Fabrizio raised the question about the prolongation from TRD to the outer
detectors (TOF, RICH, PHOS). This functionality is needed urgently in order to
do the TOF reconstruction.

Karel pointed out that the time propagation easily can be incorporated in
the Kalman filter. It is just the integral of the times at each step given 
the mass of the particle. Then one compares with the time in TOF and assigns
the probability. This question has to be discussed during the tracking 
session on Friday. Some example of reconstructed TRD data is needed by TOF.