Research Unix V7's (comparatively) long time gap from V6
Today for reasons outside the scope of this entry I found myself looking at the release dates for the various editions (versions) of Research Unix up through the pivotal V7 release. This made something about the timeline jump out at me.
Starting from the 1st Edition onward through V6, the Bell Labs CSRC set a blistering pace. V1 is dated November 3rd 1971, and was followed by V2 dated June 12, 1972, V3 dated February 1973, V4 dated November 1973 (where the kernel was written in C for the first time), V5 dated June 1974, and then V6 in May of 1975, the longest time gap up until that point but still less than a year from V5. Then the CSRC didn't pause to put together a formal 'edition' that they released until the 7th Edition (V7) in January 1979, more than three years later. This doesn't mean that the Bell Labs CSRC was idle during that time; V7 has an impressive list of developments, so they were clearly busy evolving Research Unix. They just didn't pause to make and publish their work as an edition, for whatever reasons.
(However, I believe that CSRC published about what would later be released as V7 in 1978, in what became a famous issue of the Bell System Technical Journal, Vol. 57, No. 6, Pt. 2 Jul/Aug 1978 (in scanned PDF form).)
Learning this helps illuminate to me why a relatively large number of early Unixes were partially based on V6, such as PWB/UNIX and CB UNIX (an ingredient into System III). Not only was V6 what you had to work with between 1975 and 1979, but that was enough time that people could accumulate all sorts of changes and pass them on to others. And V6 was the current (and only) externally available Unix release for long enough that people could produce things like A Commentary on the UNIX Operating System (which was first assembled a year after V6's release and then printed as a book another year later).
Although I could speculate about reasons why there was such a gap between the publication of V6 and V7, it would be just that, speculation, so I'm going to skip it. If there's an authoritative account about why, I couldn't find it with some casual Internet searches. Although some of the famous Bell Labs people are sadly deceased (cf), enough remain alive that they could be asked, if they haven't already written about it (which I suspect they have, somewhere).
Sidebar: A brief history of Research Unix licensing
According to Wikipedia, V5, V6, and V7 were all licensed to educational institutions (for little to no cost as I remember it). V6 and I believe V7 were licensed to commercial users as well. I believe that V5 was the first version to be talked about much outside of Bell Labs, and certainly it's said to be the first version released to outside people (only in educational institutions, though). Anyone who started with V5 might well have replaced it shortly afterward when V6 came out, assuming that hearing about it, getting together interested people at your institution, finding or funding a machine, and so on didn't take long enough that V6 was out by the time you were ready.