Re: 1132 printer history

[glen herrmannsfeldt]

In comp.lang.pl1 Charlie Gibbs <cgi...@kltpzyxm.invalid> wrote:

(snip)
> Most of the time we were listing short records on the (non-spooled)
> printer, which limited the demand on the tape reader. However, I
> did write one application which read everything to disk, and the
> reader cranked up to its rated 200 inches per second. Prolonged
> operation at this speed generated strong enough static charges that
> the occasional spark would induce a glitch in the channel interface,
> crashing the computer. To get the application to run, we hung a
> grounded string of paper clips such that the tape would touch it
> when being discharged from the reader, and boiled a kettle in the
> machine room to bring up the humidity.

The van de Graaf effect. I remember seeing what looked like
christmas garland on the output of 1403 printers, to discharge
the static electricity, maybe a little better than paper clips.

-- glen

[glen herrmannsfeldt]

In comp.lang.pl1 Walter Bushell <pr...@panix.com> wrote:

(snip)
> And IIRC C was developed in a very non card oriented environment. Well
> if C had been developed anywhere near an IBM environment they would
> have chosen a different comment delimiter.

There was a discussion some time ago on which C features were
borrowed from PL/I. Maybe not so many, but it seems that the
comment delimiter is one.

Also, I read not so long ago that the PL/I group didn't know what
the JCL group was using for an EOF card. I don't know who would have
changed, though.

-- glen

[hanc...@bbs.cpcn.com]

On Aug 17, 1:27 pm, "Charlie Gibbs" <cgi...@kltpzyxm.invalid> wrote:

> Yes, but think about those individual hammers.  Timing for them
> gets down into the sub-millisecond range.  Not bad for a mechanical
> device.

The Campbell Kelly book "Computer" talks about the success of the IBM
1401, a very popular computer (most popular?) of its day. The 1401
cpu was not a wonderpiece of electronics or speed, however, the 1403
printer was a wonderpiece of technology, priting extremely fast for
its day with high quality output.

The book notes that back then (late 1950s) IBM still had a lot of R&D
strength in mechanical engineering from its tabulator development.

[Dave Garland]

A lot of laser printers still use the tinsel, though it's usually
hidden in the output chute. Of course, there the static was
intentionally introduced as part of the imaging process, rather than
just being generated mechanically.

[Nico de Jong]

"Ahem A Rivet's Shot" <ste...@eircom.net> skrev i en meddelelse
news:20120817155415....@eircom.net...
> On 17 Aug 2012 13:27:35 GMT
> jmfbahciv <See....@aol.com> wrote:
>
>> Shmuel (Seymour J.) Metz wrote:
>> > In <PM0004C74...@users-ibook-g4-6.unknown.dom>, on 08/15/2012
>> > at 12:36 PM, jmfbahciv <See....@aol.com> said:
>> >
>> >>Where did you find tame peripherals?
>> >
>> > In the early days the peripheral were dogs rather than cats. Card
>> > jams, stacker jams, interlocks that didn't.
>> >
>> There was an operational art form which would diminish card jams
>> to almost none.
>
My worst experience with cards was with an IBM 2540. When reading or
punching a card, you could direct the card to any of 3 pockets. The middle
one (there are 5 pockets) could be used for read cards, and for punched
card.
Now then, when doing read and punch jobs simulteaniously, the operator was
supposed to remember if bots jobs used the middle pocket. I once saw, that
he hadnt remembered. From that day, operating both halves of the 2540 at the
same time was banned
/Nico

[Walter Bushell]

In article <877gsyl...@blp.benpfaff.org>,
Ben Pfaff <b...@cs.stanford.edu> wrote:

> Gene Wirchenko <ge...@ocis.net> writes:
>
> > On Fri, 3 Aug 2012 13:29:22 -0700 (PDT), hanc...@bbs.cpcn.com wrote:
> >
> > [snip]
> >
> >>Using analogies to everyday things seemed to be a very effective
> >>approach to teaching programming (for higher level languages).
> >>
> >>Start by describing the computer as a big dumb room with in-box,
> >>typewriter, adding machine, file cabinets, mailboxes, and a boss who
> >
> > These days, it might be that you have to describe what a
> > typewriter is in terms of a computer.
>
> Adding machine, too. I've never seen one outside of a museum.

On my first summer job I spent two weeks doing a least squares fit on
a square root Friden calculator. The poor slobs in economics did their
calculations on calculators that made you do multiplication digit by
digit.

--
This space unintentionally left blank.

