Can large u,d quark masses be inferred from the Ds→nπ decay rate?

The Ds→nπ measured branching ratio could lead to information on u,d quark masses if it is dominated by the (helicity-suppressed) Ds→ud̄ annihilation channel. The data would then suggest anomalously high mumd30 MeV masses. We find, however, that Ds→ud̄g can explain tsumB(Ds→nπ) with zero up and down quark masses. We also argue that the present Ds→nπ data may not even convincingly imply any annihilation process whatsoever.