Semileptonic Decays and CKM Matrix Elements in a TCF

We argue that a Tau-Charm factory (TCF) could improve our knowledge of CKM matrix elements since it is an incomparable tool to check the models and methods applied to extract $V_{bu}$ from $B$ decay partial widths. We report on some recent proposals to improve on parton model. Turning to exclusive decays, we compare the predictions from quark models, QCD sum rules, effective Lagrangians and lattice QCD. Quark models have anticipated on heavy quark symmetry. Their difficutly to account for the $q^2$ dependence might be partly cured by relativistic corrections. QCD sum rules and lattice seem to disagree on the $q^2$ dependence of axial form factors. We discuss the extrapolation from $D$ to $B$. Present uncertainties do not allow to predict reliably the $B\ra\pi,\rho l \nu$ matrix elements. We argue that QCD sum rules are in a good position to study $q^2$ dependence down to $q^2=0$. Lattice QCD is able to check the heavy quark scaling laws for the heavy to light semileptonic decays. It seems to confirm the increase of $A_2/A_1$ from $D$ to $B$ meson predicted by the scaling laws. Finally semileptonic decays in a TCF would give incomparable informations on $K\pi$ phase shifts, and on $K^{**}$ resonances.