Here are some screenshots of FRESHS in action.
This is how the server looks like just after starting with the sample configuration files.
This shows how multiple clients calculate the initial run of an FFS simulation in parallel, resuming their runs after delivering a configuration point to the server. Thereby, the length of a trace is preset in the server's configuration file.
The next interface position in FFS can be determined using automatic interface placement methods. Here, exploring runs are started from the last known interface position and their information is used to determine the next optimal interface position.
Ghost runs are used to bridge the waiting time if clients would be idle because of waiting for the current stage to be completed like in this case. Instead of waiting, so-called 'ghost runs' are started on the already collected points. The information of these runs are stored in a separate database and then transferred if a point with an existing ghost run is drawn during the next calculation stage.
A database editor can be used to monitor the progress of the simulation. In addition, if something goes wrong during the simulation, the database can be edited and the simulation can be resumed with the -r option of the server. The database provides a powerful source for analyzing the whole simulation data.