RE: Moving data into dynamic storage

>I wonder whether a 'User Space' may be
>an answer for dynamic storage. Anyone have
>some experience with RPG and User Spaces?

Yes.  I wrap QUSCRTUS, QUSCUSAT and QUSPRTUS in a procedure that I call in
place of ALLOC.  I pass in the space name and pointer but there's no reason
the proc can't generate a random space name, given that we really don't care
what it is.

I then wrap all the pointer math in other procs so that if I decide to
change the model (say to teraspace pointers) my mainline won't change.
  --buck