Dan,
I’m not sure how well I can answer, but I’ll try.
I think the responsible organization was known as ACIS (ACademic
Information Systems??), with BSD work primarily staffed in the Palo Alto
Scientific Center. Back then IBM had “scientific centers” separate from the
Research Division – my main contact was with Cambridge, MA and Palo Alto
centers, but I think there were 13 centers world wide.
I think you’re correct that it was called AOS and likely correct that it
was only available to academic institutions. I was primarily aware of usage at
Brown, CMU and MIT.
The sole RT I had with AOS was at my house, and I didn’t have Internet
access at the office, much less at home, so though I was in the thick of the
distributed filesystem work, didn’t do any of it hands on on AOS.
I was technically responsible for getting the NFS license for AIX and I
think you’re right that ACIS put NFS on AOS once we had the license, but would
have forgotten that if you hadn’t reminded me. I think the license agreement was
put in place in the second half of 1988, but that could easily be off by months
or longer.
Certainly much of the Andrew work in general and AFS specifically was done
with AOS. I assume that all AFS versions were made available on AOS as soon as
Kazar et al thought they were ready.
Charlie
Sent: Thursday, March 09, 2017 9:27 PM
Subject: Re: [TUHS] RT/PC-centric AIX history
Wow, this is really cool, Charlie. It puts a lot of stuff in
perspective.
I wonder if you might add a bit more detail about the BSD ports? That's
what we ran on our RTs; I seem to recall that product was only available to
educational institutions and was referred to as AOS: "Academic Operating
System." I do recall that it came with NFS, and possibly AFS version 2? It
seemed to be approximately 4.3-Tahoe based. The AFS bit is hazy....