On Mon, Jan 25, 2016 at 8:38 AM, Lawrence Stewart <stewart@serissa.com> wrote:

We had one at the Stanford Information Systems Lab sometime in 78 or 79, running (I think) V7, and we certainly didn’t do the port ourselves! We did put it on the Arpanet though, as SU-ISL. This
was a pretty weird hookup. The NCP ran on a front-end LSI-11 (or was it an 11/23?) and there was
a Very Distant Host interface home-built by Ron Crane that ran over a copper pair to the IMP at the medical school. I did the driver work to connect the 11/34 to the smaller 11 running the NCP.

If it was 78, it was probably v6+ of some sorts running Chesson's Arpanet NCP from Illinois. UNIX/TS (aka V6+++ / pre V7) sorts of oozes out via the Bell Labs' OYOC like Ted in '78 - that's what we ran at CMU since Ted brought it with him. Its a heavily hacked V6 kernel and many of what would become the v7 utilities including the a new compiler and the standard I/O library. Same was true of PWB 1.0 - which was based on most of the code. Dennis would not formally get V7 (which had an updated kernel) released until mid '79 (FWIW: The date on a number of the files in the V7 distribution tapes in Warren's archives show Aug 1, '79 - which sounds about right for when Dennis got it out).