The z/OS native C compiler does not compile the source code when I do circumflex, only logical not. I may resort to the digraph but that would make the code less readable. However, I would appreciate if you show here the JCL that compiles source code with circumflex on z/OS, maybe with a short 3 line program with that circumflex. My assumption is that you have some compile option that I do not know about (and could not find in the manual) that allow C to understand the circumflex. Ze'ev Atlas
________________________________ From: Charles Mills <[email protected]> To: [email protected] Sent: Friday, December 7, 2012 12:41 PM Subject: Re: How to load logical load x'ac' This problem is very solvable but the solution utterly depends on the details. What exactly do you mean by "when other people load this"? I'm also not sure exactly why you are using circumflex for logical not rather than for XOR. I use common source code across Windows and z/OS all the time, so I am something of a small-scale expert on this topic. I compile the same source code (translated ASCII<->EBCDIC) in both places, and transfer it from one place to the other using Ipswitch WS_FTP 12. I have no translation problems such as you describe whatsoever, so this problem is solvable. You can configure your ASCII and EBCDIC code pages in your FTP server (SBDATACONN and SBTRANS), you can specify a custom translate table on the session (SITE SBDATACONN or XLATE), and at compile time (LOCALE). You can also use digraphs ("xor") instead of "funny" characters in your source code. Charles -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Ze'ev Atlas Sent: Friday, December 07, 2012 7:37 AM To: [email protected] Subject: How to load logical load x'ac' Hi all I edit my sourcecode base in ASCII (where it comes from) and I replace the circumflex (^) that C loves with logical not (x'AC' in ASCII) that the mainframe C requires. Then I load the thing to the mainframe with wc2370 which defaults to codepage 037 and my mainframe session happily shows logical not (¬ = x'5F'). When other people load this, it becomes something like comma (,). How could I convince all flavors of EBCDIC and EBCDIC upload to recognize logical not correctly? ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
