Abuse or petulance?

[Kerr-Mudd, John]

On Mon, 29 Aug 2022 08:17:46 -0000 (UTC)

We had an arrangement with a nearby Technical College and used their
small mainframe (an Elliott 4120). Computing wasn't taught as such, but
we had an "after school club", which also met at lunchtimes. Cards were
hand-punched and for about three years I would drop them off at the
college on my way home, while picking up the previous day's cards and
lineprinter output. On average there'd be about 3/4 box of cards and a
two inch thick pile of paper each day.

The School Computer was a Ferranti Pegasus, donated by an insurance
company. It filled a classroom.

Recycled?

Yes. I was valve based!

The Forest Grammar School had a computer that featured on Tomorrow's
https://www.bbc.co.uk/archive/tomorrows-world--nellie-school-computer/zr24scw
It was an Elliott 405 according to Wikipedia and that looks about right
to me from what I remember of it.
I was not a pupil there but through various connections ended up with
various parts including a reel of 35mm magnetic tape (with sprocket
holes like 35mm film), some of the circuitry, complete with valves and
what I believe were delay lines. Most of the bits and pieces were not
much use but the valve heater transformers I acquired with their 6.3V
output were put to new use as part of a 5V power supply for playing
around with TTL ICs.
The real prize was two Cossor 1035 oscilloscopes which I managed to
repair with the aid of the circuit diagrams.

It seems the Pegasus and the 4120 were about the same vintage and the
designer of the Pegasus was ex Elliot’s. They may well have shared some
features.
When I first started in the GEC/ Marconi empire in 1979 ( ignoring two
summer jobs for them ), there were still old versions of later Elliot
machines around - mainly to support equipment in service but used by those
who had grown to love them ;-) Some were adapted versions of the military
boxes with a ‘control panel’ added etc.
My summer jobs included working on a system based on the 6100 uP, a uP
version of the PDP 8.

Now here's something worth chatting about - in the right forum, of course
- AFC.

Note xpost.

[Bob Eager]

It seems the Pegasus and the 4120 were about the same vintage and the
designer of the Pegasus was ex Elliot’s. They may well have shared some
features.

I first learned on an Elliott 4130 (like the 4120, but with hardware
floating point and memory management). Officially I learned BASIC; but
unofficially I was allowed Algol and FORTRAN, and an obscure language
called ML/I (http://www.ml1.org.uk).

I did some very naughty things on that machine, but ended up managing its
successor (at a university):

http://www.bobeager.uk/anecdotes.html

(see the first, second, fifth and sixth items).

[Roland Perry]

Post by Bob Eager
http://www.bobeager.uk/anecdotes.html
(see the first, second, fifth and sixth items).

"When the University of Kent's ICL 2960 mainframe was installed, it
came with a site engineer."

ICL had a two-tier scheme, with each site engineer being trained in the
general maintenance of the whole computer, plus specialist in one of
about six different sub-systems. Their job was to isolate which
sub-system was causing a problem, and if possible do a board-swap. Or
failing that, after 20 minutes call in the nearest (geographically)
expert engineer to assist. [Plus keeping a whole bookshelf of ring-bound
manuals up to date, and applying both routine hardware and software
patches, which arrived almost daily; plus some mechanical items like
line-printers and tape drives had a regular maintenance schedule].

The biggest sites had three shifts of multiple engineers, and spare
parts were flown around Europe at a moment's notice.

Everything except the OCP was standard across the range, although not
every site had one of everything (eg a drum - which was actually a very
large disc, and at the time the Comms Controller for remote access by
teletype wasn't very commonplace).

During 1977/78 I worked at the ICL training school in Letchworth as one
of two or three lecturers giving the "general" training course, which
was about eight weeks residential. There were other lecturers doing the
shorter expert courses, and I shadowed the one giving 2980 OCP (the most
complex of all the sub-systems).

We had one of everything to play with, and often got the feeling the
salesmen would rather ship them to customers. But that general course
was the critical path, because they couldn't hand over a new
installation until there was a site engineer available, and that
was where the shortage was, not in the factory or indeed customers
thrusting money at the company.

