Hey all,

as promised in the meeting today, here is a list of suggestions for the 
new production. All of these are issues that came up during the review of 
my analysis (although I phrased some of them in the context of moving to 
r22), and I think they equally apply to our Vub analysis.
Some of these are phrased as questions, so we should think about these.

* switch to NN based mu selector (already happened as I understand)

* store the energy/momentum for each charged track on the signal side in 
our ntuple (as I said in the meeting, something like this exists for the 
neutrals, so this is a matter of expanding this a bit, should be really 

* exclusively reconstruct and tag B->Dlnu events with D->Kpi (the idea 
of this was to test for efficiency biases in the soft neutral filter that 
we have been building, this is probably a bit more work?)

* investigate whether the cut on neutral energy should be raised (this is 
50MeV in r18, I don't know what it is in r22, but this is something we 
should look at I think, 50 MeV is quite low)

* store information of vertex probability for the signal side (is this 
easy to do, i.e. try to vertex the signal side and store the probability? 
the idea was that this could help for bkgd rejection)

* Do we need to look at PID corrections again (when moving to r22)?

* Should we lower the momentum cut for the second lepton? This cut is 
there to reject leptons from cascade decays, which have a soft momentum 
spectrum, so cutting at 1GeV for these leptons might not be optimal in 
terms of rejecting those backgrounds (rather have a lower cut to reject 
them more efficiently).