musical processing
David B. Thomas
The noises are not coming from the speaker or the monitor, and when I adjust the
CPU speed switch, the sounds speed up and slow down. They are usually scratchy
sounding, but under certain circumstances it whines or hums. Perhaps a ceramic
capacitor or resonator is piezo-ing...just a guess.
Yesterday, it finally happened. I was doing some serious crunching with a
very non-serious cruncher (dbase), when it actually seemed to be making
musical notes -- at random, but vaguely sing-song sounding, like in the old
science fiction movies.
After a few minutes of feeling "awwww, that's so CUTE!!" (like seeing a
little birdie or something), I thought that there might be some historical
basis for the 'bloop beep bleep' sounds used to suggest 'computing' in old
science fiction films.
I seem to remember hearing of a computer that was made to play Greensleves,
as a side effect of going other processing. Did the noise use to come from
nearby AM radios? The circuitry itself?
Anyway, I just had to share this! If you have an Equity 3, start listening
closely when it's crunching. This is real.
David
*** Jeff ****
the head stop at musical frequencies and plays "A Bicycle Built for Two."
I always thought that was neat, and it seems relevant.
-------------------------------------------------------------------------------
Jeff Axelrod | "If you chose not to decide, you still have made
| a choice" -Neil Peart, Rush ///
ja2...@uxa.cso.uiuc.edu | \XX/ AMIGAUSER
-------------------------------------------------------------------------------
Henry Spencer
>My Epson Equity 3 (IBM clone) at work makes little noises when doing processing.
>The noises are not coming from the speaker or the monitor, and when I adjust the
>CPU speed switch, the sounds speed up and slow down. They are usually scratchy
>sounding, but under certain circumstances it whines or hums. Perhaps a ceramic
>capacitor or resonator is piezo-ing...just a guess.
Changes in power demand can make switching power supplies make noises.
Iris workstations are notorious for this.
--
NFS: all the nice semantics of MSDOS, | Henry Spencer at U of Toronto Zoology
and its performance and security too. | he...@zoo.toronto.edu utzoo!henry
Roger Ivie
> My Epson Equity 3 (IBM clone) at work makes little noises when doing processing.
> The noises are not coming from the speaker or the monitor, and when I adjust the
> CPU speed switch, the sounds speed up and slow down.
The only machine I ever heard rumors of the cause being traced down on (the
Digilog Microterm II), it turned out to be electrical noise being picked up
by the amplifier for the speaker.
It's actually quite nifty; you can get a good feel for what the machine's
doing just by listening. For example, you can tell whether your program's
stuck in an infinite loop or actually doing loops.
I have a friend that actually went out of his way to ensure that the
speaker amplifier in one of his new designs would have plenty of opportunity
to pick up noise to duplicate this; he ran the input wires all over just
for this reason.
===============================================================================
Roger Ivie
35 S 300 W
Logan, Ut. 84321
(801) 752-8633
===============================================================================
mnyk...@cc.helsinki.fi
final chord broke the drive.. :-(
--
Matti Nyk{nen
CS Student at Helsinki U, Finland
email: mnyk...@cc.helsinki.FI
The best opinions available; get them while they're hot!
John Shipman
+--
| Anyone else have any good music stories, or adventures
| with the ol' 1130?
+--
In the mid-seventies, IBM brought out an early desktop
computer called the 5100. It had a one-line display with
room for about 24 characters. A toggle switch selected
one of two languages, BASIC and APL.
The APL was appallingly slow---to generate a random
permutation of the first five natural numbers, and then
sort it, took several seconds---but it was very solid.
Years later I heard a rumor that explained how they could
generate a solid new APL implementation in such a short
time. The 5100 used the IBM 1130 instruction set as its
internal micro-machine! So its implementation had the
robustness that comes from many years of being tortured by
canny young programmers in small colleges all over the
country.
--
John Shipman/Zoological Data Processing/Socorro, NM/jo...@jupiter.nmt.edu
``Let's go outside and commiserate with nature.'' --Dave Farber
Tim McGuire
>My Epson Equity 3 at work makes little noises when doing processing.
