On Sat, 5 Jan 2008 19:44:19 -0400, Clark Morris wrote: >On 4 Jan 2008 13:50:22 -0800, in bit.listserv.ibm-main you wrote: > >>Gil, >>We're strictly talking _JCL_-_PARM_ parameters here, where there does exist >>a 100-byte length limit imposed by JCL syntax rules. >>Passing parameter values from one called _program_ to another called >>_program_ is a completely different animal and therefore not subject to the >>100-byte limit. > How does your putative program ensure that it was invoked from JCL and not by CALL/LINK/ATTACH? Verify that it's the job step task and that R14 points to SVC 3? Is there an official way a program can determine whether it was attached by the initiator and not otherwise? For the purpose at hand, wouldn't it be simpler to do CLC PARM,=H'100'; BH ERROR?
>The fine manual used to state the limit was 144 bytes and I normally > ^^^^^^^^^ Octal? How long ago? Which FM? JCL? Supervisor Services? Other? >tested for the maximum length expected by the program and used the > Good idea. >parm length. It could get interesting if the parm was of lesser > ??? The PARM had to be exactly the expected length, neither longer nor shorter? >length than expected. This was done in both COBOL and assembler >programs. >> BTW, I wonder how Phil moved the PARM into the obtained storage area? If it was the venerable BCTR; EX MVC, he'd have problems if his program was invoked by CALL with PARM>256 bytes. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html