From mboxrd@z Thu Jan 1 00:00:00 1970 From: clemc@ccc.com (Clem Cole) Date: Sat, 23 Aug 2014 15:32:45 -0400 Subject: [TUHS] networking on unix before uucp In-Reply-To: <201408230308.s7N38kId012165@cuzuco.com> References: <201408230308.s7N38kId012165@cuzuco.com> Message-ID: Years before cu/ct show up in the AT&T distributions, we* at CMU/EE Dept we created a program called "connect" that had similar functionality. BEGIN - old guy's memories .... In the mid 70s the CS Dept "Front End" was two PDP-11's that all the serial ports in the CS offices and terminal rooms were connected too, and they were switched the PDP-10s, C.mmp and CM* over DR-11Bs IIRC. EE & CS were in a different buildings and we had not yet created the "distributed front-end" out of LSI11's (we did that in EE actually later when we started to playing with 3Meg ethernet). We did have a couple of serial lines to the other building and had to share them. Since we wanted to use the UNIX system as a our own timesharing system independent of CS, but still get to the CS machines over those shared lines, "connect" was created so we could share the very few serial connections to CS "front-end" and then we expanded it to be what we used to support the microprocessor lab. Another thing I remember from those days is that DEC DH11's serial ports were expensive (DZ's did not exist and Able Computer did not yet exist). We also had DL/KL11's - which had nasty interrupt behavior. With the development of the CS Front End CMU had it's own RS-232C interface for the Unibus called an ASLI - Asynchronous Line Interface - which were similar too KL11s but had some other improvements (I've forgotten what was different - should try to find Jim Teetor who I think was the creator). I have memories of some of first serial driver learnings chasing issues with the ASLI. I've complete forgotten the details now, but having come over from the IBM/TSS and CMU's version of TOPS-10, I remember thinking is was strange but so cool that driver was in a HLL not assembler. END - old guy's memories .... * "we" - Dan Klein, Tron McConnell, Ted Kowalski and I all hacked on it at different times. Frankly, I really do not remember who did what. Obviously it it ran on V6 and Ted's V6++ system, but I don't think it ever ran on V5. I'm pretty sure Ted took it back to Summit after his OYOC year, so I do not think cu had been done there yet. On Fri, Aug 22, 2014 at 11:08 PM, Brian Walden wrote: > Doug McIlroy wrote: > > > I was wondering if Unix had any form of networking before uucp > > > > Right from the time Unix came up on the PDP-11 it was > > networked in the sense that it had dial-in and dial-out > > modems. Fairly early on, when Unixes appeared in other > > Bell Labs locations, Charlie Roberts provided a program > > for logging into another machine. It had an escape for > > file transfer, so it covered the basic functionality > > of rsh and ftp. It was not included in distributions, > > however, and its name escapes me. Maybe scj can add > > further details. > > > > Doug > > Are you thinking of the cu (call unix) command? But that was included in > v7, > and don't think it was part of uucp. The escape was ~ So a ~. to hangup, > ~%put to send a file to the remote and ~%take to get one, and ~~ to send > a ~ > > later on, there was a ct (call terminal) command, expecting a terminal at > the end > of phone line instead of another machine. > _______________________________________________ > TUHS mailing list > TUHS at minnie.tuhs.org > https://minnie.tuhs.org/mailman/listinfo/tuhs > -------------- next part -------------- An HTML attachment was scrubbed... URL: