[TUHS] What was the first edition of UNIX that left AT&T

[TUHS] What was the first edition of UNIX that left AT&T

[Clem Cole]

I was recently asked a question that I was not sure the answer, so I
thought I would pass it to this group of folks.

What was the first edition that actually left Murray Hill and where did it
go?

My own first encounter was the Fifth edition but I know that was later.
I'm pretty sure both Harvard and MIT had it before CMU.   I'm thinking
Fourth edition went to Harvard and some other places ??NYU?? ??MIT??   Did
anything earlier than Fifth ever leave Murray Hill?

I don't think UCB or UofI got it until the Sixth edition.   I believe there
was some earlier commercial site ??CU in NYC maybe?? may have been in there
also but I have no idea what version that was.

Doug do you know?

Warren any ideas?

Tx
Clem
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170831/ca739760/attachment.html>

[TUHS] What was the first edition of UNIX that left AT&T

[Theo Pavlidis]

Unix arrived at Princeton in the early 1970's and it might have been the
third or fourth edition. The first time I heard about Unix was from one of
my graduate students who was just starting work on his thesis, so it must
have been before 1974. (He got his degree in May 1975 and this the only
date I am sure about ;-)...)

Princeton is geographically close to Murray Hill (30 miles or so) and there
was a lot of interaction between the two. I left Princeton in 1980 to join
1127 and another Princeton colleague had moved a year earlier.

Theo Pavlidis


On Thu, Aug 31, 2017 at 6:33 PM, Clem Cole <clemc at ccc.com> wrote:

> I was recently asked a question that I was not sure the answer, so I
> thought I would pass it to this group of folks.
>
> What was the first edition that actually left Murray Hill and where did it
> go?
>
> My own first encounter was the Fifth edition but I know that was later.
> I'm pretty sure both Harvard and MIT had it before CMU.   I'm thinking
> Fourth edition went to Harvard and some other places ??NYU?? ??MIT??   Did
> anything earlier than Fifth ever leave Murray Hill?
>
> I don't think UCB or UofI got it until the Sixth edition.   I believe
> there was some earlier commercial site ??CU in NYC maybe?? may have been in
> there also but I have no idea what version that was.
>
> Doug do you know?
>
> Warren any ideas?
>
> Tx
> Clem
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170831/1b7a1f4f/attachment-0001.html>

[TUHS] What was the first edition of UNIX that left AT&T

[Warren Toomey]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 936 bytes --]

On Thu, Aug 31, 2017 at 06:33:04PM -0400, Clem Cole wrote:
>   I was recently asked a question that I was not sure the answer, so I
>   thought I would pass it to this group of folks.
>   What was the first edition that actually left Murray Hill and where did
>   it go?

"A Quarter Century of UNIX" says:

... after the first paper was delivered in October 1973, you could
have the bits put on your RK05 disks. In the early spring of 1974, Lou
Katz (then at Columbia University) organized a meeting of Unix users.
Columbia had been the recipient of the first distribution—first on
disk, then on 9-track tape—in the autumn (”Cy got RK05s for the department,”
Katz told me, ”but we didn’t have a drive, so I drove down
to Murray Hill and Ken cut me a 9-track tape") The meeting was held
on May 15, 1974, in the Merritt Conference Room on the third floor
of Columbia's College of Physicians and Surgeons ...

Cheers, Warren

[TUHS] What was the first edition of UNIX that left AT&T

[Jeremy C. Reed]

> "A Quarter Century of UNIX" says:

And in Salus later book, The Daemon, the Gnu, and the Penguin (which I 
published for him), he quotes Allman as using 4th edition at Berkeley. 
Allman also told me (for my BSD history book in progress), that when he 
got involved INGRES was running on 5th Edition and he helped with the 
transition to 6th Edition.

Also McKusick's 1985 article "A BERKELEY ODYSSEY: Ten years of BSD 
history" says the version 4 tape was delivered in Jan. 1974 (and used on 
PDP 11/45) and INGRES was running the newly-available Version 5 of Unix 
in the spring of 1974 (on PDP 11/40).

(Following from my book...)

The University of California --- via the 
San Francisco Medical School Campus --- was recently
licensed by Western Electric to use Unix. (Fortunately, the
licenses from Western Electric were per-organization rather than
per-computer.)
So Fabry was able to get started with Unix much more quickly than anyone
would have imagined.\cite{fabry2}
%
% NOTE: here is the license for it: archives/1970s/UC_License_4thEd.pdf
% effective Dec. 1, 1973, licensee was the Regents of the Univ.
% NOTE: it says for the location of School of Medicine
% signed on Jan. 7 and 16, 1974
% 2.01
The license allowed use solely for academic and educational purposes.
% 4.05
In addition, the University was prohibited from sharing the software
--- including its methods or concepts --- to anyone other than the
University employees or students.
% NOTE: no version of Unix is mentioned in the license

[TUHS] What was the first edition of UNIX that left AT&T

[Clem Cole]

Thanks - I've already sent Lou a note, which he has not yet responded.  I
should have remembered he was user #1.