>
>I seem to remember hearing of a computer that was made to play Greensleves,
>as a side effect of going other processing. Did the noise use to come from
>nearby AM radios? The circuitry itself?
Back around 1976, when I was a student at LeTourneau College, we had an
ancient (even for those days) IBM 1130. One of the lab assistants, Harvey
Block (are you out there Harv?), got the bright idea of hooking a signal
tracer to the least significant bit of the accumulator extension (well,
actually the panel light for it.) The result was a lot of interesting
noises whenever the machine was processing. A couple days later, I went
down to the lab; Harvey caught me at the door and said "Watch this!".
The room was full of beginning FORTRAN students struggling with their
programs. Harvey quickly turned on the signal tracer, started his program,
and stood back as the computer played a very credible rendition of "Popcorn".
Needless to say, all other activity in the room ceased for a while.
Tim McGuire
mcg...@cs.tamu.edu
John Little - Nihon Sun Repair Depot
%
%Changes in power demand can make switching power supplies make noises.
%Iris workstations are notorious for this.
%--
%NFS: all the nice semantics of MSDOS, | Henry Spencer at U of Toronto Zoology
How about those U.K. produced ICL minis which had a speaker hooked
into the address bus (buffered/summed/amp'ed??) and constantly produced
a horrendous series of cackles and whoops.
Legend has it that `real' ops could tell exactly what the beasties were
doing by just listening to this cacophony (there were probably a fair
proportion of ops carted off by the little men in white coats after
being subject to it for days at a time, too!!)
---------------------------------------------------------------------------
| John Little - gai...@Japan.Sun.COM | Sun Microsystems. Atsugi, Japan |
---------------------------------------------------------------------------
- A fool must now and then be right by chance -
Tim Mikkelsen
> In the mid-seventies, IBM brought out an early desktop
> computer called the 5100. It had a one-line display with
> room for about 24 characters. A toggle switch selected
> one of two languages, BASIC and APL.
I think the 5100 had a small CRT (like about 5") and it did have a
toggle for selecting APL or BASIC (but I think you had to cycle power).
We got to look at one at Iowa State when they first came out and we
popped the thing open - it looked like a spaghetti factory on the
inside.
> Years later I heard a rumor that explained how they could
> generate a solid new APL implementation in such a short
> time. The 5100 used the IBM 1130 instruction set as its
> internal micro-machine! .......
I understood that the 5100 was based on a small 360 chipset. (Which
also had a reasonable APL implementation.)
A single line 24 character machine (which might be what you are thinking
of) was a machine that HP introduced in the mid 70s called the
9825/9831. The 9825 was an HPL machine (instead of APL). HPL (Hewlett
Packard Language) is a language based on earlier calculator programming
languages) machine. The 9831 was the same computer with a different
system ROM (which was located in a removeable drawer on the right-hand
side) that let it speak BASIC. They were pretty cute. These machines
were based on a 16 bit CPU that HP made called the BPC (binary processor
chip) which was a re-implementation of the HP 1000 computer line (which
was around in the late 60s and descendants are still being sold today).
Tim Mikkelsen
Never Kid A Kidder
motors in it (I'm not sure what they're for; separate rollers or
something?), and when it's plotting, it comes out with the most
amazing little whispers, like Hewey, Dewey and Lewey in Silent
Running. Very musical (well I think so).
Mahesh Subramanya
Not quite the same thing, but waaaays back (well, mid 70s), I remember
working on a Dec 1090. The *high* speed line printer was capable of
some pretty fancy overstriking, repeat strokes, etc, during the processs
of which it would generate the most amazing shrieking sound (during line
feeds) machine gun rattles (overstrikes), and chattering (normal action).
Apart from the massive thrill of generating line printer pictures, I had
managed to generate quite a few documents which would *play* some pretty
neat tunes on the printer. My favourite was the Indian national anthem
( a rather atonal song, which was still quite distinctive). I also
distinctly remember being hauled on the carpet for doing that.....
