On Wed, 20 Aug 2014 20:14:55 -0400, Thomas Kern wrote: > >I have used XMITIP to send PDFs from a z/VM system. The PDFs that I >created to be send out were created as ASCII data then they were >transmitted at S/MIME attachments of pure "DATA" to Windows recipients. >The PDFS have been "compressed and encrypted" in a PDF fashion but it is >all ASCII data. > I'm astonished and curious. XMITIP is available for both z/OS and z/VM? I'm more familiar with VM SMTP than z/OS -- we haven't the latter operational.
I believe the SMTP VM/task accepts EBCDIC files in the spool as either plain text or NETDATA format; decodes the NETDATA, and translates EBCDIC to ASCII for network transmission. Does it treat apparent binary attachments specially, suppressing EBCDIC<->ASCII conversion? Is the code page pair constant within SMTP (037<->819?), regardless of locale of the host system? What about the LF/NL nightmare. I know CMS Pipelines translates 0x25<->0x0A; z/OS iconv 0x15<->0x0a. The SMTP server will add a network line separator, ASCII <CR><LF> after each record (someone said FB 80). And this all gets undone at the other end!? You said ASCII data. Does this mean no code points above 0x7F? I'm curious. Will someone volunteer to send me a brief PDF attachment, via XMITIP, to my address above? (LISTSERV would discard it.) <minor rant>When I email to my VM account, the MIME headers may say character set is ISO8859-1, but the document has obviously been translated to EBCDIC. Shouldn't the MIME headers be adjsted to reflect this translation, e.g. to IBM-1047?</minor rant> Thanks, gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
