ITER and DO Loops (Was: %EOF, %FOUND and the infamous Dou loop)

Actually, any loop has the potential to become endless if an ITER is
placed before the code that sets the condition being tested by the DO.  As
these contrived examples show . . .

C               EVAL    y = 10
C               EVAL    x = 0
C               DOU     (x > y)
C               ITER
C               EVAL    x = x + 1
C               ENDDO

C               EVAL    y = 10
C               EVAL    x = 0
C               DOW     (x < y)
C               ITER
C               EVAL    x = x + 1
C               ENDDO

neither of these loops will end.  Real code will never have an
unconditional ITER inside a loop but it is easy to imagine a maintenance
programmer adding a conditional ITER when some condition occurs that
results in x never being incremented.  Oops!

ITER is just a disguised GOTO and thus should be carefully considered
before being discarded!!  ITER is useful in many cases but then GOTO can
also be useful.  The current wisdom suggests that GOTO be avoided.  I
suggest that the impact of any branch operation must be carefully
considered.  These sorts of bugs (endless loops) are frequently caused by
ill-considered quick fixes.

And please! I don't want a heap of appends showing how ITER and LEAVE and
LEAVESR etc. are contributing to cleaner code.  It is far too early in the
year for that sort of drivel, and it just ends up being a series of
opinion pieces and the only opinion that matters is mine :)

Regards,
Simon Coulter.

--------------------------------------------------------------------
   FlyByNight Software         AS/400 Technical Specialists
   http://www.flybynight.com.au/

   Phone: +61 3 9419 0175   Mobile: +61 0411 091 400        /"\
   Fax:   +61 3 9419 0175   mailto: shc@flybynight.com.au   \ /
                                                             X
                 ASCII Ribbon campaign against HTML E-Mail  / \
--------------------------------------------------------------------