hancock4@bbs.cpcn.com writes:
> I believe IBM always made the source code available for its mainframe
> operating systems. Competitors could and would use it for supplemental
> utility programs. They would write links and exits to/from the
> operating system for maximum program efficiency.
Until June 1969 (a date known in the IBM mainframe community as "New
World") IBM (with a very few exceptions) didn't even copyright its
software, and did not charge for it. The price was bundled into the
charges for IBM hardware. That's why you can find the source for
pre-New World MVS and VM on the Internet, and run them in the Hercules
S/370 emulator on a PC.
After New World, the combined pressure of the IBM mainframe clone
manufacturers and the Justice Department antitrust lawsuit gave IBM
the opening to unbundle software and begin charging what were then
extremely high prices. (On the day of the New World announcement IBM
released four "program products". A headline in a subsequent issue of
_Computerworld_ read "SURPRISE! Software costs as much as a
printer!". The reference was to one of the four program products,
Generalized Information System (GIS), which had a monthly charge
(running forever) of ~US$1200 (in 1969 dollars!), which was about the
same as the monthly rental fee for a 1403-N1 1100 lpm printer.
Don't take the above price as exact; it's been 36 years ... <grin>
However ... even after New World many of the program products still
offered an option for the customer to obtain the (copyrighted) source
code. A few years later, however, the PHB contingent at IBM decided
that it was a Bad Thing to allow mere customers to see the source
code, with the result that IBM implemented the Object Code Only (OCO)
policy.
IBM insisted that there was no need for customers to see the source or
*gasp* modify it to meet their organization's requirements because IBM
was providing defined interfaces that gave customers all they needed.
(Does this sound like the attitude of a certain software vendor in
Redmond?)
One other consequence of the OCO policy was that the customers could
no longer debug the problems that were encountered when using the IBM
products. One industry observer (Melinda Varian, I think, but I'm not
sure and I've not talked to Melinda in many years) commented that with
the OCO policy IBM had fired its most productive systems support
staff: the unpaid (by IBM) customers.
Joe Morris
Joe Morris (jcmorris@mitre.org)