PedestrianPositionProcessor and the OSM

I'm currently writing a new processor to write out all foot steps i.e. including those smalls steps in between one time step. I found out that we write out a little dubious positions using the PedestrianPositionProcessor in combination of the OSM: Assume the pedestrian starts his step from p1 to p2 at time t and the step requires dt seconds, then we write out p2 at time t* with t <= t* < t+dt. So what we see on e.g. the postVis and OnlineVis is the future. In my opinion, if we write out at simulation time t*, we should only write out steps which have been completed i.e. t+dt <= t*. Whats your opinion @hm-mgoedel, @hm-kleinmei, @hm-lehmberg ?