> Is it multi-platform.
Sort of. I edit, compile and unit test on a little white box. The only
environment in which it "really" runs is z/OS. The fopen()'s work on the little
white box but (1) the little white box does not know about TSO userids and (2)
I suspect the fopen()'s may be surrounded by #ifdef __MVS__.
I just looked at the code. No #ifdef, but a real if based on a platform switch.
The filename is the same in both cases, but z/OS gets an lrecl=80, noseek that
the little white box would not understand. The parsing of the filename from the
config file or command input is radically different white box versus Z. A lot
of platform-dependent screwing around with member names, DD:, assumed filetype
suffixes, and so forth. I see the following comments in the code but I would
have to do some analysis to be sure they are accurate. Let's assume they are.
// Under z/OS all names are prefixed with //
// Under z/OS all names that do not begin with * or / or have a ":" in the
second position are quoted
"Second position" would appear to mean [2] which is really the third character
such as in DD:. The product uses * to mean "allocate a SYSOUT dataset."
Charles
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf
Of Paul Gilmartin
Sent: Tuesday, September 25, 2018 11:49 AM
To: [email protected]
Subject: Re: TSO PREFIX, FTP and z/OS upgrade
On Tue, 25 Sep 2018 11:27:18 -0700, Charles Mills wrote:
>POSIX(ON) tends to change everything.
>
>I suppose I should qualify my remarks about fopen() by saying I am a big
>believer in quotes and fully-qualified names. Of course, I am developing an
>"enterprise, production" product, not a user productivity tool. I suspect all
>of my "legacy" fopen()'s use quotes.
>
>The product runs POSIX(ON) FWIW.
>
Is it multi-platform?
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
