On Fri, 11 Sep 2009 02:36:21 -0500, Andrew Armstrong wrote: > >edited it with ISPF editor and found to my astonishment that if a record was >exactly 8 bytes long it was saved with a blank appended - making it 9 bytes >long! Not good for a data file! (that particular data file anyway). > >Not sure if that "feature" is still lurking - I vaguely recall it had >something to do with not upsetting the PL/I or COBOL compiler which >sometimes expected sequence numbers in cols 1 to 8. > It's apparently fixed in z/OS 1.7. Yaaaay!
I can't imagine that PL/I's or COBOL's expectation of sequence numbers is somehow assisted by adding a blank in col. 9. I can only imagine that some product had a defect that was triggered only by records with a data length of 8 (Subtract 13 from length, then EX MVC?), and that it was circumvented by changing the editor rather than by fixing the problem where it lay. Shamefully irresponsible; I suppose such an engineering decision was most plausible if the offender was ISPF itself. And ISPF Copy (3.3) no longer trims trailing blanks. Yaaaay again! -- 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
