Three ways of dealing with rescattering effects in B+/- --> K-o pi(+/-) are
compared, in order to determine the weak phase gamma from these processes
and B+/- --> K+/- pi(o). We find that neglecting these contributions altoge
ther may involve sizeable errors in gamma, depending on the rescattering am
plitude and on the value of a certain measurable strong phase. We show that
an attempt to eliminate these effects by using the charge-averaged rate of
B+/--->(K+/-Ko) suffers from a large theoretical error due to SU(3) breaki
ng, which may be resolved when using also the processes B+/--->pi(+/-)eta(8
).