On Monday 02 March 2009 23:18:47 you wrote:
>
> Yes, 100% reproducible. But just in case ... what kind of state
> machine bug might produce this? The machine is running and I can do
> the same thing from the local deployer, IIRC.

Some years ago, we had the SM queries crashing because of the ever changing
'current' state pointer. I see now the SM code fixed this (it copies current,
then checks validity, then returns current->getName() ), so then it's CORBA.
We're converting a const string& to a string for use in a return value, the
bug might be in the conversion code

I need a RTT with -DCMAKE_BUILD_TYPE=Debug gdb trace or to reproduce it here
in order to fix this...

Peter