Paper tape readers (was Re: IBM 1403 Printer)
Charlie Gibbs
r...@yrloc.ipsa.reuter.COM (Robert Bernecky) writes:
>Oh, and on a slightly different note: Someone was muttering about
>paper tape readers: I think Burroughs held the record for the best
>paper tape reader in creation: The reader on the B205 and the B220
>(with drum and tube mainstore, respectively) would read 1000 characters
>per second. That may not sound impressive, but it plenty fast, b'wana.
I once worked on a Univac 9300 that had a Regnecentralen RC-2000
(some sort of Danish) paper tape reader kludged on to it. It would
crash the system whenever you turned it on, but once it got going it
would fairly scream. The 2000 in RC-2000 presumably referred to its
top speed of 2000 characters per second. Most programs that used it
were G/L proof lists that would read 10-character records and print
them on our 600-lpm bar printer, so it never got a chance to run at
full speed - until I added a "summary" option to the program just so
it could. It was really something to see that thing shoot tape
through at 200 inches per second.
>And the kicker was: It could stop on ONE character while doing so!
>
>The last IBM paper tape reader I saw had trouble stopping within a
>few feet of tape...
Our machine didn't stop right away either, but that didn't
matter - it contained a 256-byte buffer (a little core stack!) that
it would try to keep approximately half-full. When you loaded a
tape you'd hit a button that would cause it to read the first chunk
of tape into the buffer, and it would satisfy the computer's requests
at channel speeds while reading more tape as it had to.
One quirk of this reader was that it ignored the sprocket holes.
Since most of the tapes we read were punched with odd parity, it
didn't matter because there was always at least one punch in one of
the other tracks. It wouldn't read nulls too well, though. The one
client who gave us tapes punched in ASCII left several inches of
totally blank tape (except for the sprocket holes) between records -
the reader would skip over these sections at top speed looking for
the next byte.
This machine had no reels - you put a roll of tape into a
compartment and fed it under the head, where a capstan and pinch
roller pulled it through. The tape, once read, shot out of the
machine and into a heap on the floor unless you had a garbage bin
positioned to catch it. We'd then rewind the tape with one of
those hand-cranked winders (carefully, lest static electricity
build up and crash the system - but read on).
We finally got an application where the reader had a chance to
run flat-out. It was some sort of list of books out of print, which
was typed on IBM Selectrics that had paper tape punches attached.
These tapes were punched in IBM's "correspondence code" - since I
had to write the program to read them, I found out just how the code
came to be. I was pretty grossed out when I discovered that the
tape code was a bit-by-bit representation of the tilt and rotation
of the golf ball to print the corresponding character.
Anyway, my program would suck these tapes into the computer
at top speed - a couple of hundred feet per roll flying through the
reader at 200 inches per second. What we didn't realize was how
much static electricity this would generate. We got some almighty
zaps, each of which would crash the system. We wound up hanging a
grounded chain of paper clips by the machine where the tape would
hit it on the way out, and had a kettle steaming away in the
machine room, trying to get the humidity up enough to bleed off
the static charges. The things you gotta do sometimes...
Charli...@mindlink.UUCP
I'm trying to develop a photographic memory.
Douglas W. Jones,201H MLH,3193350740,3193382879
by Charli...@mindlink.UUCP (Charlie Gibbs):
>
> I once worked on a Univac 9300 that had a Regnecentralen RC-2000
> (some sort of Danish) paper tape reader kludged on to it. It would
> crash the system whenever you turned it on, but once it got going it
> would fairly scream. The 2000 in RC-2000 presumably referred to its
> top speed of 2000 characters per second.
But the RC-4000 was a computer system made famous (in the US at least)
by Per Brinch Hansen in his book Operating System Principles
(Prentice-Hall, 1973). This suggests, to me at least, that the RC-2000
paper tape reader was probably the system tape reader from a computer
that had the same name.
Both Brinch Hansen's textbook and his design for the RC-4000 operating
system were quite influential in the development of early post-Unix
operating systems. Unfortunately, the post-Unix generation of operating
systems seems to have been forgotten in the rush to spread the crowning
glory of 1968 vintage operating system technology over the entire world.
Doug Jones
jo...@cs.uiowa.edu
Jakob Nielsen
Jones,201H MLH,3193350740,3193382879 writes:
>From article <49...@mindlink.UUCP>,
>by Charli...@mindlink.UUCP (Charlie Gibbs):
>>
>> I once worked on a Univac 9300 that had a Regnecentralen RC-2000
>> (some sort of Danish) paper tape reader kludged on to it. It would
>> crash the system whenever you turned it on, but once it got going it
>> would fairly scream. The 2000 in RC-2000 presumably referred to its
>> top speed of 2000 characters per second.
