Plus ça change...

One of the explanations frequently given for the emergence and success of the free software movement is that it is simply one particular expression of a pre-existing tendency to share code.

In an effort to get started on an embryonic project concerning the social history of The User, I just finished reading Atsushi Akera's meticulously researched 2001 paper, "Voluntarism and the Fruits of Collaboration: The IBM User Group, Share." (As far as I can tell, it's only available online if you have access to a Project MUSE subscription.) While Akera seems primarily interested in the roots of professionalism in the computing industry, I was struck by a number of his observations that speak to FOSS.

Share was founded as a group of corporate users of the IBM 704 mainframe, generally composed on the individual level by directors of computing installations. The group did ground-breaking work to develop notions such as "systems programming" and "operating systems," -- Akera notes that

during this period the terms themselves were being defined in specific relation to the emerging labor structure of computing installations

-- and developed a compiler for a mathematics-intensive language, PACT, that ended up being eclipsed by FORTRAN.

Among the interesting moments in the paper:

[The focus on systems programming] followed from the computer center directors' commitment to operating efficiency. . . . As if to reinforce this orientation, Share forbade the exchange of applications programs. . . . Share used this distinction to mark the boundaries of proprietary interest.

which certainly foreshadows the logic of differentiating and non-differentiating technology deployed by Bruce Perens in his First Monday piece on "The emerging economic paradigm of Open Source." Also,

Share was established with an eye towards productive collaboration. . . . The unsatisfactory quality of IBM's programming tools and particular dispersion of knowledge provided a technical rational for collaboration.

which is a (probably unnecessary) bit of support for Raymond's principle about programmers and their itches.

One of the most interesting points Akera makes has to do with corporate voluntarism and anti-trust law:

[T]he group arguably transgressed a legal boundary by coordinating development work. By asking each installation to take on specific "assignments," Share entered a gray area that could have brought on costly litigation. Share representatives therefore began to stress the voluntary character of the work . . . . the rhetoric of voluntarism began to pervade Share's letters and proceedings.

So one implicit function of FOSS projects is in fact to provide legal cover for corporate contributors.

Finally, a note about the early connections between collaboration and reputation:

[S]ince all of Share's programs were attributed to their authors, the very distribution of programs served as a mechanism for credit. During the mid-1950s, when programmers could lay claim to no professional society, Share provided many programmers with the single most important forum in which to develop and demonstrate their skills. . . . The group also made it easier to identify . . . talent by establishing common measures of skills and abilities.

Certainly an article worth seeking out.