Information Technology Reference
In-Depth Information
The heterogeneity of software, its inherent messiness, and permeabil-
ity, are everywhere apparent in the historical documents. Compared to
the history of computer hardware, which is characterized by regular and
remarkable progress, the history of software is replete with tension,
confl ict, failure, and disillusionment. The fi rst commercial computers
had been out for only a few years when the availability of useful and
reliable software was identifi ed as one of the critical bottlenecks hinder-
ing the expansion of the industry.
20
Unlike computer hardware, which
was constantly becoming smaller, faster, and cheaper, software always
seemed to be getting more expensive and less reliable. By the early 1960s
industry observers and corporate managers increasingly warned against
a growing “software gap” as well as a sense of “frustration,” “disen-
chantment,” and “disillusionment” with electronic computing provoked
by problems associated with the rising costs of software development.
21
By the end of the decade many were talking openly of a looming soft-
ware crisis that threatened the health and future of the entire com-
mercial computer industry. For the next several decades, corporate
managers, academic computer scientists, and government offi cials would
release ominous warnings about the desperate state of the software
industry with almost ritualistic regularity.
22
In fact, what is most striking
about much of the literature from the supposed Golden Age of the com-
puter revolution is how contentious it is, how fraught with anger and
anxiety. In an industry characterized by rapid change and innovation,
the rhetoric of the software crisis has proven remarkably persistent. The
Y2K crisis, the H1-B visa debates, and recent concerns about the loss
of programming jobs to India and Pakistan are only the most recent
manifestations of the industry's apparent predilection for apocalyptic
rhetoric.
To many observers of the computer industry, reconciling the two
dominant but opposing views of the history of computing—the glorious
history of computer hardware and the dismal history of computer soft-
ware—often has been diffi cult, if not impossible. The seeming paradox
between the inevitable progress promised by Moore's Law and the per-
petual crisis in software production challenges conventional assumptions
about the progressive nature of computer technology. This is perhaps
the most signifi cant lessons to be learned from the history of software:
There is no Moore's Law for software technology
. But the real problem
with software is not so much that it is “hard” (as computer scientist
Donald Knuth famously declared) but rather that it is inherently con-
tested; the problem was generally not that the software itself did
