> Was this with RECFM=F or RECFM=V? The "SYSIN" input data set was read (by the program that was executed by the JCL submitted via the INTRDR) with a DCB that specified RECFM=VB,LRECL=252,BLKSIZE=256. It used the LOCATE form of the GET macro, so it just looked at the RDW to see how long a record it actually had been pointed to.
> To this end, is there anywhere documentation on how > the attributes of the SYSIN data set as perceived by > the problem program reading it are determined by the > JCL atributes of the internal reader and the > characteristics of the instream data? Not that I am aware of. The only thing I have seen is the APAR text, which, if I remember correctly, describes some algorithm that attempts to provide an intelligent default, such as RECFM=V if there are any records in the in-stream data set that were written with such a DCB to begin with. Maybe that works. I dunno. The very low-level code in JES2 that actually picks apart the SPOOLed in-stream data set records returns a record and its length. What gets done with the length ultimately depends on code at least one layer up, past the QSAM DCB-to-VSAM ACB/RPL compatibility interface, which looks at what the DCB says. RDWs will be manufactured if the DCB indicates RECFM=V. Thus, one can read a standard FB-80 in-stream data set with a RECFM=V[B] DCB, and all will be well. Every record will have a length of 84 in the RDW (in the simple-minded instance where the JOB was written to the INTRDR with an LRECL=80 RECFM=F[B] DCB or ACB). -- WB ---------------------------------------------------------------------- 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