************************MVS? EX-TER-MI-NATE!!**************************
Mahesh Subramanya INTERNET: mah...@darwin.cc.nd.edu
Senior Analyst
Office of University Computing NeXT: mah...@numenor.cc.nd.edu
University of Notre Dame Voice: (219) 239-5600 x6421
Notre Dame, IN 46556
************************************************************************
Chris Behrens
> managed to generate quite a few documents which would *play* some pretty
> neat tunes on the printer. My favourite was the Indian national anthem
> ( a rather atonal song, which was still quite distinctive). I also
> distinctly remember being hauled on the carpet for doing that.....
>
I have heard my father talk of such "songs" played by the printers on
some comp. system. Tiger Rag was the song he mentioned.
How is this done ??? Is there some way to guess the particular sounds
or pitches that will be made without hours and hours of trial and error
and ribbon and paper ???
Just curious...
Chris Behrens Chr...@Hubcap.Clemson.Edu
Terry Kennedy, Operations Mgr
> I have heard my father talk of such "songs" played by the printers on
> some comp. system. Tiger Rag was the song he mentioned.
I remember an implementation of the "1812 Overture" for System/360. You
needed a 360/40, 2 2501 card readers, a 1442 punch, 4 2410 tape drives, a
1403 printer on a 2821 control unit and 2 table radios. The 2821 had to
have one of the interlocks disabled and a different pattern loaded into
it's control memory (core, of course 8-).
I may have an audo recording of this (reel-to-reel) around somewhere. The
playing of this tune was grounds for immediate dismissal, as the interlock
you bypassed in the 2821 was the one that prevented *all* the 1403 hammers
from firing at once. This had the effect of destroying some of the output
transistors in the 2821 (but you got really *great* cannon shots from the
printer). Of course, you had to fabricate really creative excuses for the
IBM rep when he discovered that all 136 60-milliamp fuses in the 1403 were
blown. We generally said "Must have been a voltage surge from that really
bad thunderstorm yesterday".
Terry Kennedy Operations Manager, Academic Computing
te...@spcvxa.bitnet St. Peter's College, US
te...@spcvxa.spc.edu (201) 915-9381
Tim McGuire
>Years later I heard a rumor that explained how they could
>generate a solid new APL implementation in such a short
>time. The 5100 used the IBM 1130 instruction set as its
>internal micro-machine! So its implementation had the
>robustness that comes from many years of being tortured by
>canny young programmers in small colleges all over the
>country.
I remember the APL implementation on the 1130. You essentially had to use
the 1130 as a "personal computer". It didn't make one very popular with the
FORTRAN crowd who were used to batch processing. Anyone out there who
actually used 1130 APL?
Tim McGuire
mcg...@cs.tamu.edu
--- .sig under construction
Douglas W. Jones,201H MLH,3193350740,3193382879
by t...@hpfcbig.SDE.HP.COM (Tim Mikkelsen):
>
> I think the 5100 had a small CRT (like about 5") and it did have a
> toggle for selecting APL or BASIC (but I think you had to cycle power).
> We got to look at one at Iowa State when they first came out and we
> popped the thing open - it looked like a spaghetti factory on the
> inside.
>
> also had a reasonable APL implementation.)
>
about that machine is that the reason IBM was building it was that they
had lots of surplus production technology for their 360 series of mainframes
since they had just switched over to the 370 generation of machines.
The IBM Office System 6 word processors were also supposed to house
left-over 360 parts.
Doug Jones
jo...@herky.cs.uiowa.edu
Douglas W. Jones,201H MLH,3193350740,3193382879
> I have heard my father talk of such "songs" played by the printers on
> some comp. system...
This is true. To understand how it was done, you have to understand the
old chain and drum printer technologies. In the following, I will
oversimplify and overgeneralize a bit:
Chain printers had one hammer per print position on the page, with a chain
of type slugs that ran horizontally across the print line. Usually, the
chain had at least three alphabets in it. IBM used print trains instead of
chains; a print train worked the same way, except that you could easily
change the train in a printer, and you could drop the train, scrambling
the slugs because they were only loosely coupled to each other.
Anyway, with chain and train printers, as the appropriate character passed
its print position, the hammer would fire, hitting the back of the paper
and banging the paper into the ribbon into the passing type slug. When the
printer was out of adjustment, the characters tended to shift horizontally
a bit.
