Bug 316958 - drkonqi hangs indefinately while sending crash report
Summary: drkonqi hangs indefinately while sending crash report
Status: RESOLVED FIXED
Alias: None
Product: drkonqi
Classification: Applications
Component: general (show other bugs)
Version: 2.1.5
Platform: openSUSE Linux
: NOR normal
Target Milestone: ---
Assignee: Unassigned bugs mailing-list
URL:
Keywords:
: 317137 (view as bug list)
Depends on:
Blocks:
 
Reported: 2013-03-18 10:24 UTC by kdebugs.anon134
Modified: 2013-03-29 10:09 UTC (History)
4 users (show)

See Also:
Latest Commit:
Version Fixed In: 4.10.2
Sentry Crash Report:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description kdebugs.anon134 2013-03-18 10:24:34 UTC
This problem (not a crash) is reproducible every time.

The network connection was and is working properly at the time of the malfunction.

KDE version 4.10.1 "release 545"
Distro is openSUSE 12.3 (x86_64)
OS is Linux 3.7.10-1.1-desktop x86_64

- What I was doing when the application malfunctioned:
1) opened keditbookmarks
2) immediately did Folder > Recursive Sort
3) keditbookmarks immediately crashed
4) drkonqi started
5) clicked "Report Bug" button
6) Dialog "Welcome to the Reporting Assistant" appeared.  Clicked the button labeled "Next".
7) Dialog "What do you know about the crash?" appeared.  Clicked the radio button labeled "Yes".  Changed drop-down menu to "Every time".  
8)  Checked box for option item "What i was doing when the application "Bookmark Editor" crashed".  Clicked button labeled "Next".
9)  Dialog "Fetching the Backtrace (Automatic Crash Information)" appeared.  Backtrace loaded Clicked the button labeled "Next".
10)  Dialog "Login into the KDE Bugtracking System" appeared.  My account detailes automatically filled in.  Clicked the button labeled "Login".  
11)  Dialog "Look for Possible Duplicate Reports" appeared.  clicked line item "308823 [Duplicate report] Crash on recursive sort".  Clicked button labeled "Open selected report".
12)  New dialog window "loading information about report 308823" appeared.  Dialog box labeled "Nested duplicate detected" appeared.  clicked button labeled "No, let me read the report I selected".
13)  New window titled "Bug Description", "Showing bug 308823" appeared.  clicked button "Suggest this crash is related".
14)  New dialog titled :Related Bug Report Appeared".  Clicked radio button labeled "Completely sure: attach my information to this report".  Clicked button labeled "Continue".
15)  Returned to window labeled "Look for Possible Duplicate Reports".  Clicked the button labeled "Next".
16)  Window labeled "Enter the Details about the Crash" appeared.  Entered the following into the text window:
"- What I was doing when the application crashed:
1) opened keditbookmarks
2) immediately did Folder > Recursive Sort
3) keditbookmarks immediately crashed

keditbookmarks frequently crashes during recursive sorting of bookmark folders in any location.

The crash can be reproduced every time.

My entire system was installed using official openSUSE packages from openSUSE repositories."
17)  Clicked the button labeled "Next".
18)  Window labeled "Preview the Report" appeared.  Clicked the button labeled "Next".
19)  Window labeled "Sending Crash Report" appeared.  Busy indicator begins rotating, and continues to rotate indefinitely.  drkonqi is not frozen nor crashed - the busy indicator just rotates indefinitely, and the crash report never finishes getting sent.

My entire system was installed using official openSUSE packages from openSUSE repositories.
Comment 1 Jekyll Wu 2013-03-18 13:16:45 UTC
I'm impressed by this detailed report, but frankly I have no idea why it stucks. I can only routinely blame the network connection or some temporary bugs.kde.org high load problem , since drkonqi works well in sending crash information to bugs.kde.org as usual, which can be observed and confirmed by querying  crash reports created recently.

