Gil, Can you pass a type= memory file? I think it it has persistence.. I have tried it in C ...and know you can create it..
Scott On Dec 1, 2017, 4:46 PM -0500, Paul Gilmartin <[email protected]>, wrote: > On Fri, 1 Dec 2017 14:55:08 -0600, John McKown wrote: > > > > > And the OS feature I miss most in UNIX is temporary data sets. > > > > Yes, very much. My method, in general, is to create a "temporary" file in > > the ${TMP} directory. Open it. Unlink the file entry. Use the FILE * and > > pass it around to other programs so that they can access the data. > > > It's easier than that. In: > http://pubs.opengroup.org/onlinepubs/9699919799/functions/tmpfile.html#tag_16_618_03 > ... In some implementations, a permanent file may be left behind if the > process calling > tmpfile() is killed while it is processing a call to tmpfile(). > > This is the exception that proves the rule that once tmpfile() nas completed > killing the calling process leaves no permanent file behind. tmpfile() ought > to > unlink the file before returning. > > Unlike most UNIXen, the earliest OMVS violated this, deleting the temporary > file > only when the process exited. I suspect this was an accommodation to > implementing > temporary files as Classic data sets. I believe it's better now. Don't know > about > POSIX(OFF). > > As an alternative, you could pass just the descriptor. > > Of course, this doesn't cross job step boundaries. Alas, JCL doesn't support > temporary UNIX files. > > -- gil > > ---------------------------------------------------------------------- > 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
