On Fri, 17 Oct 2008 08:35:28 +0200, Hunkeler Peter (KIUK 3) wrote: >Its alreqdy 8:30am here but it might still be too early >to understand your sample. The result shown indicates > I didn't understand it last night.
>that you implemented an "append", which of course is >equal to "insert after last record". Not what the OP >wanted. > >But why are you reading the data set in first? ALLOC >with MOD followed by EXECIO DISKW accomplishes the same. > >I don't understand what the do loop is for. > I don't understand what the "parse" instruction inside the unnecessary do loop is for. I was disappointed. I had envisioned something like what a colleague told me long ago. As an undergraduate, he had learned on a Scientific Data Systems (later XEROX) Sigma system. He explained that program source files, necessarily line-numbered, were stored in something like a VSAM KSDS file, with the line numbers as keys. So it was possible to update, insert, and delete records in place. Of course, this behavior is hard to distinguish from an editor's reading the entire DSORG=PS file into main storage, updating it in the buffer, and rewriting it. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html