David, > > The "I" in ILE stands for integrated, just try >and call a CEE procedure from ILE CL. > I'm sure you are thinking of some specific situations, but in general I do quick and dirty testing of many CEE APIs using CL. One fragment is: DCL &DATE *CHAR 8 DCL &LILDATE *CHAR 4 CALLPRC CEEDAYS PARM(&DATE 'MM/DD/YY' &LILDATE *OMIT) > > I still don't understand the re-compile bit. I >was under the impression that observable AS/400 programs had the >unique ability to be regenerated to support system enhancements. > To simplify horribly, observability is great for situations like having a program specify an 'ADD' and at different points in time the underlying machine has a CISC ADD, a RISC ADD, and a floating point assisted ADD. As the underlying machine changes (CISC to RISC; 48-bit to 64-bit; N registers to N*M register) the machine can simply re-translate the 'ADD' to generate new instructions using new system facilities. In the case of a pointer vs a value on the stack however, it would be closer to say the programmer went "oops", that 'ADD' really should have been a 'MULTIPLY'. Observability does not help in this situation (and please, no one go down the path that the machine could theoretically analyze the stream to find the exact instances of the "oops" and change the instruction stream because my earlier append really did gloss over some of the finer points of what is happening in favor of expressing the ending result). Bruce Vining +--- | This is the Midrange System Mailing List! | To submit a new message, send your mail to "MIDRANGE-L@midrange.com". | To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com. | Questions should be directed to the list owner/operator: email@example.com +---
As an Amazon Associate we earn from qualifying purchases.
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.