Writing to be cited or really to be read: towards the dynamics of open-access publishing