Dear colleagues, thank you for sharing your opinions and experiences. I'm aware that tersing first or using an alternative block-preservation method is the right approach (as we are doing it for years), but it wasn't the case in this scenario.
The data only exists as a file on PC at the moment (z/OS part is gone). By browsing, it's visible that the data is still there and by comparing to a valid DUMP data set it's clear that the formatting is "a bit" wrong. There are no EBCDIC translation issues. @Brian Westerman - scenario was transmitting specific data sets to PC, for re-transmitting to another system later on (please don't comment the business process:) @Peter Hunkeler - Thank you for giving more details on this problem! This was my initial idea also - to reblock the dataset using a program. Just wondered if anyone has a better solution or an existing program for such a case before writting my own. @CM Poncelet - original was transfered using a TN3270 Windows client in binary mode (IND$FILE). @Dave Gibney - you are quite right, see above :) Jerry Mainframe Operations ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
