I just fixed the bug that crippled the track reconstruction for the last month. Some relatively-new variables in the magnetic field map code weren't being initialized if the magnetic field map was being generated rather than being read in from an EVIO file. The b1pi histograms should look normal next time. - Paul