On Aug 18, 2012, at 11:36 AM, John Gilmore wrote:
IBM had the considerable advantage, circa 1964, of having had
substantial experience with the inadequacies of the fix-handling
machinery it had put together for IBSYS.
It came to terms early with the perhaps regrettable but ineluctable
frequency of errors in tested code. Other vendors persisted for long
in the notion that they could somehow eliminate all errors before they
distributed code.
Moreover, IBM (a few unsuccessful excursions aside) has always dealt
with professional programmers rather than, say, the end users of
Windows. An analogue of SMP/E would not, I suspect, serve Microsoft
well in its dealings with most of its end users.
John Gilmore, Ashland, MA 01721 - USA
John,
IBM has gone the MS route with the OE component (IMO) they no longer
ship the csect in question, they ship the entire module.
which imo makes maintenance (just the receive part a real PITA).
There maybe other issues that IBM isn't talking about (I wish they
would at least acknowledge the issue and say they are working on it).
I just find it disconcerting that to fix one issue in say JAVA they
ship essentially a new version (replacement module).
I am agreement that an end user would be baffled by SMPE constructs.
But I think they(MS) could hide most of it. Of course I am sure one
of the bright guys would find out about it and complain bitterly
about it. Its too far into the distribution to change (but it would
be better in my opinion) than the way its done now.
Ed
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
