Hi Vadim,
in this case, i am using for/next to avoid result of allocation that spread to all members of jobgrade. in my logic script, for/next is like function to loop for all possibility of that employee whose has related jobgrade.
when i use :
option 1: (with for/next loop)
*DIM JOBGRADE WHAT=%JOBGRADE_LOOP%; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
this will loop all members of jobgrade one by one.
*DIM JOBGRADE WHAT=JG1; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
*DIM JOBGRADE WHAT=JG2; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
*DIM JOBGRADE WHAT=JG3; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
*DIM JOBGRADE WHAT=JG4; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
*DIM JOBGRADE WHAT=JG5; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
this issue only the performance.
option 2: (without for/next loop)
*DIM JOBGRADE WHAT=bas(JOBGRADE); WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
this will be:
*DIM JOBGRADE WHAT=JG1, JG2, JG3, JG4, JG5; WHERE=<<<; USING=NO_JOBGRADE; TOTAL=<<<;
which means in WHAT, the value will be rollup (this is no issue cause, the relationship of employee and jobgrade is 1:1) and it will allocate USING NO_JOBGRADE then store in WHERE place, which all members of jobgrade.
the issue of option 2 is the result is spreaded into all member of jobgrade, which expects only for JG1.
the reason why i am not using *REC, i think it kinds of not possible, perhaps you have any ideas. the requirement perhaps like:
WHAT WHERE USING
INTERCO NO_INTERCO >>> <>NO_INTERCO
TIME <every month> <<< NO_TIME
thank you.