[Walter Bushell]

In article <m3y5lef...@garlic.com>,
Anne & Lynn Wheeler <ly...@garlic.com> wrote:

> "Charles Richmond" <nume...@aquaporin4.com> writes:
> > ISTM that a "dataset" is IBM-ese for a disk file... And I think
> > "DASD" means hard disk storage. IBM has their own vocabulary for
> > things... For example, what most computer systems refer to as "system
> > calls", IBM has to call "supervisor calls". And of course we have the
> > PSW (program status word). :-)
>
> DASD is for direct access storage device ... generic term, it predates
> hard disks dominating the technology ... dates from when "drums" were
> more common ... drums even being used for main processor storage.
>
> dataset ... again predates disks becoming dominate technology.

I must mention the story of Mel the real programmer.

<http://www.cs.utah.edu/~elb/folklore/mel.html>

There were giants on the Earth in those days.

[Walter Bushell]

In article <k0k0ok$l9$1...@dont-email.me>,
"Charles Richmond" <nume...@aquaporin4.com> wrote:

> <hanc...@bbs.cpcn.com> wrote in message
> news:fc8161c5-df8b-4830...@l17g2000yqg.googlegroups.com...
> On Aug 16, 8:47 am, jmfbahciv <See.ab...@aol.com> wrote:
>
> >> One day our one, and only, disk drive had a head crash.
> >> The lead manager mounted the backup disk. Schreech.
> >> So, to run all of the FORTRAN kiddie decks, someone loaded
> >> the compiler from the cabinets of cards we had while the IBM
> >> FS had the disk parts laying all over the room.
> >
> >Student jobs had very low priority. If something like happened, the
> >students would just to have to wait or come back later. This was a
> >problem for commuting students, who didn't have the option of
> >returning to the college at night.
>
> At our college, students would be told to see Helen Hunt. The operator
> would say: "You have to go to hell and hunt!!!" ;-)
>
> --
>
> numerist at aquaporin4 dot com

Any references to Mike Hunt?

[Walter Bushell]

In article
<7ea68172-612d-4c52...@e29g2000vbm.googlegroups.com>,

hanc...@bbs.cpcn.com wrote:

> The Campbell Kelly book "Computer" talks about the success of the IBM
> 1401, a very popular computer (most popular?) of its day. The 1401
> cpu was not a wonderpiece of electronics or speed, however, the 1403
> printer was a wonderpiece of technology, priting extremely fast for
> its day with high quality output.
>
> The book notes that back then (late 1950s) IBM still had a lot of R&D
> strength in mechanical engineering from its tabulator development.

Much of the 1401 work was glorified tab shop work and many times the
sorts were done on tab equipment.

The 1401 was also much employed as an I/O device for larger confusors.

[John W Kennedy]

On 2012-08-17 07:55:48 +0000, Shmuel (Seymour J.) Metz said:

> In <502d17d1$0$6063$607e...@cv.net>, on 08/16/2012
> at 11:54 AM, John W Kennedy <jwk...@attglobal.net> said:
>
>> Define "early".
>
> 1950's through 1963.
>
>> The 1403-N1 certainly did it,
>
> That came later.
>
>> A 2401-7 could do 320000 bytes per second,
>
> Still slow, and not in the early mainframe days.
>
>> while the 2301 drum was fast
>> enough to require special double-bandwidth 2860 channels.
>
> Are you sure you're not thinking of the later 2305-1 disk, which
> required a 2-byte 2880? The 2301 was only 1.25 MB/s, and A22-6895-2
> doesn't seem to mention any special feature being required on the
> 2860.

Hmmm.... Looks like I am -- but not entirely. It wasn't a bandwidth
doubler, like the feature used by the 2305, but there was an optional
2860 feature which specifically addressed attaching the 2301, although
it seems to have addressed only the 360/67 and 360/85 at the time my
reference was printed. I gather it allowed the channel to tell the CPU
to temporarily leave storage alone, because the channel would be
stressing it too hard. Also, it seems that the 2301 could only be used
on channels 1 and 2.

--
John W Kennedy
"...if you had to fall in love with someone who was evil, I can see why
it was her."
-- "Alias"

[John W Kennedy]

On 2012-08-17 07:22:49 +0000, Shmuel (Seymour J.) Metz said:

> In <k0khi6$lf6$1...@speranza.aioe.org>, on 08/17/2012
> at 04:33 AM, glen herrmannsfeldt <g...@ugcs.caltech.edu> said:
>
>> At least the later IBM manuals explain Fortran IV as extensions to
>> Fortran 66, and use shading to mark the extensions.
>
> That would have to be on S/360, probably OS/360 FORTRAN G and H.

