The new observational O/Fe versus Fe/H abundance relationship for halo stars is studied in terms of several models of chemical evolution for the solar neighborhood. Nucleosynthesis products from type I (both Ia and Ib) and Type II SNs are taken into account. The behavior of the O/Fe ratio for halo and disk stars is well reproduced by assuming (1) a lower iron production in SN II than in previous theoretical prescriptions, (2) the coalescence by gravitational wave radiation of two CO white dwarfs as the scenario for type Ia supernovae, and (3) stars in the Wolf-Rayet stage as progenitors of type Ib supernovae. Nevertheless, the best agreement with the observations is obtained by adopting an IMF favoring massive star formation only at very early epochs in the life of the Galaxy. Model predictions from other plausible scenarios for the origin of type I supernovae are also discussed.