>
>But the RC-4000 was a computer system made famous (in the US at least)
>by Per Brinch Hansen in his book Operating System Principles
>(Prentice-Hall, 1973). This suggests, to me at least, that the RC-2000
>paper tape reader was probably the system tape reader from a computer
>that had the same name.
Actually, the RC2000 was first used on the Gier computer from Regnecentralen.
The RC2000 was introduced in 1963 and was extremely advanced for its time. I
remember using an RC2000 when I first learned programming in 1973. It was still
impressive then, 10 years later. (Just think if there is any computer equipment
from 1981 which you would still find impressive).
The RC2000 was an especially fast paper tape reader because it did not slow
down the tape once it had stated moving through the reader at 18 km/h (11 mph).
The reader had internal electronic buffers to store the characters until they
were requested by the CPU which was often slower than the reader!
Some additional info on the RC2000 is given (in Danish) by Kurt Andersen on pp.
55-56 in the facinating book on Danish computer history "Niels Ivar Bech - en
epoke i edb-udviklingen i Danmark", P. Sveistrup, P. Naur, H.B. Hansen, and C.
Gram, Eds., Data Publishers, Copenhagen 1976.
W.C. Bulte
We want to find out if the thing still works allright.
Is there anybody out there who can help us to information on this terminal?
E.g. what kind of interface, sync/async etc.
Any help appreciated
_ ______________________________________________________________________
| / o / / _ Wilko Bulte Domain: wi...@idca.tds.philips.nl
|/|/ / / /( (_) uucp : [mcsun,hp4nl]!philapd!wilko
* Philips Information Systems Nederland phone: 055-432372 fax: 055-432103
____________________________________________________________________________
Niels Svennekjaer
It is quite funny to see that a part of my hobby pop up here, I use a
part of my sparetime to fix computers for the technical museum in
Aalborg, Denmark, and we have, among a lot of other computer stuff, a
RC4000 Computer, or in another way it take so much space in the
museum as the rest of the computers together (we have a GEIR-II, a
texas thing, a pdp-4 (?), a old old IBM something, and some more RC
computers, RC7000, RC-Piccoline,RC Piccolo)
If you come to Denmark, Aalborg go to the "teknisk museum" and ask if
you can see it. It stands in a closed room only used to amuse people
as me to get it to work, we have talked about making a little computer
museum, but due to the museum political ...... we can not do this.
As the time have past, is there only one person in Aalborg who knows
enough about the RC4000to get it to run 'stabile', and he (not me)
have so little sparetime to give, so we are 3-4 young (20-30years)
people trying to understand how to do it. but it is not easy, when we
are not supposed to power it up when he is not present...
But the RC7000, with its nova-2 or nova-200 cpu from DG is funny this
one I have known from 1982 (it was the first computer I tried, and I
got my own in 85), and have dismantled it so many times.... And it
still runs (sometimes) but we more and more have to fix the chips
ourself, because there exist no spare parts anymore, we got the last
from Regnecentralen two years ago...
Due to my 26 years of age it is funny to see the RC line of computers
DASK -> GIER -> RC 2000/4000/7000/8000/9000
and the small computers Piccolo,Piccoline,Partner,Rc900
Today Regnecentralen is called RC International and cooperate with
ICL.
+------------------------+---------------------------------------------------+
| _- | Greetings from Denmark, University of Aalborg |
| __- - | (Where the "*" is) |
| / *- +---------------------------------------------------+
| | --_ _--_ | Niels Svennekjaer (n...@iesd.auc.dk) |
| | -- \( ) | {uunet}!mcsun!sunic!dkuug!iesd!nrs |
| \ / __ |_ /O | |
| | | ( ) _--o +-| JEG GIK MIG OVER S0 OG LAND, DER M0DTE JEG EN |
| | | -- / (_)O |O| "GAMMEL" MAND. HAN SAGDE SAA OG SAGDE SAA ..... |
+ Denmark ---------------+---------------------------------------------------+
--
+------------------------+---------------------------------------------------+
| _- | Greetings from Denmark, University of Aalborg |
| __- - | (Where the "*" is) |
| / *- +---------------------------------------------------+
| | --_ _--_ | Niels Svennekjaer (n...@iesd.auc.dk) |
| | -- \( ) | {uunet}!mcsun!sunic!dkuug!iesd!nrs |
| \ / __ |_ /O | |
| | | ( ) _--o +-| JEG GIK MIG OVER S0 OG LAND, DER M0DTE JEG EN |
| | | -- / (_)O |O| "GAMMEL" MAND. HAN SAGDE SAA OG SAGDE SAA ..... |
+ Denmark ---------------+---------------------------------------------------+