@kagan @jascha As a historian of science I can state that this claim hasn’t stood up since at least the 1950s and that the introduction of NDAs into the grant process about 20 years ago has made hiding the results more common than we might expect.
The most glaring recent example being the climate scientists at Shell who never disclosed their research but most commonly it happens in medical research.

@petealexharris Vendor lock in also dates to the late 50s. “copyleft” licensing only became possible when copyright law changed in the early 80s to allow software copyright, but even it’s 40 years old.
Interestingly, vendors actually encouraged sharing free software, through user groups like IBM Share and DEC DECUS, before the law changed.
The business software market has always been lucrative.

@petealexharris Not sure what you mean by “appearance” but free software has been around since the late 1950s, when compilers were first passed around.