Dr Konqi still misbehaving - advice needed

[Ian Wadham]

Hi guys,

You may remember the recent fracas with Dr Konqi suddenly refusing to submit
crash reports, because Bugzilla software (bugs.kde.org) had changed to a new
version that no longer recognised cookies.

The problem was fixed by me and the patch was released in KDE 4.14.2 and 4.14.3.
It has also been forward-ported to KF5/Frameworks by Hrvoje Senjan.

But now the problem is reappearing in the field…  See:
   https://bugs.kde.org/show_bug.cgi?id=337742#c30

This should have been fixed, see:
   https://git.reviewboard.kde.org/r/120431/ and also
   https://git.reviewboard.kde.org/r/120876/

Germano Massullo, who wrote Comment 30 re the bug, is able to reproduce
a crash in Gwenview, which then fails to get added by Dr Konqi to an existing
bug report.  The error message from Bugzilla is the same as it was before and
Germano sends a screenshot of that, among Comments 30 to 38.

The only thing I cannot get is the log/stderr output of Dr Konqi, which might
tell me what the heck is going on.

It occurs to me that Dr K is being started via kdeinit4 in Germano's case, rather
than by forking and execing, and so its log/stderr output is going "somewhere
else", not to Germano's stderr.

So my immediate and most urgent question is how can I get Germano to
recover that log output from Dr Konqi?  He is using Fedora 20, KDE 4.14.3.

Cheers, Ian W.

PS. I work on Apple OS X and my recollection of where Linux puts such
output is hazy.  I have re-visited and re-tested my fixes, using Apple OS X,
and they are still working fine.