Drum printers had one hammer per print position, and printed similarly,
but instead of a chain, they had a rotating drum with the entire alphabet
repeated a few times around the drum at each each print position. A 132
column printer supporting 64 printable characters would have 64x132
individual type elements on the surface of the drum. A badly adjusted
drum printer would print with characters offset vertically from where they
were supposed to go.
To see how a tone could be printed on a chain printer, imagine printing
an entire line of the same character. As the type slug for that character
zips across the page, each hammer hits it in sequence. On a 300 Line Per
Minute printer, each line can be printed in 1/5th of a second, so that
means that in 1/5 of a second, 132 hammers hit the same character, giving
a frequency of 660 hz (actually the frequency was higher because I left
out the time it takes to advance the paper from my calculations).
To see how a tone could be printed on a drum printer, imagine printing
the entire alphabet on a line. Since all the A's are printed first, then
all the B's, then all the C's, printing a row of A's just produces a nice
loud bang, but printing the alphabet produces a ringing tone.
Other frequencies can be obtained from other printed lines. For example,
on a chain printer, printing the alphabet in order, in the direction of
chain motion, will produce a higher tone (or a loud bang in the print
slug spacing equals the print column spacing, but usually the spacing was
about 3 print columns). Printing the alphabet in reverse order would
produce a lower tone.
Line printer musicians would study the physics of their printers to
determine the character sequences that would produce each note of a
reasonable musical scale, then prepare print files to create the music.
As a musical instrument, the old impact line printers were awful. The
notes all had shrill, non-harmonic, overtones because of the steel on
steel impacts of hammer on type-slug (poorly padded with paper and ribbon),
and the beat of the mechanism advancing the paper at a few lines per
second didn't usually contribute anything useful.
On the other hand, it didn't take a trained ear to pick out the tune.
The musical notes were harsh, but their tones were clear enough that
when a printer started making music instead of merely deafening everyone
around it, people noticed!
Doug Jones
jo...@herky.cs.uiowa.edu
Bernie Cosell
The 'how' is very simple -- how the guys that did so actually figured
it all out is something that I don't know. The real printer that
spawned the first hacks of this type that I saw was the venerable
1403. If I remember right it was a band printer (although the idea is
the same for a drum printer):
The printer has a print hammer in each column, and a chain whizzing
around with the printing-dies on it. Your line-to-be-printed was
loaded into a buffer in the printer, and as the chain whizzed, when the
die for the required character lined up with the print hammer for the
column it was needed in, the hammer fired. Well, what if you had a
complete line-full of 'E's? Well, the print die for E would whiz
across, and every print hammer would fire in a neat left-to-right
"arpeggio". What if you printer "ABCDEFG...."? Well, nothing much
would happen until the chain was in exactly the right position, and
then EVERY hammer would fire in a gigantic WHAM.
Thus, whether you want it to or not, everything you print will have a
characteristic sound. It is not hard (in theory, at least --- lord
knows how the hackers actually figured this all out) to print exactly
the right characters to print so that the hammers will fire at a
particular frequency [440 hammers/second for 'A', etc...]; but making
the chars convenient, you can get higher frequencies, by going for
optimal inconvenience you can make lower frequencies. By having
multiple-hammers hit at the same time you can vary the overall
amplitude. Even polyphonic is only a little bit trickier.
/Bernie\
David B. Thomas
>The only machine I ever heard rumors of the cause being traced down on (the
>Digilog Microterm II), it turned out to be electrical noise being picked up
>by the amplifier for the speaker.
Tonight I finally had the energy to remove the cover from the pc and poke
around. Sure enough, the noise was coming from the speaker, but was not
affected by the volume control -- it's leaking into the speaker amp!
David
Steven Bellovin
> I remember the APL implementation on the 1130. You essentially had to use
> the 1130 as a "personal computer". It didn't make one very popular with the
> FORTRAN crowd who were used to batch processing. Anyone out there who
> actually used 1130 APL?
Depends on what you meant by ``used''. Do you remember the keyboard?
ARRGHHH!
That's better, I've calmed down. Let me explain the keyboard situation.
