> Please advise. I know nothing, but I'd guess that if $QIOW itself is happy, then the problem is not in the $QIOW parameters themselves, but in their contents, if that makes any sense. For example, if "chan" were bad, then I'd expect an error directly from $QIOW, but if some goofy parameter value is hidden inside "atr", then I might expect a bad IOSB status. > But with the print statements I couldn't find cause of the problem. They didn't help me, either. One statement is not a complete test case, and my psychic powers are too weak to provide one here. My advice would be to supply a complete test case if you expect anyone else to debug your code. Are you still trying to set an absurd value for the global buffer count? (And you thought that starting a new thread for the same problem was a good idea?) http://h30499.www3.hp.com/t5/x/x/m-p/6129533 > Why would anyone ever want to creaate a file with SYS$QIO(w) ? Info-ZIP UnZip does it. See [.vms]vms.c in any source kit. Recent UnZip code shows some of the quirks involved in using ODS5 extended file names with $QIO. Of course, in this case, I have no idea what's behind "name_desc" (or practically anything else).