Plus FORTRAN F (DOS/360) and 44PS FORTRAN. Probably TSS/360 FORTRAN, too.

--
John W Kennedy
"Though a Rothschild you may be
In your own capacity,
As a Company you've come to utter sorrow--
But the Liquidators say,
'Never mind--you needn't pay,'
So you start another company to-morrow!"
-- Sir William S. Gilbert. "Utopia Limited"

[Charles Richmond]

"Shmuel (Seymour J.) Metz" <spam...@library.lspace.org.invalid> wrote in
message news:502deffe$3$fuzhry+tra$mr2...@news.patriot.net...
> In <m3y5lef...@garlic.com>, on 08/16/2012

> at 07:52 PM, Anne & Lynn Wheeler <ly...@garlic.com> said:
>
>>DASD is for direct access storage device ... generic term, it
>>predates hard disks dominating the technology ... dates from when
>>"drums" were more common ... drums even being used for main
>>processor storage.
>

> By the time IBM coined the term DASD, no new machines were using drums
> as main storage. Disks were more common by then, but drums had better
> performance.
>
> OTOH, nobody ever accused the noodle picker of having better
> performance :-(
>

At a PPoE, the finance department worked on a latter-day 370 type computer
(I forget the number of the box). They had a "noodle picker" that they used
to conserve disk space. If a file had *not* been accessed for two days, it
would automatically be sent to noodle-picker hell. The programmers wrote a
program that would be run every morning and would access each one of their
files... thus preventing their files from being sent to the noodle picker
backup. :-)

[Charles Richmond]

"Shmuel (Seymour J.) Metz" <spam...@library.lspace.org.invalid> wrote in

message news:502def01$2$fuzhry+tra$mr2...@news.patriot.net...
> In <k0k0br$uak$1...@dont-email.me>, on 08/16/2012

> at 06:39 PM, "Charles Richmond" <nume...@aquaporin4.com> said:
>
>>ISTM that a "dataset" is IBM-ese for a disk file
>

> No. See, e.g., IBlM System/360 Operating System Supervisor and Data
> Management Services, C28-6646-0.

>
>>And I think "DASD" means hard disk storage.
>

> No. It means direct access storage device, not just disk.

>
>>IBM has their own vocabulary for things
>

> So does every other vendor.

I contend that IBM had *more* of a specialized vocabulary... specific to IBM
systems. Other vendors of hardware had their special language, but much of
it was shared among the vendors. For example, most vendors called a "system
call"... a "system call". IBM seemed perverse to me in this... that *had*
to have their own word for almost everything!!!

>
>>what most computer systems refer to as "system calls", IBM has to
>>call "supervisor calls".
>

> First, it's "system call" that is the neologism, and second, a system
> call is a software artifact, not an instruction.
>

I did *not* intend to say or imply that a "supervisor call" was an
instruction. If I did, I appologize. And of course it is "system call" and
"supervisor call". This does *not* materially affect the point I was
making. DASD, by the time I was using computers in the late 1970's...
almost always meant hard disk storage, regardless of the broader meaning.

[glen herrmannsfeldt]

In comp.lang.pl1 Charles Richmond <nume...@aquaporin4.com> wrote:

(snip)

> I did *not* intend to say or imply that a "supervisor call" was an
> instruction. If I did, I appologize. And of course it is "system call" and
> "supervisor call". This does *not* materially affect the point I was
> making. DASD, by the time I was using computers in the late 1970's...
> almost always meant hard disk storage, regardless of the broader meaning.

I know at least one (two, actually) 2301s that were running in
the early 1980's.

The 2305 fixed head disk doesn't fit the usual definition of
hard disk storage, but is, in fact, a disk.

-- glen

[John W Kennedy]

On 2012-08-17 15:32:46 +0000, Walter Banks said:

> "Shmuel (Seymour J.) Metz" wrote:
>
>> In <PM0004C74...@users-ibook-g4-6.unknown.dom>, on 08/15/2012
>> at 12:36 PM, jmfbahciv <See....@aol.com> said:
>>
>>> Where did you find tame peripherals?
>>
>> In the early days the peripheral were dogs rather than cats. Card
>> jams, stacker jams, interlocks that didn't.
>

