads and nothing has changed. Expected Results: System profile should be deleted.
Re-tested: Still exists in Red Hat Linux 7.3.
*** Bug 64928 has been marked as a duplicate of this bug. ***
Something to note: On the RHN side of things, we're seeing formvars getting submitted to completely different pages than they're intended for. I.e., the "delete a server" page might get the formvars from archiving scheduled actions...
Found another good testcase showing Konqueror messing up real bad... http://www.icrontic.com/index.php?page=public/hardware&hardwareId=364 Try clicking through the pages of this review, then trying the links on the left side of the page. Very often with Konqueror you end up at a page showing MySQL login errors. You never see those errors in Mozilla. (It seems that their PHP code depends on client submitted data... tsk tsk.)
On RHN, we are still seeing dozens of tracebacks a day from Konqueror users due to formvar manglement -- and whomever has classified this bug as "low" hasn't fully considered the potentially ugly ramifications of passing an incorrect server ID to a "delete server" request page; therefore, I've moved the priority back to "high". If someone decides to set it back to "low" again, please explain the rationale to me. Is there any relief in sight to this issue? It's now two months old.
This has been fixed (at least it works perfectly here with 3.0.2-0.cvs20020624.1). Please let me know if you can still see this problem with current versions.
I haven't seen any problems like this in Limbo either. Safe to close?
Can errata be issued for KDE fixing this issue in Valhalla? Many people will still be using 7.3 for a long time, and I would hate for them to use a dangerously broken browser with this bad behavior.