RE: Match program object to source member; CAN'T use source last update timestamp in object

>>operationally identical comparison between two
>>program objects seems to me too subjective
>>for quick a program fix
>
>Gene, I'm somewhat surprised that you feel
>this is subjective.

Hi Dan,
  I'm not in Gene's league, but I would be surprised to find that there is a
way for a simple program that can determine functional equivalence.  Imagine
compiling a program where the compiler generates MI calls to some OS
function.  Now a PTF gets applied, supplying a new, speedier version of that
OS function.  Re-compiling the program _without any source changes_ will
result in MI calls to the new, enhanced OS function.  The two programs will
produce exactly the same results but the MI will not be the same.

I can think of dozens of ways the compiler might generate different MI
instructions that are functionally equivalent, depending on PTF level,
optimiser settings and compiler options in effect.  The only way that I can
think of to determine functional equivalency is to run both programs with
the same set of test data and compare the results.  Or write an OS400
emulator that can simulate executing an RPG program.

Now I'm off to read chapter 38 and be humbled.
  --buck