Hi,
I won't be able to attend tomorrow's meeting due to a concurrent 
practice talk session which I am organizing. Here is an update on the 
partial branching fraction measurement from mx-q2:

1) by using the new reweighting files provided by Dominique for the 
charmless exclusive BFs, we are now able to compute the associated 
systematic effect instead of taking the numbers from BAD540. We did this 
for both the CLEO and BELLE ellipses. The net effect is a small increase 
of the uncertainty when using CLEO numbers, and a more substantial 
increase when using the BELLE numbers. Daniele has an explanation for 
this effect, and he will report it at the meeting.

2) the theoretical error for the CLEO ellipse was still computed by 
using the most distant (mb,a) points from the central values. Now all 
the available points on the ellipse are taken into account and the 
theoretical systematic uncertainty is defined as the maximum (positive 
and negative) deviations from the default values when moving along the 
ellipse. Please remember we are talking about partial branching 
fractions, so the theoretical uncertainty in this case is a second order 
effect with respect to other ones.

You will find more quantitative results in

http://www.slac.stanford.edu/~bozzi/tabelle.ps

there are 3 tables in that file. The first one is a copy of Table 2 of 
the current BAD976, which I put for reference. The second table is 
obtained by using CLEO numbers and by revaluating the theoretical and 
signal modeling uncertainty as explained above (for this reason, all 
columns but \sigma_{theo} and \sigma_{sig} are unchanged with respect to 
the first table). The third one uses BELLE numbers. Each row represents 
a q2 cut, the mx cut is 1.7 GeV.

Our conclusion is that we can safely switch to the Belle ellipse.
Regards, Concezio and Virginia.