Predictably I suppose, recompilation gets my vote.  The issues
involved are technical and not management ones, and bureaucratizing
them never helps.

Development takes some time, and linking the development version of a
PL/I compiler to that in current production use is always a bad  idea.
 It ensures that retrograde technology and performance will be wired
into newly developed systems.  (This may happen anyway, of course; the
use of the best  translator is a necessary but not a sufficient
condition for high performance.  That use can be, often is,
perfunctory.)

I am also suspicious of Jan Vanbrabant's esclusion of homologation
from this discussion.  The word is derived from the ancient Greek verb
homologein, to approve, which becomes homologare, to agree, in fairly
late Latin.  (It has a special meaning in Scots law, where it is used
to characterize a process of removing minor defects from contracts,
the remediated versions of which are then given the force of law.)

If, as I suspect, homologation here has to do with ensuring that a
systems meets its functional specifications, it is relevant.

John Gilmore, Ashland, MA 01721 - USA

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to