By the way, the query shows that you have added (basically the same) crash information onto bug 308823 for 7~8 times today.   Please don't do that in the future: 

1. Adding the same information multiple times does not provide extra help, but only create noise. Once is good enough.
2. bug 308823 is already marked as duplicate of bug 258505, so you should follow the suggestion of drkonqi (it should have) and add your crash information onto bug 258505 instead of bug 308823.

Actually, your own activities today show that drkonqi works well for you.
Comment 2 kdebugs.anon134 2013-03-24 10:47:19 UTC
I added multiple crash reports becuase I was not aware the reports were being sent, as it appeared that drkonqi was not finishing the job.

I did not intentionally send multiple copies of the same report, nor would I in the future.  From now on, when drkonqi appears to be hanging (as it is again now), I will check https://bugs.kde.org before I submit the same report again.
Comment 3 kdebugs.anon134 2013-03-24 10:53:48 UTC
Hello Jekyll Wu,

I added my bug report from above to bug 258505 as you suggested.
Comment 4 Richard Llom 2013-03-24 11:43:04 UTC
Hello,
I have the very same problem, my report from bug 317137:

Up to the final upload page, everything works fine in DrKonqi: debugging, looking for bugs/duplicates (which is AWESOME btw, thanks a lot for this!!!), etc. However, on the final upload page it starts uploading the report, but never finishes. DrKonqi is still running, so you can just cancel and end drkonqi.

On the bug report, the attachment is fully uploaded and there is also comment which is however stripped (i.e. not fully uploaded).

I had this now two times where I attached my report to an existing bug:
bug 315649 comment #7
bug 297502 comment #17


Reproducible: Always

--
Strange that I didn't found this report while reporting mine...

Also looking at other DrKonqi reports/comments we don't seem to be the only ones.
Comment 5 Richard Llom 2013-03-24 11:43:43 UTC
*** Bug 317137 has been marked as a duplicate of this bug. ***
Comment 6 Jekyll Wu 2013-03-25 04:10:45 UTC
Git commit b5a8845c515fa6f6b40d58748f05b626d2356f17 by Jekyll Wu.
Committed on 25/03/2013 at 04:53.
Pushed by jekyllwu into branch 'KDE/4.10'.

Connect to the new signal, instead of the non-existent one

I made a stupid mistake in the final revised patch for review 108411,
which makes drkonqi never realize the attachment has been successfully
uploaded to the bugzilla.
FIXED-IN: 4.10.2

M  +1    -1    drkonqi/reportinterface.cpp

http://commits.kde.org/kde-runtime/b5a8845c515fa6f6b40d58748f05b626d2356f17
Comment 7 Jekyll Wu 2013-03-25 04:42:30 UTC
Really sorry for this stupid mistake.
Comment 8 Richard Llom 2013-03-25 10:42:44 UTC
(In reply to comment #7)
> Really sorry for this stupid mistake.

Anyway, thank you that you took a look at it and fixed it so fast!
Comment 9 Martin 2013-03-29 08:23:28 UTC
Maybe this could be posted to planetkde.org or on the KDE page as information.
Comment 10 Richard Llom 2013-03-29 09:58:24 UTC
(In reply to comment #9)
> Maybe this could be posted to planetkde.org or on the KDE page as
> information.
You mean to blame Jekyll ;-) or to inform the users?
I guess you mean for the latter... Well, there must be /quite/ a few users affected by this, but only very few find its way to bko, so I guess it just doesn't matter that much.

Also there would be (IMHO) very little information content in this message: "Well, we had an regression 4.10.1 but it is already fixed in .2." (which will come in a few days).

This would have been useful weeks ago (where I also was looking around, what is up with DrKonqi), but now?
Comment 11 Martin 2013-03-29 10:09:43 UTC
I never intended to blame anyone, least of all Jekyll. I just thought that it could be quite useful. Unfortunately, the idea to check the release schedule came to late (more specifically, directly *after* my comment... ;-) ). Sorry for the noise.