* FTOF geometry in FastSim: we've changed the detector layout 
according to the new design -- see picture linked to the agenda.
This picture has been produced using the root macro written by
Matteo months ago; I've assumed it's still working fine. The xml
modification still needs to be committed.

* The barbox structure (needed around the FTOF quartz sectors like
  for the FDIRC) has a small X0: about .8% per wall according to
  the DIRC NIM paper. It seems it is not simulated in FastSim for
  the FDIRC => we can probably neglect it as well for the FTOF.

* We have standalone G4 jobs running in batch to estimate
the # of photo-electrons as a function of p, theta and
particle type. A LUT will be built using these results
to make this number more realistic in FastSim.

* We're also working on making the FTOF resolution more
realistic. Two paths are followed in parallel:
  - simulate the reconstruction using G4 standalone data
as input; parameterize the results and inject them to FastSim
  - choose a realistic (and possibly a bit conservative)
resolution per photo-electron; define the timing error as
resolution / sqrt( # of ph.e. ), the number of ph.e. coming
from the LUT described above
Method 1 would be the best but it depends on various 'new'
things => it may not converge in time for the production
which is why we have method 2 as backup.

* We're working on providing tools (pulls, etc.) to quantify
the quality of the FTOF contribution to the PID likelihoods.

* When the above steps are completed, one could tune the LH selectors.