> The best card jam *incident* that I saw was a high speed third party
> card reader that used a grooved belt to pick and hold the card as it
> traveled around two 6 inch pulleys and back to a stacking tray. This
> thing was fast. The first pass read the cards and store the source on
> mag tape and then the following passes the source was read from
> the mag tape. .
>
> On the day in question the the cover on the reader was open I was
> one of many rubber necking this new and wonderful card reader
> installed for evaluation by the distributor. It slurped up decks of
> cards at great speed and then it happened. On the run between
> the pulleys something was lifting the card off the belt and turning
> them into high speed projectiles. That wouldn't have been so bad
> but the arm stabilizing the second pulley just happened to be in the
> way and neatly cut each card in half lengthwise. Confetti time.
>
> The operator aborted the job remembering the information was
> on the intermediate mag tape in a desperate effort to yank success
> out of the jaws of failure. They checked the tape the file was there
> they could even tell which one it was. It was later when the tried
> to read the tape that they discovered that the program was
> tokenized by the compiler.
>
> Two days later I saw an unshaven grad student rekeying from an
> old marked up listing.
>
> The distributor did not make the sale.

RCA tried to sell us their video terminals to replace our IBM 2260s.
Now, granted that 2260s were already kind of antiquish when they came
out (they used sonic delay lines to store pixels), and the RCA
character generation was gorgeous (it used some kind of character
stencil in the beam), but when I saw that /all/ the trim pots were in
front, I knew they was a bad idea. Sure enough, while my software was
still in the pilot stage, one imploded for no particular reason.

(There were other problems. The cheap version of the hub used a
commutator, so that A) you couldn't tell which terminal had sent the
incoming query, and B) the CPU /had/ to respond within a certain time
frame or the response would go to the wrong terminal; the entire access
method (so to speak) /had/ to be expanded as a macro of thousands of
lines within user code; it did a wait spin on the ECBs; and the "access
method" code altered ECB completion codes in the main thread (and not
under the IRB), so that there were windows when the completion code
wouldn't yet have the documented value; and some of these ex-post-facto
"completion" codes didn't have the X'40000000' [event complete] bit
set. Oh, and the "access method" code wasn't reentrant, either.)

--
John W Kennedy
"The poor have sometimes objected to being governed badly; the rich
have always objected to being governed at all."
-- G. K. Chesterton. "The Man Who Was Thursday"

[Charles Richmond]

"glen herrmannsfeldt" <g...@ugcs.caltech.edu> wrote in message
news:k0luml$p21$2...@speranza.aioe.org...

Indeed, there was a time when PL/I was IBM's "language of choice". And PL/I
did use the "/*" and "*/" as comment delimeters. So it would seem that C is
*not* the real culprit here...

[Charles Richmond]

<hanc...@bbs.cpcn.com> wrote in message
news:43d6ef24-ab9b-4ea4...@z9g2000yqf.googlegroups.com...

On Aug 16, 6:17 pm, Gene Wirchenko <ge...@ocis.net> wrote:

>> >Start by describing the computer as a big dumb room with in-box,
>> >typewriter, adding machine, file cabinets, mailboxes, and a boss who
>>
>> These days, it might be that you have to describe what a
>> typewriter is in terms of a computer.
>

>Yes, the analogies have to be changed. Of course, young people have
>computer experience already. They have to be weaned from video games
>to doing real work.
>

<funny story>

Little Kid: Dad, what is radio???

Dad: Well, radio is kind of like TV, without the pictures.

Little Kid: Huh!!! What will they think of next???

</funny story>

[Charles Richmond]

<hanc...@bbs.cpcn.com> wrote in message
news:144466c8-4b43-4538...@y1g2000yqc.googlegroups.com...

On Aug 16, 6:28 pm, Ben Pfaff <b...@cs.stanford.edu> wrote:

>> Adding machine, too. I've never seen one outside of a museum.
>>

>Maybe the classic adding machines, but kids do use calculators.

I grew up in a small town/country region. In the general store, the owner
had an antique cash register. He had an electro-mechanical calculator he
used to figure up the bill... and then just used the cash register to
"deposit" the final amount.

When I was a really little kid, I thought the adding machine was just a
typewriter that only typed numbers. The idea that a machine could actually
do arithmetic seemed impossible to my young mind!!!

[Charles Richmond]

"glen herrmannsfeldt" <g...@ugcs.caltech.edu> wrote in message

news:k0khn0$lf6$2...@speranza.aioe.org...
> In comp.lang.pl1 Ben Pfaff <b...@cs.stanford.edu> wrote:
>
> (snip)
>
>> Calculators aren't out of place yet, but soon. My coworkers
>> think it is very strange that I keep a calculator on my desk (an
>> HP RPN calculator, at that, but that's not why they think it's
>> strange).
>
> I noticed not so long ago that the HP28C is on the list
> of graphing calculators allowed for the SAT.
>
> Old enough to be in a museum, but still allowed.
>

Today, for around $10 US, you can buy a "scientific calculator" that uses
less power and is better... than the scientific calculator that cost a
couple of hundred dollars back in the 1970's. Back in the 1960's, if you
told someone that there would be calculators that *not* only would multiply
and divide... but would do logrithms and calculate trig functions like sine
and cosine... they would be sure you were "off your rocker"!!!

