Comment on "Decoy State Quantum Key Distribution"

The main claim by H.K. Lo et al that they have for the first time made the decoy-state method efficiently work in practice is inappropriate. We show that, prior to our work, actually (and obviously) none of proposals raised by H.K. Lo et al can really work in practice. Their main protocol requires infinite number of different coherent states which is in principle impossible for any set-up. Their idea of using very weak coherent state as decoy state doesn't work either by our detailed analysis. The idea implicitly requires an unreasonablly large number of pulses which needs at least 14 days to produce, if they want to do QKD over a distance of 120-130km.