On Mon, Jan 25, 2016 at 11:18 AM, Pete Turnbull <pete(a)dunnington.plus.com>
wrote:
The University of Leeds did something similar - an
11/34 with a lot of
Emulex serial lines and an RX02 to boot from, was connected to their
Amdahl. I
Indeed back in the day that was a very popular PDP-11 configuration at a
lot of places (except UCB who had a giant patch board in each building).
Noel, didn't MIT have something running SupDup?
CMU called this configuration the "terminal front end" or just "Front
End"
(FE). I know commercially the Timeshare guys did this with their PDP-10s -
they called them "terminal switches" but started with CMU EE/CS version.
I also remember walking into a computer room in one of the big banks in
NYC and the only DEC equipment was PDP-11 running the terminals -
everything else in the room was big blue.
One of my CMU classmates that went to IBM friends told me that the way the
Series-1 finally got funded at IBM was to try to sell against DEC in just
that market. He said that it was why the S1 was IBMs first pure ASCII
machine and had RS232C ports (when it first came out IBM did not even
make/sell an terminals that talked to it - most customers were buying
VT-100 or clones).
Also, the CMU Front End was originally two system depending if you were
talking to the University's main Computer Center or to the CS/EE Dept
systems. Both FE's were 11/20s originally, then later 40e's but not
connected to each other. CMU had also designed it's own serial port for
them which we called an ASLI or Asynchronous Line Interface because DH-11
and DL11 ports were too expensive at the time. As more and more "large
systems" systems (read UNIX based 11's and Vaxen) showed up on campus
(around the time as I was leaving) the "Distributed FE" was being developed
in EE, originally using LSI11s and 3Mb Xerox ethernet. It was then further
cost reduced to 8085s on multibus boards by Andy Bechtolsheim (which he
later redesigned to use a 68k at Stanford - ie. the Sun board traces it
roots to the PDP-11 being used an embed terminal front-end :-)
Anyway - the key point being made is that the DEC sold a large number of
embedded PDP-11 as a popular way to driving terminals and modem pools into
larger systems. No DEC SW ran on them -- they were "purpose built." DEC
pretty much owned that business until finally cheap microprocessors and
cheap ethernet connections displaced them for the tasks. But in truth,
that really was limited because by the the personally computer and
workstations had begun to replace the "glass TTY."
Clem