APL requires a large (and unusual) character set. The 1130 had a keyboard
stolen from an 029 keypunch. (In fact, it even provided input in
card code. But for some reason, they left the multipunch key off
on the 1130...) The 029 didn't have nearly enough keys, hence the 1130
didn't, either. So some brilliant soul assigned *3* meanings to each
key, and decreed two keys to be shift. One shifted up, and the other
shifted sideways. (The images come from the little decals you were
supposed to glue to the front of each key.) If you hit a shift key
twice, it locked you in that mode. Fortunately, the 1130 was reasonably-
well equipped with blinking lights, so the accumulator and extension
registers would indicate the current shift status of the keyboard.
That said, the rest of the APL system was really a waltzing bear sort
of thing. I'm still amazed they crammed that much code into so little
space, even with the disk I/O they had to do for almost anything. But
that made things run *very* slowly. Still, it was better than the
drain-bamaged FORTRAN the machine had.
Werner Icking
>In article <97...@hubcap.clemson.edu>, chr...@hubcap.clemson.edu
>(Chris Behrens) writes:
>> I have heard my father talk of such "songs" played by the printers on
>> some comp. system...
>This is true. To understand how it was done, you have to understand the
>old chain and drum printer technologies. In the following, I will
>oversimplify and overgeneralize a bit:
[... the very good and clear explanation deleted ...]
One small addition: you could create the rhythm by the vertical advance of
the paper. Bigger spacing made a bigger rest, smaller spacing a smaller.
Overprinting made the fastest sequence of tones, but sometimes you had to
insert at least a small advance of the paper to avoid cutting the paper.
--
Werner Icking ick...@gmdzi.gmd.de (+49 2241) 14-2443
Gesellschaft fuer Mathematik und Datenverarbeitung mbH (GMD)
Schloss Birlinghoven, P.O.Box 1240, D-5205 Sankt Augustin 1, FRGermany
Brian Kantor
three bits of the accumulator were wired to an audio amp that fed a
speaker in the console. Normal jobs chuckled along as it ran; when a
job batch finished the operating system ("UCSD Presto") would sound a
bosun's whistle. It got positively obnoxious when it was waiting for
operater input or a tape mount.
"Piano in D" was a nifty one; it set a clock interrupt then halted so
that the D-register console buttons were active, then started up again
and looped a bunch of times to make an appropriate note in the
speaker. 48-key piano...
But the best was having it play "The Stars and Stripes Forever". It
used the console speaker for the fife, one to three tape drives (with
their loud vacuum flapper valves) as the bass section, and back in the
days when the printer was directly attached, the printer was the
percussion section.
A few days before we dismantled the beast, I made a super-8 stereo
sound film of it playing the tune. It's still around here
somewhere... I may still even have a copy of the boot tape, I think.
- Brian
Dave Hsu [Contractor]
>I have a program for my Commodore 64 the bangs the disk drive head into
>the head stop at musical frequencies and plays "A Bicycle Built for Two."
Hopefully you won't run this too often. Probably the biggest problem
with the original 64 (besides having a poorly shielded video section) was
that the old 1541(?) drives were poorly ventilated, and if you let the
drive heat up enough, the head stop setscrew would loosen up and boom,
your stepper alignment lost its zero on the next boot.
-dave
--
Dave Hsu Sun Microsystems MS 14-40 (415) 336 5253 in a pinch:
h...@Eng.Sun.COM 2535 Garcia Avenue, Mountain View, CA 94043 h...@eng.umd.edu
"Are you familiar with Twiggly, the gnome who guards the entrance
to the Cave of Knowledge?" "Ah, yes yes. Twiggly!"
Douglas W. Jones,201H MLH,3193350740,3193382879
by url...@smurf.sub.org (Matthias Urlichs):
>
> In alt.folklore.computers, article <58...@bbn.BBN.COM>,
> cos...@bbn.com (Bernie Cosell) writes:
> <
> < What if you printer "ABCDEFG...."? Well, nothing much ... until the
> < chain was in ... the right position, and then ... a gigantic WHAM.
> <
> Many printers didn't have ABCDEFG, but a more reasonable character
> distribution.
