Why does MQSeries convert character 0x1A to 0x7f and character 0x1C to 0x1A when the sender's CCSID is 819 (ISO8859-1) and the receiver's CCSID is 437 (PC-ASCII US)?Every table that I have seen for ISO8859-1 shows character 0x1A to be the SUB control character and 0x1C to be the FS control character. How can I keep this conversion from happening? I looked at the ccsid conversion tables in the MQSeries\conv\table directory and sure enough, the table being used (033301B5.tbl) converts the characters as I explained above. Any suggestions? Can I just modify the 033301B5.tbl file on the MQSeries Server machine?