hi fredrik !
unfortunately i do not know an easy way to solve your problem but since you are using only float-inlets (hope that i am right (?)) you still have a chance
create your inlets/outlets like
for (i=0;i<arg;i++) { x->x_outlet[i] = outlet_new(&x->x_ob, &s_float); floatinlet_new(&x->x_obj, &(x->x_f[i])); x->x_f[i]=0; }
when you send a float to one of your (right) inlets, its value will be written directly into memory (x->x_f[n]), to routine will be executed. getting something on the first inlet will start your procedure and you then can access all the floats that were set before. this will do most of the times.
for is something like proxy_new() in max, which i belive is lacking in pd - right?
being no max-user i never heard of this
mgf.fd.asdf IOhannes