[Shmuel Metz]

In <PM0004C77...@ac83b7da.ipt.aol.com>, on 08/17/2012

at 01:27 PM, jmfbahciv <See....@aol.com> said:

>There was an operational art form which would diminish card jams to
>almost none.

CRBE? <g, d & r>

Even with good humidity control and proper jiggling of the cards, jams
were common on everything from the 407 to the 2540. Then there was
what we affectionately referred to as the downtime card reader on the
RCA 3301 :-(

--
Shmuel (Seymour J.) Metz, SysProg and JOAT <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action. I reserve the
right to publicly post or ridicule any abusive E-mail. Reply to
domain Patriot dot net user shmuel+news to contact me. Do not
reply to spam...@library.lspace.org

[Shmuel Metz]

In <20120817.08...@twoplaces.co.uk>, on 08/17/2012
at 09:11 AM, si...@twoplaces.co.uk (Simon Turner) said:

>The Delightful Rod is far from the only person who posts to afc via
>news.individual.net,

Still, if they can't or won't control the morphing of their trolls
their trolls, it's tempting. Think of it as an extension of the UIP.

>Much better to quietly killfile all Rod's assorted sock puppets, and
>add new ones each time they pop up.

If they pop up from the same provider, it's easier to just add an
entry from the provider one time.

[Shmuel Metz]

In <icehn5b...@home.home>, on 08/17/2012
at 11:44 AM, Dan Espen <des...@verizon.net> said:

>Lacking some kind of disciplined build system, it makes sense to
>include JCL in the same dataset as the source.

Not when you have multiple jobs with different compile options. Not
when the editor sets defaults based on the low qualifier of the DSN.

[Charlie Gibbs]

In article <k0makj$upe$1...@dont-email.me>, nume...@aquaporin4.com

(Charles Richmond) writes:

> "Shmuel (Seymour J.) Metz" <spam...@library.lspace.org.invalid>
> wrote in message news:502def01$2$fuzhry+tra$mr2...@news.patriot.net...
>
>> In <k0k0br$uak$1...@dont-email.me>, on 08/16/2012
>> at 06:39 PM, "Charles Richmond" <nume...@aquaporin4.com> said:
>>
>>> ISTM that a "dataset" is IBM-ese for a disk file
>>
>> No. See, e.g., IBlM System/360 Operating System Supervisor and Data
>> Management Services, C28-6646-0.
>>
>>> And I think "DASD" means hard disk storage.
>>
>> No. It means direct access storage device, not just disk.
>>
>>> IBM has their own vocabulary for things
>>
>> So does every other vendor.
>
> I contend that IBM had *more* of a specialized vocabulary... specific
> to IBM systems. Other vendors of hardware had their special language,
> but much of it was shared among the vendors. For example, most
> vendors called a "system call"... a "system call". IBM seemed
> perverse to me in this... that *had* to have their own word for
> almost everything!!!

Yes, while much of it might be explained as "independent creation",
IBM's terminology had a definite NIH feel to it. "Pel" vs. "pixel",
"alphameric" vs. "alphanumeric", "IPL" vs. "boot"... their CEs even
said "HDI" (head-disk interference) instead of "head crash".

A glossary translating IBMese into English would probably be
considerably larger than its counterparts for other vendors.
If I were sufficiently paranoid I might even suggest that
IBM did this in an attempt to confine its users to its own
portion of computing's Tower of Babel.

--
/~\ cgi...@kltpzyxm.invalid (Charlie Gibbs)
\ / I'm really at ac.dekanfrus if you read it the right way.
X Top-posted messages will probably be ignored. See RFC1855.
/ \ HTML will DEFINITELY be ignored. Join the ASCII ribbon campaign!

[Shmuel Metz]

In <k0mdej$fqb$1...@dont-email.me>, on 08/17/2012

at 04:35 PM, "Charles Richmond" <nume...@aquaporin4.com> said:

>Back in the 1960's, if you told someone that there would be
>calculators that *not* only would multiply and divide... but
>would do logrithms and calculate trig functions like sine and
>cosine... they would be sure you were "off your rocker"!!!

