Hi, probably this is not the best place for such a question but i think that most of you have the right knowledge of fortran (better than me of course) to help me.

I'm writing a genetic algorithm code which makes use of external modules. The code is like the following (i put between quotes "" the not important stuff):

PROGRAM example

USE external_module

IMPLICIT NONE

INTEGER(KIND=my_int_kind) :: "integer stuff"

REAL(KIND=my_real_kind) :: "real stuff"

TYPE(my_der_type) :: type_one,type_two

"Allocation of some components of type_one which have the pointer attribute to have variable dimension"

CALL initialize_type(type_one)

CALL show_me(type_one)

type_two=type_one

CALL modify_type(type_two)

CALL show_me(type_one)

END PROGRAM example

What actually happens is that after having modified type_two, type_one also is modified. I know that there could be thousands reasons for this (some of them have already been checked) but, what i would know is that if, from what i wrote of my code, there is something wrong which i'm doing with the derived types and that determines the wrong modification. I'm using Compaq Visual Fortran 6.0.

Thanks

Even if anyone has the answer for my question, maybe some of you could be interested in it (which i found in a manual). Also because i have another question.

The assignment:

type_two=type_one

when the derived type has an allocatable array of pointers as a component, is actually equivalent to

type_two=>type_one

so it is equivalent to a target assignment and this is why type_one was changing with type_two.

Now, the question is:

Because i'm going to modify my whole program to handle this, do you know if assigning my derived type component by component (instead of a single =) is going to produce the same result or what i need for it?

Thanks