http://www.redbooks.ibm.com/redbooks/pdfs/sg247000.pdf page 45 : « Few people know that open source was the business model that software began with! In the 1960s, nobody would buy a computer (a huge investment at that time) that was not immediately ready for some use. Software had to be given away by manufacturers as “a way to sell the hardware faster,” and free of charge for that reason. The source code was distributed so that anybody could change it. At the time, nobody would or could use a computer without having programming skills.
In the 1970s, legislation in many countries decided that giving away software for free was unfair competition toward new companies that were developing competitive software and selling it for a living. Therefore, software had to be billed by law. It was still lawful to give the source for free as an option to anybody who bought a software product, and IBM did just this.
In the late 1970s, centralized support to end-users began to get difficult because they did not know whether they used “vanilla” software (object code from IBM) or software modified by their IT group. Since only very few people ordered the source, but were causing most of the work, IBM had to adopt an object-code-only (OCO) distribution mode. This may have been one of the reasons universities began to switch from VM/CMS to UNIX, because UNIX code was available to play with. »