Make no bones about it, I had a calculator like that in the 1960's,
and so did most of my classmates.

[Shmuel Metz]

In <k0makj$upe$1...@dont-email.me>, on 08/17/2012

at 03:47 PM, "Charles Richmond" <nume...@aquaporin4.com> said:

>I contend that IBM had *more* of a specialized vocabulary... specific
>to IBM systems. Other vendors of hardware had their special
>language, but much of it was shared among the vendors. For example,
>most vendors called a "system call"... a "system call".

No. In fact, at the time IBM introduced the S/360, no vendor called it
a system call. It was enough to give one the screaming MME's.

>DASD, by the time I was using computers in the late 1970's...
>almost always meant hard disk storage, regardless of the broader
>meaning.

I take it you didn't have a 3850.

[Shmuel Metz]

In <k0lnrr$9nn$1...@speranza.aioe.org>, on 08/17/2012
at 03:26 PM, glen herrmannsfeldt <g...@ugcs.caltech.edu> said:

>For one, it takes up a lot less space. To read directly, it has to
>be, or at least had to be, FB 80. (The Fortran compilers I remember
>required FB 80.)

VB worked fine for PL/I.

>I did use TSO for one summer, and I believe that I did it the same
>way as with WYLBUR. As well as I remember, though, they were FB 80
>so the reason above doesn't apply.

TSO EDIT has a table of characteristics by data set type. When you
keep your source code in the appropriate PDS, you automatically get
the right settings, which are not always the ones for JCL.

[Anne & Lynn Wheeler]

glen herrmannsfeldt <g...@ugcs.caltech.edu> writes:
> I know at least one (two, actually) 2301s that were running in
> the early 1980's.
>
> The 2305 fixed head disk doesn't fit the usual definition of
> hard disk storage, but is, in fact, a disk.

re:
http://www.garlic.com/~lynn/2012k.html#85 1132 printer history
http://www.garlic.com/~lynn/2012k.html#88 1132 printer history


most of the cp67 360/67 timesharing systems ran with one or more 2301s
(drums).

there was also 2303 drum ... i never saw any ... similar to 2301 drum.
2303 drum transferred one r/w head for approx. 300kbyte/sec. 2301 drum
transferred four r/w heads simulataneously for approx 1.2mbyte/sec.
(each 2301 "track" data capacity was four times that of 2303).

cp67 used the tss/360 "page" format for 2301 ... nine 4kbyte pages
formated over a pair 2301 "tracks" (i.e. fifth record spanned the end of
first track and start of the following track).

when cp67 was first delivered to univ last week of jan1968 ... all i/o
operations FIFO single transfers at the time. 2301 throughput would peak
approx. 80 page-transfers/sec (each 2301 page transfer would have an
avg. of half-rotational delay). 2314 disks would get degrade quickly

I implemented ordered seek queuing for disk requests ... and would chain
together multiple page requests in single transfer in rotational
optimized order. I could get close to 9 4k page transfers per two
revoluations ... @ 3600rpm/min, 60rps, close to 30*9=270 transfers/sec
(1.08mbyte/sec thruput) peak.

Ordered seek queuing change (in place of FIFO) was for all i/o requests
... both cp page i/o and virtual machine queued i/o requests ... adding
ordered seek queuing resulted in cp67/cms showing much more graceful
(user perceived) degradation as load increased.

--
virtualization experience starting Jan1968, online at home since Mar1970

[Shmuel Metz]

In <k0mbea$oga$1...@speranza.aioe.org>, on 08/17/2012

at 09:00 PM, glen herrmannsfeldt <g...@ugcs.caltech.edu> said:

>The 2305 fixed head disk doesn't fit the usual definition of hard
>disk storage, but is, in fact, a disk.

What usual definition? It has a stack of disks with magnetic coating
and R/W heads to record data on them. Would you call a UNIVAC FastRAND
drum a disk drive just because the heads moved?

[Peter Flass]

On 8/17/2012 11:35 AM, Walter Bushell wrote:
> In article <502cddd1$79$fuzhry+tra$mr2...@news.patriot.net>,

> Shmuel (Seymour J.) Metz <spam...@library.lspace.org.invalid>

> wrote:
>
>> In <icobmez...@home.home>, on 08/13/2012
>> at 03:34 PM, Dan Espen <des...@verizon.net> said:
>>
>>> An every day C program needs a DLM.
>>
>> By the time C was invented cards were on their way out.

