POSIX run-time option vs TSO and Batch

2006-01-19 Thread Wendell Simmons
I have a C application that uses Unix Services and therefore has to run with POSIX ON - logic has been added to the C program to display the setting of the POSIX option. This C application is invoked from a COBOL program that is linked with a CEEUOPT module that has the POSIX option turned ON. Th

Re: POSIX run-time option vs TSO and Batch

2006-01-30 Thread Wendell Simmons
Found the problem. Thanks to everyone for their input. The problem is that in the chain of modules being invoked, one of them is doing a COBOL "CALL ISPLINK" with a SELECT PGM option. This distorts the enclave inheritance for the POSIX ON option for later modules in the chain. Later modules wil

Re: POSIX run-time option vs TSO and Batch

2006-01-31 Thread Andy Robertson
On Mon, 30 Jan 2006 17:42:33 -0600, Wendell Simmons <[EMAIL PROTECTED]> wrote: >Found the problem. Thanks to everyone for their input. > >The problem is that in the chain of modules being invoked, one of them is >doing a COBOL "CALL ISPLINK" with a SELECT PGM option. This distorts the >enclave i

Re: POSIX run-time option vs TSO and Batch

2006-01-31 Thread Bill Manry
On 19 Jan 2006 11:19:08 -0800, [EMAIL PROTECTED] (Wendell Simmons) wrote: >I have a C application that uses Unix Services and therefore has to run >with POSIX ON - logic has been added to the C program to display the >setting of the POSIX option. I know this is resolved, but I'm curious about th

Re: POSIX run-time option vs TSO and Batch

2006-01-31 Thread Paul Gilmartin
In a recent note, Bill Manry said: > Date: Tue, 31 Jan 2006 16:25:10 -0800 > > I know this is resolved, but I'm curious about the foregoing. Most UNIX > services are available regardless of the LE POSIX option. Are you sure you > need POSIX(ON)? > Doesn't the POSIX setting affect the r

Re: POSIX run-time option vs TSO and Batch

2006-02-01 Thread Wendell Simmons
Yes, POSIX ON is definitely required. Without it, I experienced the following: ++ The C application uses Unix Services and therefore has to run with POSIX ON - logic has been added to the C program to display the setting of th

Re: POSIX run-time option vs TSO and Batch

2006-02-01 Thread Bill Manry
On 31 Jan 2006 17:06:28 -0800, [EMAIL PROTECTED] wrote: >Doesn't the POSIX setting affect the resolution of pathnames? >where normally fopen("FRED") refers to a file relative to the >current working directory, with POSIX(OFF) it refers instead >to a DDNAME? Admittedly, this is CRTL behavior, not

Re: POSIX run-time option vs TSO and Batch

2006-01-20 Thread Andy Robertson
On Thu, 19 Jan 2006 13:19:00 -0600, Wendell Simmons <[EMAIL PROTECTED]> wrote: >I have a C application that uses Unix Services and therefore has to run >with POSIX ON - logic has been added to the C program to display the >setting of the POSIX option. > >This C application is invoked from a COBOL

Re: POSIX run-time option vs TSO and Batch

2006-01-20 Thread Wendell Simmons
In response to your question about the highest level module in the enclave being the one that determines the POSIX option; If I fix the highest level module in the enclave in the TSO environment (module VORPTMEN) to have POSIX ON via a specially linked version of CEEUOPT, I get the following error

Re: POSIX run-time option vs TSO and Batch

2006-01-20 Thread Andy Robertson
On Fri, 20 Jan 2006 11:14:12 -0600, Wendell Simmons <[EMAIL PROTECTED]> wrote: >In response to your question about the highest level module in the enclave >being the one that determines the POSIX option; > >If I fix the highest level module in the enclave in the TSO environment >(module VORPTMEN)