So, my guess is that Lou had 4th edition at this point.

Although from your comments, I'm a little surprised Berkeley went back that
far, give than BSD (1.0) was all sixth edition and post Ken's sabbatical.
The Industrial Liaisons Office (ILO) in EECS which ran all of that had
existed for a number of years (since the late 1960s).  They were already
distributing things like SPICE and SPLICE.  So the idea of giving away
there SW work was really well ingrained in the Berkeley way of doing things
in EECS.

BSD (1.0) in fact used the original SPICE distribution process in the ILO
(as I suspect Ingress would have also a few years later).   While folks
think of the later CSRG stuff, that was later - once UCB got the support
contract from DARPA.  Originally it was just the fruits of the labors of
the folks in EECS being package and distributed to the folk what worked
with them as managed by the ILO (which was originally folks like AT&T, IBM,
HP, Tektronix, DEC, Intel, Fairchild, *etc*...)

So it means that UCB was hacking privately without taking to Katz@ NYU, or
the Columbia and Harvard folks for a while.   I need to ask Lou what he
remembers.   UCB was not connected to the Arpanet at this point (Stanford
was), so it's possible Ken's sabbatical openned up some channels that had
not existed.   [UCB does not get connected until ing70 gets the
vdh-interface up the hill to LBL's IMP as part of the Ingress project and
that was very late in the 70s  - not long before I arrived].

On Fri, Sep 1, 2017 at 3:05 PM, Jeremy C. Reed <reed at reedmedia.net> wrote:

> > "A Quarter Century of UNIX" says:
>
> And in Salus later book, The Daemon, the Gnu, and the Penguin (which I
> published for him), he quotes Allman as using 4th edition at Berkeley.
> Allman also told me (for my BSD history book in progress), that when he
> got involved INGRES was running on 5th Edition and he helped with the
> transition to 6th Edition.
>
> Also McKusick's 1985 article "A BERKELEY ODYSSEY: Ten years of BSD
> history" says the version 4 tape was delivered in Jan. 1974 (and used on
> PDP 11/45) and INGRES was running the newly-available Version 5 of Unix
> in the spring of 1974 (on PDP 11/40).
>
> (Following from my book...)
>
> The University of California --- via the
> San Francisco Medical School Campus --- was recently
> licensed by Western Electric to use Unix. (Fortunately, the
> licenses from Western Electric were per-organization rather than
> per-computer.)
> So Fabry was able to get started with Unix much more quickly than anyone
> would have imagined.\cite{fabry2}
> %
> % NOTE: here is the license for it: archives/1970s/UC_License_4thEd.pdf
> % effective Dec. 1, 1973, licensee was the Regents of the Univ.
> % NOTE: it says for the location of School of Medicine
> % signed on Jan. 7 and 16, 1974
> % 2.01
> The license allowed use solely for academic and educational purposes.
> % 4.05
> In addition, the University was prohibited from sharing the software
> --- including its methods or concepts --- to anyone other than the
> University employees or students.
> % NOTE: no version of Unix is mentioned in the license
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/tuhs/attachments/20170901/3c134750/attachment.html>

[TUHS] What was the first edition of UNIX that left AT&T

[Jeremy C. Reed]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1448 bytes --]

On Fri, 1 Sep 2017, Clem Cole wrote:

> So it means that UCB was hacking privately without taking to Katz@ NYU, or
> the Columbia and Harvard folks for a while.   I need to ask Lou what he
> remembers.   UCB was not connected to the Arpanet at this point (Stanford
> was), so it's possible Ken's sabbatical openned up some channels that had
> not existed.   [UCB does not get connected until ing70 gets the
> vdh-interface up the hill to LBL's IMP as part of the Ingress project and
> that was very late in the 70s  - not long before I arrived]. 

Allman told me that Mike O'Malley had an ARPA connection at UCB that was 
axed a few years before the INGRES link. So yes, I think no Arpanet 
connection during the early BSD development work. (Losing this 
connection may have had some controversy, but I don't know the details.)

Fabry told me that O'Malley used Unix for his (EECS) Artificial 
Intelligence research projects before he discovered it (so before the 
October 1973 Symposium).

RFC 402 of Oct 1972 has a ARPA network participant Michael O'Malley of 
University of Michigan Phonetics Laboratory. Also this draft report at 
http://digitalcollections.library.cmu.edu/awweb/awarchive?type=file&item=355330 
about the ARPA speech recognition project lists M. H. O'Malley at UCB 
and says the principle investigator from Univ. of Michigan moved to UCB.  
(I never go ahold of him to see if had any other relevance to my BSD 
story.)

[TUHS] What was the first edition of UNIX that left AT&T

[Clem cole]

I did hear back from Lou Katz - user #1.  Indeed the first version that escaped the labs was the 4th edition. 

Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite. 

