On Fri, 11 Oct 2013 15:25:27 -0700, Charles Mills wrote: >Right. "when it is read" could equally well mean "when it is read originally >by JES2" or "when it is read finally by the problem program." > >In fact, the more I think about it, the more I like the first interpretation. >From JES2's point of view, that's the "reading" part. The other part is not >"reading," it's passing the data to the problem program. > >That's not to say that's what the doc writers meant, or how it actually works. > This matter of shifting/overflowing/truncation is all so chaotic that perhaps it would be better to allow substitution in instream data sets only when RECFM=V.
Hmmm ... In the Bad Old Days, I used to make insertions in punched cards: DUP up to the point of insertion; press hard with my thumb on the card in the read station so the capstan skidded; type inserted data; DUP the rest of the card. Conversely to delete. I wonder how many warranties I voided? It was the university's equipment. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
