Tim gave a status report on the work he did with a reweighting code from W. Placzek that allows to reweight Korlaw events 'a la' YFSWW. It has been shown at the Lisbon workshop that the polar angle distribution of W-s as generated by YFSWW are less peaked in the forward region that KoralW. The effect, of the order of 2%, might induce strong modifications to TGC results. Placzek provided a code that allows reweighting KoralW events without generating new YFSWW data set. This code requires as input the 4-momentum of all SR photons and of the 4-fermons. Tim had to face some technical problems. The code cannot be run on the minis because of the integer compression (1 MeV rounding of energies). The code s now implemented and running on POTs. Results are excepted before the editorial board session of the 189 GeV TGC paper (14^th Dec.)

The new release of KoralW includes non-factorisable O(alpha) virtual corrections. Effect of those corrections have been estimated by several authors: Beenakker et al, Denner et al (the RacoonW team): they induce a distortion of the W mass spectrum but have little effects on selection efficiency and on angular distributions. The induced mass shift on the W was estimated to be of the order of 5 MeV. The exact corrections can be approximated by the so-called 'screened coulomb' ansatz which is now implemented in the new release of KoralW. In the 'screened coulomb' ansatz, the coulomb interaction between Ws is reduced by a factor (1-b)**2. Andrea has developed a reweighting code that allows reweighting old KoralW events according to this modified coulomb correction. At generator level, a Breit-Wigner fit to the W mass spectrum shows a 5MeV downward shift. However, after reconstruction and kinematical fir, the shift is 11MeV. This has been confirmed by Oliver, how performed tests on 50 MC pseudo-experiments, the mass shift ranges between 11 (4q) and 13 MeV (e, m). The blow up of the effect between generator level and full analysis has to be understood.

Stephane reported on the technical test being performed together with Opal. Charged TGC of the two experiments will be combined at the OO level, including systematics. The results of 2D fits will be compared to those obtained from the LEP combination procedure used for Moriond and Osaka. To really evaluate biases on fitted values and errors induced by the LEP procedure, a study on MC pseudo experiments has to be performed (not foreseen for the time being!!!).

Patrice presented two transparencies from Djamel, who corrected a 'feature' in his fitting program. As the result, the expected W mass error from the tau channel at 189 GeV improved and is now below 200 MeV (198.5) for 200pb-1.

David presented a status of the improvements on the Gauge Neutral Couplings from the ZZ channel. More events have been used to compute the expected cross section from the YFSZZ generator and the parameter setup has been changed (aem, etc..). As a result, the 95% C.L. on f4g went from [-0.31,+0.31] to [-0.27,+0.28]. Selection efficiency variation with couplings has to be taken into account.

Guillaume presented some studies he performed on the variation of the expected error on the W mass from e and m channels as a function of the NN output or/and the kinematical fit Chi2 cut values. The NN cut has not been optimised from some time. As the expected error does not vary with the NN cut value. It was suggested to Guillaume to choose same cut as for the cross section : the one which maximises the product efficiency*putity. Some background events were missing for the electron channel. Variation of the expected error with Chi2 cut value, was also shown. Conclusions are expected for the next meeting.