>
> And IIRC C was developed in a very non card oriented environment. Well
> if C had been developed anywhere near an IBM environment they would
> have chosen a different comment delimiter.
>

Probably not. PL/I used "/* ... */", first I believe, but other
languages used that also. Remember Denis came from a Multics
background, and I assume was quite familiar with PL/I.

--
Pete

[Peter Flass]

On 8/17/2012 11:44 AM, Dan Espen wrote:
> Shmuel (Seymour J.) Metz <spam...@library.lspace.org.invalid> writes:
>
>> In <k0khc3$kf1$1...@speranza.aioe.org>, on 08/17/2012
>> at 04:29 AM, glen herrmannsfeldt <g...@ugcs.caltech.edu> said:
>>
>>> In all those cases, you still need the /* even though the cards
>>> are virtual.
>>
>> If you're using an interactive editor then why would you want your
>> source code to be in an in-stream dataset rather than a member of a
>> separate PDS? If you're using TSO, then there's an even stronger
>> incentive to keep source separate from JCL.

>
> Lacking some kind of disciplined build system, it makes sense to
> include JCL in the same dataset as the source.
>

> That way you don't have to keep track of 2 things, Ie. the source
> code and the correct build JCL.
>
> Most places do have build systems, but there are always lose ends,
> that require special JCL or are used for testing.
>
> Anyway, point is, there are reasons, even if they aren't the best ones.
>
>

Usually the reasons are one or more of:
1. We've always done it that way.
2. It's quicker to set up.
3. That's the only way I know.

--
Pete

[Peter Flass]

On 8/17/2012 12:13 PM, Charlie Gibbs wrote:
> In article <8e354ccb-5bc0-499d...@wi5g2000pbc.googlegroups.com>,
> robin....@gmail.com (Robin Vowels) writes:
>
>> On Aug 17, 2:57 am, "Charlie Gibbs" <cgi...@kltpzyxm.invalid> wrote:
>>
>>> In article
>>> <8e3f3caa-bd37-4746-915c-b4d588736...@d4g2000yqb.googlegroups.com>,

>>>
>>> hanco...@bbs.cpcn.com (hancock4) writes:
>>>
>>>> On Aug 16, 8:47 am, jmfbahciv <See.ab...@aol.com> wrote:
>>>>

>>>>> I prefer working with cards than paper tape.
>>>
>>>> Agreed, because cards are random access and easier to sight read.
>>>
>>> Tape is just as easy to read - you just learn different patterns
>>> of holes.
>>>
>>>> Card readers tended to take in a program faster than a 10 char/sec
>>>> Teletype could do.
>>>
>>> But not as fast as a dedicated tape reader that does 2000
>>> chars./second. :-)
>>
>> Where was that?
>> Our S/350 had a paper tape reader. Delivered without spooling,
>> it ran at 1,000 cps.
>> When spooling was subsequently added, its maximum speed was only
>> 500 cps.
>>
>> On the other hand, the English Electric paper tape reader (early
>> 1960 or late 1950s) ran at 1,000 cps.
>
> Our 9300 had a Regnecentralen RC-2000 paper tape reader interfaced
> to it. This was a photocell-based unit with a 256-byte internal
> buffer; read requests were filled from this buffer, which the
> reader tried to keep about half full by controlling the speed
> of the servo motor that drove the capstan. There were no reels -
> you placed a roll of tape into a compartment, fed it under the
> read head, and the tape was ejected into a large bin, from which
> you rewound it by hand.
>
> Most of the time we were listing short records on the (non-spooled)
> printer, which limited the demand on the tape reader. However, I
> did write one application which read everything to disk, and the
> reader cranked up to its rated 200 inches per second. Prolonged
> operation at this speed generated strong enough static charges that
> the occasional spark would induce a glitch in the channel interface,
> crashing the computer. To get the application to run, we hung a
> grounded string of paper clips such that the tape would touch it
> when being discharged from the reader, and boiled a kettle in the
> machine room to bring up the humidity.
>

What? You didn't put a timer in the program to stop the reader every
little while?

