As I read the description of FILEDESCRIPTOR in the Reference, I can't tell whether the descriptor is assumed to contain "records" separated by x'15', as <OE and >OE do, or whether the stage assumes an unstructured stream as HFS(REPLACE) does and it's the programmer's responsibility to use (DE)BLOCK TEXTFILE to separate records.
I hope the behavior is HFS-like rather than OE-like so that FILEDESCRIPTOR can process an unstructured stream. Does this merit a RCF? Also, I don't believe the description makes it clear enough that when FILEDESCRIPTOR is the first stage it reads from the descriptor and passes records to the next stage, and when not the first stage it accepts records from the preceding stage and writes then to the descriptor. Thanks, gil
