Service program signature violations?

["Jon Paris" <Jon.Paris@xxxxxxxxxxxxxx>]

 >> Yuk!  I thought that that was ridiculous from the outset.  Was there a
rational reason for this design decision?

Probably but I wasn't party to those discussions or I'd have remembered it!
<grin>  Best I can think of is that there is no guarantee that the compiler
will output the procedures in sequence and/or the optimizer may move
procedures around in the module for greater efficiency.  For example I would
assume that a procedure that is not exported would probably be relocated
within the module for proximity with its callers.  I don't know for sure but
that is my best guess.  Having the proc names exported in alpha sequence at
least is predictable (if annoying).  I'll copy a friend in Rochester on this
and see if there is another reason behind this.

 >> What are the chances of fixing this so that it works the way it should?

None I suspect.  It will be working as designed and some folks may already
rely on it.  One of the bains of a compiler/linker writers life is that if
you get it wrong you can only do it over - not fix it.  By the time you know
it's wrong at least 50 people will have decided it's a feature!! (and of
those 50 at least 10 will work for IBM's biggest and most influential
customers!).  Trust me - been there - done that - got the APARs!!!

Jon Paris
Partner400