Hello folks,
while writing the paragraph on X and neutrino four-momentum reconstruction
I reviewed the code and the logic and I have a couple of comments:
 1) B0rectrk is a bit map and not an index. The way it is implemented
right now it is hardwire and relies on the assumption of a single
candidate being reconstructed. I think we should do something slightly
safer ...
 2) in the case of identified leptons in the X system , when
reconstructing the X 4-momentum,  we assign the electron and muon mass.
Since our analysis assumption is that we have only one lepton in the event
(and this needs to be true in the signal), would it be wiser to assign the
pion mass to these candidate? If there is actually an additional lepton in
the background the reconstructed MX will be pulled up, which is not bad...
 3) do I read the code correctly that we are using the 2C fit with the
equal B mass constraint? Also reading the code it looks like that when you
turn the smearing on you turn off this additional constraint. I have no
clue what this means (I guess it is a testing configuration) but it sounds
wierd

	ciao
	ric
P.S. I committed a revised version of chapter 3 according to what I think
we agreed upon. Let me know if you are not happy with it.