Materiel Command and the Materiality of Commands: An Historical Examination of the US Air Force, Control Data Corporation, and the Advanced Logistics System - History of Computing: Learning from the Past - page 94

competing against CDC for ALS [20]. 15 CDC was awarded the initial computer and
software contract for $87.4 million in April 1972 [21]. This was one of the very larg-
est government computer contracts to that time, but well justified to AFLC leaders—
they believed ALS would result in savings of $250 million, through automation and
greater efficiencies, in the succeeding decade [22]. It was also, by far, CDC's largest
contract to date [23].
The Control Data contract called for 21 CDC Cyber 70 computers, three each at
the AMAs, with the remaining six at the Logistics Command headquarters at Wright-
Patterson. The agreement specified a complex monthly lease with right-to-purchase
structure for the stipulated seven-year life of ALS [21]. Approximately 700 AFLC
personnel were working on ALS by this time at Wright-Patterson and about 400 at
other AMAs, and these numbers escalated rapidly in succeeding years. The Air Force
would develop the central control system and applications software, while CDC
would contribute a transaction-based operating system named ZODIAC. The central
control system, written in COBOL, was an essential piece of system software standing
between ZODIAC and any applications software [24].
The CDC Special Systems Division ALS team, managed by Laccabue, was over-
seen by a high level ALS Program Management Board at CDC that included Robert
Price, President, Special Systems and Services Division, and other senior CDC execu-
tives [25]. Laccabue joined CDC as a programmer analyst in 1960 after a short stint
as a dynamics engineer at Convair. He had participated in the preparation, bidding,
and early phases of a CDC Special Systems Division project for the Worldwide Mili-
tary Command and Control System. This involved discussion and preliminary
research on transaction oriented operating systems. His experience, on the promising
but young technology of transaction oriented operating systems, made Laccabue a
strong choice for CDC's ALS project manager position [18].
Given CDC leaders' uneasiness with the central control system software being
programmed in COBOL and with some other Air Force specifications, they carefully
crafted their proposal and the lease-to-purchase contractual terms. Purchase credits
were greater if the Air Force executed a purchase at an earlier date. In effect, if the
Air Force delayed in purchasing the CDC Cyber 70 computers (beyond the initial
target date), it would end up paying significantly more money for the systems.
Control Data contract vice president H. D. Clover was the mastermind behind this
structure that protected CDC if the AFLC could not deploy the system successfully,
and on schedule. As Laccabue later recalled, the GAO subsequently prohibited gov-
ernment contracts being structured this way. Laccabue emphasized that this was not to
pull one over or be unfair to the Air Force, which CDC hoped to do future business
with; it was merely a protective mechanism given what Control Data leaders saw as a
risky project and CDC's dependence on the Air Force for project success [18].
Unlike the initial harsh response CDC received at Hanscomb field at the pre-
contract phase, some top AFLC officials subsequently understood CDC's concerns.
However, they thought these challenges could be readily overcome by AFLC data
processing personnel. The Air Force had thousands of programmers and initially the
focus was on central processing unit (CPU) cycle time: beating benchmarks. The Air
Force was far from alone in the late 1960s and early 1970s in underestimating the
15 RCA and Burroughs did not attempt second round benchmark testing.