In the last ten years, the number of models of circumstellar envelopes has
exponentially increased, both concerning evolved stars and forming stars wi
th their planetary systems. The main progress was intoduction of chemistry
and coupling with dynamics. However, paradoxically, the nucleation theories
did not evolve at the same rate; moreover, dynamics and nucleation are rar
ely analysed at the same degree of accuracy: one is of them is always mostl
y drawing attention. At the same time observations showed the presence of d
ust grains at a lot of sites where it could not be expected from usual quas
i equilibrium theories. Self consistent systems are omnipresent. Thus, it i
s important to distinguish causes and effects, even if the phenomena are ar
e strongly coupled.