Subject: Re: the old days From: Rudd Canaday Date: Thu, 27 Oct 2011 08:40:06 -0700 To: Kristaps Dzonsons CC: Tom Van Vleck Tom, Feel free to use anything from our correspondence. Also, feel free to use my name in connection with it. I don't remember what Bob Morris' involvement was in Multics. I don't think he was involved with UNIX at all until after I left. So far as I know, his UNIX involvement, if any, was peripheral. Bob may have ported runoff to the Murray Hill 635 (actually a 645, I think, by that time). So far as I know he had no role at all in the NIKE TSS or the Whippany 635. I don't think any of the Murray Hill people had any involvement with NIKE TSS or the Whippany 635 (except me, of course). I've said that runoff may have been ported to the Whippany 635 as part of the NIKE project. I do not specifically remember this and it may very well never have happened. If it did happen, I believe it was done by the NIKE TSS team at Whippany (the GE contractors), not by Bob Morris. I believe that it is a coincidence that the contractors who worked on NIKE TSS were from GE. I don't think there was any connection with the fact that GE was part of the MULTICS effort. There were a number of contractors from GE working on the NIKE ABM system. Unfortunately, I have no documentation or code from the NIKE TSS. It probably exists in the NIKE archives at Bell Labs. We worked with the code on punched cards and I don't remember it ever existing on disk. Probably it did exist on tape. As a side note, we began the NIKE TSS effort before the 635 was installed at Whippany. So, for the first couple of months, every night two team members would take all the punched cards to Boston at 5PM and run them on the 635 at MIT. We'd return on the redeye so the results of the run were available to the team by about 9AM the next morning. It was fun. - Rudd On Oct 26, 2011, at 3:34 AM, Kristaps Dzonsons wrote: > Tom, Rudd, > > Thanks for your comments! This fills in a previously-unknown port of > Saltzer's code. Before beginning, would either of you object if I > include relevant correspondence in the record? (Of course, I omit all > mail headers but date, subject, and email-obscured participants.) > > I have a few questions, which I'll ask as-is instead of in-line. > > A Big Pre-Unix Mystery is of the GE-635 and the involvement of Bob > Morris, who's mentioned inconsistently throughout the secondary > sources but is completely absent from the primary. > > Rumour has it that he ported RUNOFF to a GE-635 (which one is > unknown). McIlroy estimates this effort was in '67 or '68, as he knew > of Morris' port, but nothing more than its existence. If Saltzer > published RUNOFF in '64 and you moved on to Multics in '66, your > GE-635 Nike TSS RUNOFF would fall between. McIlroy reimplemented > runoff with BCPL in '69. Can you fill in the gaps, here? Was Morris' > 635 port in fact your own? > > Lastly, do you know of a source-code trail for the GE-635 Nike TSS > port? > > Thanks again, > > Kristaps