Historical note: If I'm not mistaken: Bruce Perens put freedom-to-use (non-discrimination against persons, groups, or fields of endeavor) in the DFSG (1997-07-05) *before* RMS added freedom 0 to the FSD (some time between 1999-02-24 and 1999-04-30). #FreeSoftware
it sounds like the sort of thing that would be mentioned in the GNU Bulletin. I don't have recordings nor perfect memory, but I'm pretty sure that in the first speech by RMS that I had occasion to attend, in 1996, freedom 0 was already old news, as in well established as a requirement
since the goal of the DFSG was not to make up new requirements, but to codify the allegedly too-abstract freedoms into concrete, objective requirements, it seems weird to as much as hypothesize that they'd make a requirement that didn't reflect what was already present in the FSD: it would detract from the stated goal
I suppose grepping GNU and Debian mailing list archives from early 1990s for 'four freedoms' will get plenty of hits. I started getting involved with GNU in 1993, and reading a lot about it since September 1991, and the phrase 'three freedoms' doesn't ring any bells. it feels to me like it was *always* 'four freedoms', since I first became acquainted with them, but I recall reading and hearing about how the freedom to run for any purpose was added after the other three
fantastic! thanks for the data points. interestingly, the very opening paragraph mentioned "rights to use, copy, modify, redistribute", so it was arguably already there, but was absent in issue 1, so it was considered and added between them. it was not presented as a freedom nor permission that one gets along with a free program, though, which is accurate per copyright laws
Issue 1 (Feb 1986) through Issue 20 (Jan 1996) have a definition with just "two specific freedoms". In Issue 21 (July 1996) that became "three specific freedoms". It remained 3 freedoms through the final issue, Issue 24 (Mar 1998).
that it spoke of 'three freedoms', and earlier versions spoke of 'two freedoms', sort of proves that my memory was faulty. not that I needed further proof of that ;-) sorry about the misleading and incorrect recollections
@lxo And I'm sorry about misleading that Perens wrote about right to use before RMS did. At the time I was just stupidly over-confident, but glad to know that Cunningham's law still works :)