>
The chain printers I used solved this problem by having the print slugs
in the chain spaced in something other than a one-slug-per-column spacing.
The one-slug-per-column spacing guarantees that there will be a pathological
character sequence that goes WHAM when you print it. By making the print
slugs farther apart, the WHAM goes away. If I remember correctly (from the
many times I cleaned the chain on the printer I used through most of the
1970's, the characters were spaced something like 2.5 print columns apart,
but they were in alphabetical order.
As a result, you could get a little wham by printing every other character
in every fifth column, for example:
A C E G I K M O Q S U W Y
This did no harm at all to the printer. I've presented this example
assuming that the chain moves from right to left (as it did on the printer
that I used) and that characters pass each point in alphabetical order
(also true of the printer I used).
The following sequence would print two small whams in quick sequence, where
the time between whams was 1/2 the time it took a print slug to get from
one column to the next:
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
This could be extended to give 4 whams in a row, with the following, but
after that, you'd have to change tactics:
AABB CCDD EEFF GGHH IIJJ KKLL MMNN OOPP QQRR SSTT UUVV WWXX YYZZ
==================
And does anyone remember the GE Terminet 1200 printing computer terminal?
It was one of the first 1200 baud terminals, and unlike many of its
competitors in the early 1970's, it printed both upper and lower case. The
print mechanism was essentially a scaled down version of the chain printer,
except that the chain was a rubber belt, with spring steel type fingers
sticking up out of the edge of the belt.
When it printed, it sounded like a creaky spring of the sort old rocking
chairs sometimes contain, and with a pair of pliers and a bit of patience,
you could pull the type fingers out of the belt and change the colating
sequence. (The only use I found for the latter was to get the convention
for zero and capital-oh more reasonable).
Doug Jones
jo...@herky.cs.uiowa.edu
Matthias Urlichs
< would happen until the chain was in exactly the right position, and
< then EVERY hammer would fire in a gigantic WHAM.
<
distribution.
And, when you figure out how to make every hammer fire at once, strange things
happen -- once, a long long time ago, I looked at the print chain of one of
these old things when it was on preventive maintenance. When the repair guy
noticed, I told him I was going to make up the slowest-printing text and the
fastest-printing, and compare speeds. He said OK, have fun...
A few hours later, I gave the operator a deck of cards, and could she print
them please?
On these cards was a nice printer speed arpeggio, culminating in what I
thought was the fastest speed, firing all hammers simultaneously.
You guess what happened -- the print speed went _way_down_ because (a) some
fuses blew, and (b) the chain broke.
I had been making up some real music, but they didn't allow me to try it.
I wonder why... ;-)
And then there was the TRS-80 at school... It had a reed relay on the cassette
drive to turn the motor on and off whcih ine could abuse as a speaker. Until
the contacts fused, that is. :-(
--
Matthias Urlichs -- url...@smurf.sub.org -- url...@smurf.ira.uka.de
Humboldtstrasse 7 - 7500 Karlsruhe 1 - FRG -- +49+721+621127(Voice)/621227(PEP)
Michael Ardai
-In article <1990Jul12....@ux1.cso.uiuc.edu- ja2...@uxa.cso.uiuc.edu (*** Jeff ****) writes:
--I have a program for my Commodore 64 the bangs the disk drive head into
--the head stop at musical frequencies and plays "A Bicycle Built for Two."
-
-Hopefully you won't run this too often. Probably the biggest problem
-with the original 64 (besides having a poorly shielded video section) was
-that the old 1541(?) drives were poorly ventilated, and if you let the
-drive heat up enough, the head stop setscrew would loosen up and boom,
-your stepper alignment lost its zero on the next boot.
In that case, open the 1541, loosen the two screws holding the stepper
and slowly twist it back and forth until it works. I had to do this a
few times on mine :-) By the way, never stack 1540 and older 1541 drives.
The bottom one can get so hot that its case will melt to the one above.
