I made TWO errors talking about "INTDATE" It is a compile-time, not a run-time, option. (So again, if you don't compile on both LPARs, then it won't impact you).
Also, it impacts only COBOL intrinsic functions - not LE callable services (but setting it to "LILLIAN" will make the intrinsic functions give the same "answers" as the callable services. Sorry about the errors. > -----Original Message----- > I don't know what you wanted to "hear" from IBM. However, if > your question is whether differences between these options > will change run-time behavior, the answer is DEFINITELY yes. > > 1) TRUNC - is a compile-time option. Do you do compiles on > both LPARs? If not, a change won't impact anything. If so, then: > TRUNC(OPT) is DEFINITELY a better performance option. > TRUHNC(STD) *will* change the behavior of applications that > have binary (USAGE COMP, COMP-4, or BINARY - not COMP-5) > where the individual fields can have values larger than the > PICTURE specifies. (e.g. 32000 in PIC 9999 COMP) field. > > 2) INTDATE is a run-time option that ONLY impacts CALLs to LE > "date" features. (Such calls can be explicit or implicit by > use of COBOL date intrinsic functions). > If you use LE callable services in non-COBOL (e.g. PL/I or > Assembler) applications, then you PROBABLY want to use > "LILLIAN". If you are a COBOL-only shop, then using ANSI > will give you "portable" results that conform to the ANSI Standard. > > *** > > Is this the type of information you are looking for? > > This really IS all described (in more detail) in the IBM > documentation. > > <[EMAIL PROTECTED]> wrote in message > news:<OF2F330E87.D953EB36-ON85257140.00629F35-86257140.006482F > [EMAIL PROTECTED]>... > > I'm merging two lpars and, not surprising, there are some > customization > > differences in COBOL and LE. Most are easily changed and > don't really > > affect how jobs run. However I have two differences that > concern me. > > > > LPARA (the sending lpar) has INTDATE=ANSI and LPARB (the > receiving lpar) > > has INTDATE=LILIAN. Also LPARA has OPT as the default for > TRUNC while > > LPARB uses STD. I've asked IBM about these but they just quote me > > sections of the COBOL manuals dealing with the options. > They don't say > > anything about what might happen when I change one value to another. > > > > Has anyone ever made these changes? I'm more scared about > INTDATE; TRUNC > > is more easily explained. > > > > Thanks > ---------------------------------------------------------------------- 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