> On Sep 1, 2017, at 7:16 PM, Clem cole <clemc at ccc.com> wrote:
> 
> Interesting. If O'Malley had a connection wonder what it was connected too on both sides.  It had to be to lbl but the Vdh was a piece of shit even in the ingvax days.  The first version was even worse.  On the ucb side I wonder.  It would not have been Unix because UofI did the original arpanet code and that was for v6.  There was never a pdp10 at ucb so I wonder if it was one of the CDC machines which were the primary systems until Unix came to be.  
> 
> Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite. 
> 
>>> On Sep 1, 2017, at 5:59 PM, Jeremy C. Reed <reed at reedmedia.net> wrote:
>>> 
>>> On Fri, 1 Sep 2017, Clem Cole wrote:
>>> 
>>> So it means that UCB was hacking privately without taking to Katz@ NYU, or
>>> the Columbia and Harvard folks for a while.   I need to ask Lou what he
>>> remembers.   UCB was not connected to the Arpanet at this point (Stanford
>>> was), so it's possible Ken's sabbatical openned up some channels that had
>>> not existed.   [UCB does not get connected until ing70 gets the
>>> vdh-interface up the hill to LBL's IMP as part of the Ingress project and
>>> that was very late in the 70s  - not long before I arrived]. 
>> 
>> Allman told me that Mike O'Malley had an ARPA connection at UCB that was 
>> axed a few years before the INGRES link. So yes, I think no Arpanet 
>> connection during the early BSD development work. (Losing this 
>> connection may have had some controversy, but I don't know the details.)
>> 
>> Fabry told me that O'Malley used Unix for his (EECS) Artificial 
>> Intelligence research projects before he discovered it (so before the 
>> October 1973 Symposium).
>> 
>> RFC 402 of Oct 1972 has a ARPA network participant Michael O'Malley of 
>> University of Michigan Phonetics Laboratory. Also this draft report at 
>> http://digitalcollections.library.cmu.edu/awweb/awarchive?type=file&item=355330 
>> about the ARPA speech recognition project lists M. H. O'Malley at UCB 
>> and says the principle investigator from Univ. of Michigan moved to UCB.  
>> (I never go ahold of him to see if had any other relevance to my BSD 
>> story.)
>> 
>> 

[TUHS] What was the first edition of UNIX that left AT&T

[Jeremy C. Reed]

> > On Sep 1, 2017, at 7:16 PM, Clem cole <clemc at ccc.com> wrote:
> > 
> > Interesting. If O'Malley had a connection wonder what it was 
> > connected too on both sides.  It had to be to lbl but the Vdh was a 
> > piece of shit even in the ingvax days.  The first version was even 
> > worse.  On the ucb side I wonder.  It would not have been Unix 
> > because UofI did the original arpanet code and that was for v6.  
> > There was never a pdp10 at ucb so I wonder if it was one of the CDC 
> > machines which were the primary systems until Unix came to be.

Regarding the ARPA network at UCB, see ARPANET Completion Report at 
http://walden-family.com/bbn/arpanet-completion-report.pdf PDF page 199 
"ARPA NETWORK, LOGICAL MAP, SEPTEMBER 1973" shows LBL with 316-based 
IMP. The next page (June 1974) shows it (LBL IMP) linked with UCB with a 
PDP-11 "v".

I don't know the key for "v" but maybe means "very distant host" as 
defined in Report 1822 Appendix F (PDF page 151 at 
http://walden-family.com/impcode/BBN1822_Jan1976.pdf ), If that is what 
it is, the BBN 1822 Section F.3.4 suggests "[e]ither an EIA RS232C 
interface or the special Bell 303 [modem] interface can be used. Speeds 
up to 230.4 kilobits/second are permitted. ... At the Host site there 
will be a mating full-duplex modem."

"UCB" was removed from the Oct. 1976 map, but it still shows a PDP-11 
"v" link. And the remote link removed from map in March 1977.

Allman told me the INGRES link was a "VDH (Very Distant Host) interface, 
which was essentially a 9600 baud modem that took up most of a rack, 
plus a rather large card that plugged into the UNIBUS on the PDP-11."

He also told me the Unix v6 Arpanet code was from San Diego. (I still 
don't know about the earlier connection but I assume was not on Unix.)

[TUHS] What was the first edition of UNIX that left AT&T

[Noel Chiappa]

    > From: "Jeremy C. Reed"

    > I don't know the key for "v" but maybe means "very distant host"

Yes. See:

  http://mercury.lcs.mit.edu/~jnc/tech/jpg/ARPANet/L77Dec.jpg

and look at the MIT-44 IMP (upper right center). It's listed as having a
PDP-11, with the /v, and that machine (LL-ASG, 1/44) was definitely on a VDH
(it was not in Tech Sq). (A VDH was basically an IMP modem interface
hardware-wise, but made to look like a host at a high level within the IMP
software.)

    > He also told me the Unix v6 Arpanet code was from San Diego.

Err, he may have gotten it from San Diego, but they didn't write the code, it
was written at UIll. See:

  http://minnie.tuhs.org/cgi-bin/utree.pl?file=SRI-NOSC

which contains a copy of the code, which came to me via NOSC in SD.

      Noel