I once saw a stack of three 1540s that were fused together :-)
/mike
-Dave Hsu Sun Microsystems MS 14-40 (415) 336 5253 in a pinch:
-h...@Eng.Sun.COM 2535 Garcia Avenue, Mountain View, CA 94043 h...@eng.umd.edu
\|/ Michael L. Ardai ar...@bu-pub.bu.edu
--- ---------------------------------------------------------------
/|\ ...!sun!teda!maven.dnet!ardai (preferred)
John Shipman
+--
| I understood that the 5100 was based on a small 360 chipset.
| (Which also had a reasonable APL implementation.)
|
| A single line 24 character machine (which might be what
| you are thinking of) was a machine that HP introduced in
| the mid 70s called the 9825/9831. The 9825 was an HPL
| machine (instead of APL).
As for whether the 5100 was based on the 1130 or the 360,
I have no hard evidence, just rumors. Can anyone nail
this down?
I'm quite sure about the IBM brand and the APL language,
though. I saw the machine at an APL convention in Anaheim
(1974?). At that time, I was working on an APL
implementation for the ill-fated HP 300 (not the same
machine as the 3000), so -hp- paid my way to the convention.
--
John Shipman/Computer Science Department/New Mexico Tech/Socorro, NM 87801
(505)835-5301; jo...@jupiter.nmt.edu
Jeremy J. Epstein
> Tim Mikkelsen (t...@hpfcbig.SDE.HP.COM) writes:
> +--
> | I understood that the 5100 was based on a small 360 chipset.
> | (Which also had a reasonable APL implementation.)
> |
> this down?
An interesting hypothesis. The 5100 was released in 1974 I believe...
the price tag for the base model was $18,000. The machine could be
lugged around, so in a sense, based on both price and size it could
be considered an early personal computer.
The first announcement I saw said $1800, not $18,000...I had almost
convinced my father to buy me one until I found the real price.
--
Jeremy Epstein
TRW Systems Division
703-876-8776
jje%vir...@uunet.uu.net
Matthias Urlichs
ar...@bass.bu.edu (Michael Ardai) writes:
<
< [...] By the way, never stack 1540 and older 1541 drives.
< The bottom one can get so hot that its case will melt to the one above.
< I once saw a stack of three 1540s that were fused together :-)
Shit happens. ;-)
In my case, I ran a BBS system on my C-64.
It worked really well, although the things I had to do to it to make that
so-called computer behave like into a decent BBS belong in alt.hackers.
And then, after two months of uninterrupted operation, the power supply looked
kind of saggy -- it was starting do melt. Because a lump of plastic with some
metal inside conducts heat even worse, it started to flow off my desk faster
and faster.
Luckily, a cheap-and-dirty power supply for it needed only a 7805, a
transformer (of course), and two condensators. Those were the days...
Michael Hart
>In article <16...@yenta.alb.nm.us> d...@yenta.alb.nm.us (David B. Thomas) writes:
> ... got the bright idea of hooking a signal
>tracer to the least significant bit of the accumulator extension (well,
>actually the panel light for it.) The result was a lot of interesting
>noises whenever the machine was processing.
The first computer I ever learned to fix was an (ancient!) Burroughs
B-825(I think), used by the USAF as an early warning system. This
guy was a real MONSTER!! Power consumption was 65 KW/Hr !!!!
Each cabinet was 4'x4'x6-7'tall, and there were between 6 & 12 !~!!!
I remember that the memory cabinets were really sophisticated; they
were dual memory units - each cabinet had 2 ....64Kb..... _CORE_
memory units. WoW! .:-)
Anyway, the main console had a speaker and amp in it; the input to
the amp was from the LSB of the main accumulator (register, for you
youngsters out there!). After some time, you could tell the
approximate state of the machine by listening to the "tune" coming
from the console.
Years later, when I worked on various DEC PDP based equipment,
we would often take an FM radio into the computer room, tune it
up & down the dial until we found the CPU. This actually turned
out to be a valuable troubleshooting tool, at times.
--
-------------------------------------------------------------------------------
Michael G. Hart ha...@blackjack.dt.navy.mil / mh...@dtrc.dt.navy.mil
DTRC/DoD | "Wherever you go, there you are."- me
DISCLAIMER: If you want the Navy's opinion, talk to Secretary Cheney.