One of the drivers for the newly introduced baby of the range - the
2950, was that it might be simple/reliable enough that it didn't
actually need full-time site engineers.

[Norman Wells]

Post by Bob Eager
http://www.bobeager.uk/anecdotes.html
(see the first, second, fifth and sixth items).

  "When the University of Kent's ICL 2960 mainframe was installed, it
   came with a site engineer."
ICL had a two-tier scheme, with each site engineer being trained in the
general maintenance of the whole computer, plus specialist in one of
about six different sub-systems. Their job was to isolate which
sub-system was causing a problem, and if possible do a board-swap. Or
failing that, after 20 minutes call in the nearest (geographically)
expert engineer to assist. [Plus keeping a whole bookshelf of ring-bound
manuals up to date, and applying both routine hardware and software
patches, which arrived almost daily; plus some mechanical items like
line-printers and tape drives had a regular maintenance schedule].
The biggest sites had three shifts of multiple engineers, and spare
parts were flown around Europe at a moment's notice.
Everything except the OCP was standard across the range, although not
every site had one of everything (eg a drum - which was actually a very
large disc, and at the time the Comms Controller for remote access by
teletype wasn't very commonplace).
During 1977/78 I worked at the ICL training school in Letchworth as one
of two or three lecturers giving the "general" training course, which
was about eight weeks residential. There were other lecturers doing the
shorter expert courses, and I shadowed the one giving 2980 OCP (the most
complex of all the sub-systems).
We had one of everything to play with, and often got the feeling the
salesmen would rather ship them to customers. But that general course
was the critical path, because they couldn't hand over a new
installation until there was a site engineer available, and that
was where the shortage was, not in the factory or indeed customers
thrusting money at the company.
One of the drivers for the newly introduced baby of the range - the
2950, was that it might be simple/reliable enough that it didn't
actually need full-time site engineers.

And you criticise me for allegedly posting 'off topic'!

Some hypocrisy there, methinks.

[Kerr-Mudd, John]

On Tue, 30 Aug 2022 09:17:07 +0100

Post by Norman Wells

Post by Bob Eager
http://www.bobeager.uk/anecdotes.html
(see the first, second, fifth and sixth items).

  "When the University of Kent's ICL 2960 mainframe was installed, it
   came with a site engineer."
ICL had a two-tier scheme, with each site engineer being trained in the
general maintenance of the whole computer, plus specialist in one of
about six different sub-systems. Their job was to isolate which
sub-system was causing a problem, and if possible do a board-swap. Or
failing that, after 20 minutes call in the nearest (geographically)
expert engineer to assist. [Plus keeping a whole bookshelf of ring-bound
manuals up to date, and applying both routine hardware and software
patches, which arrived almost daily; plus some mechanical items like
line-printers and tape drives had a regular maintenance schedule].
The biggest sites had three shifts of multiple engineers, and spare
parts were flown around Europe at a moment's notice.
Everything except the OCP was standard across the range, although not
every site had one of everything (eg a drum - which was actually a very
large disc, and at the time the Comms Controller for remote access by
teletype wasn't very commonplace).
During 1977/78 I worked at the ICL training school in Letchworth as one
of two or three lecturers giving the "general" training course, which
was about eight weeks residential. There were other lecturers doing the
shorter expert courses, and I shadowed the one giving 2980 OCP (the most
complex of all the sub-systems).
We had one of everything to play with, and often got the feeling the
salesmen would rather ship them to customers. But that general course
was the critical path, because they couldn't hand over a new
installation until there was a site engineer available, and that
was where the shortage was, not in the factory or indeed customers
thrusting money at the company.
One of the drivers for the newly introduced baby of the range - the
2950, was that it might be simple/reliable enough that it didn't
actually need full-time site engineers.

And you criticise me for allegedly posting 'off topic'!
Some hypocrisy there, methinks.

Norman likes to argue; he can carry on doing so in unnm, if they still
want to keep him happy.



FU to unnm only