Version:           svn (using Devel) 
OS:                Linux

At the moment i am experiencing a lot of crashes (konqueror/dolphin) and have to abuse the kbugbuster to some extent. The first thing i want to know is: has this crash been reported by someone else. So i select that crash is reproducable and i have additional information, so that the reporter app allows me to report a bug at which point it finally searches the BugDB. What would be a lot more practical if it would search for the backtrace in the bugDB as first step. Also, most times i do not file a report even if no similar backtrace/duplicate bug is found simply because it is a 'random' crash, sometimes here, sometimes there and hard to be nailed down.
In my opinion it would make sense if the app would push the backtrace to the server, so anonymous crash statistics would give a better picture of which crash bugs are in the wild and with which QT/KDE versions specifically.

Reproducible: Always



Expected Results:  
So to sum up duplicate bug search should be first in crash reporting workflow, and collecting crash stats without need to file a bug might be a good information source for developers. Not?

OS: Linux (x86_64) release 3.2.0-rc1
Compiler: x86_64-pc-linux-gnu-gcc