--
-------------------------------------------------------------------------------
Michael G. Hart ha...@blackjack.dt.navy.mil / mh...@dtrc.dt.navy.mil
DTRC/DoD | "Wherever you go, there you are."- me
DISCLAIMER: If you want the Navy's opinion, talk to Secretary Cheney.
Tim Mikkelsen
> about that machine is that the reason IBM was building it was that they
> had lots of surplus production technology for their 360 series of mainframes
> since they had just switched over to the 370 generation of machines.
>
> The IBM Office System 6 word processors were also supposed to house
> left-over 360 parts.
In the mid 70s, I visited a mass-mailing outfit and they had rows of
'smart' line printers. In talking with the technical people, these
were re-furbished 360s rebuilt to work as printer controllers hung off
of 370s. (Pretty clever - better than than taking them to the land
fill :-).
Tim M
Michael Meissner
Yep, I spent many an afternoon hacking away at the IBM 1130 we had at
the Baltimore Polytechnic Institute (public high school). I was there
from fall 1971 to spring 1973, but I didn't take computers until my
second year. While I did primilarly use Fortran in a batch card
environment, towards the end I did use APL and do all of the typical
'I can do this in one line of programming' that goes along with APL.
I remember once being left alone with instructions to shut the machine
down when I was done. I did turn off the master power switch, only to
discover that you really had to shut off the periphials first
(everybody assumed that since I hung around, I knew how to shut things
down), and that I trashed the APL master disk. As I recall, we got a
backup copy of the disk from a local community college.
--
Michael Meissner email: meis...@osf.org phone: 617-621-8861
Open Software Foundation, 11 Cambridge Center, Cambridge, MA
Do apple growers tell their kids money doesn't grow on bushes?
da...@exactly.me
Scott Lurndal
>On Wednesday, July 18, 1990 at 6:21:00 PM UTC-7, Brian Kantor wrote:
>> Our CDC3600 had a number of musical programs for it, since the top
>> three bits of the accumulator were wired to an audio amp that fed a
>> speaker in the console. Normal jobs chuckled along as it ran; when a
>> job batch finished the operating system ("UCSD Presto") would sound a
>> bosun's whistle. It got positively obnoxious when it was waiting for
>> operater input or a tape mount.
>> that the D-register console buttons were active, then started up again
>> and looped a bunch of times to make an appropriate note in the
>> speaker. 48-key piano...
>> used the console speaker for the fife, one to three tape drives (with
>> their loud vacuum flapper valves) as the bass section, and back in the
>> days when the printer was directly attached, the printer was the
>> percussion section.
>> sound film of it playing the tune. It's still around here
>> somewhere... I may still even have a copy of the boot tape, I think.
>> - Brian
>
>center consultants invited me into the computer room and showed me the comp=
>uter playing Stars and Stripes forever using the computer sounds (spinning =
>tapes, line printers, etc.). It was fantiastic! I think it was a Burrough=
>s 6700 mainframe computer at that time. If you have a video of this, it wo=
>uld be great if you could post it on YouTube. Otherwise, this will just be=
Quadibloc
Charles Richmond
> The prototype of the 5100 in the Smithsonian used an 1130 emulation; as this did not produce adequate results, production models used 360 emulation for APL instead. Both emulated the System/3 for BASIC.
>
one instance... when the two got different answers when evaluating the
same expression. How could that be... if the 5100 emulated APL\360???
--
numerist at aquaporin4 dot com
Quadibloc
J. Clarke
wrote:
>The 5100 ran a modified VS APL, not the earlier APL/360.
some compromises required.
gloga...@gmail.com
> The 5100 ran a modified VS APL, not the earlier APL/360.
hanc...@bbs.cpcn.com
"simulation". In IBM parlance, "simulation" meant that a
software program simulated another machine. In the prep for
S/360, IBM found that simulators ran rather slowly. In contrast,
"emulator" referred to hardware features, such as microcode, that
mimicked another machine. IBM found that microcode emulation ran
fast and was suitable for S/360 to act as older machines, a major
selling point.
Has anyone used a S/360 emulating a 7090 series machine, especially
with floating point work? How did that work out?