I agree. Heaps of experimentation and first-class work were lost because they happened before open source became mainstream. Especially in industry, where licensing models sealed work off from community adoption.
Open source was basically the default in the beginning, then declined in the 80s and 90s which prompted FSF and GNU to emerge, finally picking back up quickly in the 21st century.
This sounds like a massive case of post hoc ergo propter hoc.
> Open source was basically the default in the beginning,
Prior to the rise of the GPL and other formal F/OSS licensing models, there was some explicitly-dedicated-to-the-public-domain software, and lots of lax enforcement of copyrights in software, and, especially prior to 1974, some doubt about the copyright status of software. Before automatic copyright in 1978, and especially before 1974, lots of software was probably not copyrighted (because it took an active step to do so and because there was doubt about whether doing so would have any legal effect.)
But I don't think open source was ever the norm for software once it was clear that copyright protection was available, and certainly not once it became automatic.
