We discuss the treatment of radiative transfer through a two-component discrete stochastic mixture. A broken cloud field can be modeled as such a stochastic mixture, with the clouds and clear sky representing the two components. The integral equation approach of Titov for a Markovian mixture is shown to be equivalent to a differential model introduced in the kinetic theory literature. Simplifications and extensions of this model are also discussed. The simplifications include a renormalized equation of transfer, as well as various diffusive models. The extensions involve more accurate descriptions utilizing additional radiative transfer equations, as well as allowance for non-Markovian mixing statistics.