--
Pete

[Peter Flass]

On 8/17/2012 2:01 PM, Nico de Jong wrote:
>
> "Ahem A Rivet's Shot" <ste...@eircom.net> skrev i en meddelelse
> news:20120817155415....@eircom.net...
>> On 17 Aug 2012 13:27:35 GMT
>> jmfbahciv <See....@aol.com> wrote:

>>
>>> Shmuel (Seymour J.) Metz wrote:
>>> > In <PM0004C74...@users-ibook-g4-6.unknown.dom>, on 08/15/2012
>>> > at 12:36 PM, jmfbahciv <See....@aol.com> said:
>>> >
>>> >>Where did you find tame peripherals?
>>> >
>>> > In the early days the peripheral were dogs rather than cats. Card
>>> > jams, stacker jams, interlocks that didn't.
>>> >

>>> There was an operational art form which would diminish card jams
>>> to almost none.
>>

> My worst experience with cards was with an IBM 2540. When reading or
> punching a card, you could direct the card to any of 3 pockets. The
> middle one (there are 5 pockets) could be used for read cards, and for
> punched card.
> Now then, when doing read and punch jobs simulteaniously, the operator
> was supposed to remember if bots jobs used the middle pocket. I once
> saw, that he hadnt remembered. From that day, operating both halves of
> the 2540 at the same time was banned
> /Nico
>

I was going to say I didn't think this was possible, but the reader and
punch were independent devices - 00C and 00D as I recall. I don't know
what HASP would have done with this, but then you couldn't
stacker-select input cards with HASP (I believe).

--
Pete

[Peter Flass]

On 8/17/2012 3:29 PM, Walter Bushell wrote:
>
> Much of the 1401 work was glorified tab shop work and many times the
> sorts were done on tab equipment.
>

Some (many?) 1401s were tape-only. Given the choice I might opt for
sorting the cards too. In addition, most likely the jobs were
originally all-tab jobs, and the 1401 had been inserted in place of an
accounting machine or some-such.


--
Pete

[Joe Morris]

<hanc...@bbs.cpcn.com> wrote:
>On Aug 16, 9:12 pm, Robin Vowels <robin.vow...@gmail.com> wrote:
>
>> One day, our S/360 model 50 halted with an over-run error.
>> Turned out that this message only happened when some 100 re-tries
>> at reading a disk track had failed.
>> IBM's remedy was to increase the number of re-tries to around 200.

>> The error occurs when too many DMAs take place simultaneously,
>> and there are no spare memory cycles available to handle a disk
>> transfer.

>That sounds strange. I can't help but suspect there was a physical or
>programming problem; that many retries is an awful lot.

It would help to know which OS was being used, and (if OS/360) if memory was
so tight that SER0 was selected for machine error recovery: that would
result in a disabled wait state following a machine or channel check.

But I don't see the link between ERP and channel overload. If I'm reading
the situation right this is one I/O operation (or at least I/O operatinons
on one data stream on disk ("a disk track had failed"). That should have
put the operation in to a loop:

* Issue DASD read CCW chain (Seek/search/TIC *-8/ReadData)
* Get an error
* make several attempts to read the data, including
numerous head-shake steps but get an I/O error
* Loop until loop limit hit

...but all of the I/O operations would have been sequential, with each one
ending the channel operation before the next one can be started. Even if
somehow the channel could be overloaded, that's a defined failure
("overrun") which raises Unit Check with the appropriate sense bits and
isn't considered to be either a CPU or channel failure.

Remember that the S/360 line (with the exception of the /85) had no block
mux channels, so any disk drive must have been on a selector (meaning that
the channel cannot support more than one active I/O request); this would
mean that the ERP logic shouldn't have been able to generate enough of a
load to bother the CPU, even though the /50 had internal channels. This
situation sounds more like there was a hardware problem either in the /50
itself or the control unit for whatever DASD was involved (probably either a
2841 or the CU portion of a 2314).

Joe

[Anne & Lynn Wheeler]

Peter Flass <Peter...@Yahoo.com> writes:
> I was going to say I didn't think this was possible, but the reader
> and punch were independent devices - 00C and 00D as I recall. I don't
> know what HASP would have done with this, but then you couldn't
> stacker-select input cards with HASP (I believe).

HASP drove the physical reader and punch ... if the reader was ready it
read cards and if it had cards to punch (all this asynchronously with
application execution & spooling) ... both cases it sent cards into
their respective stacker 1.

i did a 2540 driver for the univ class registration ... that read cards
into (middle/central, shared) stacker 3 and optionally punched cards
into same stacker ... but had to be done w/o hasp. the registration
cards (were all plain manilla) were read and went into stacker 3 ... if
a problem was found, a blank card was punched behind it (the cards in
the punch had colored stripe across the top). after processing with all
class registration cards back into the card trays ... "problem"
registration was easily identified by the following colored card.

[Anne & Lynn Wheeler]

an issue was multiple channels with multiple 2314s operating
concurrently on different channels with the channels having to
share/interleave the memory bus

cp67 would have periodic problems. to simulate virtual machine channel
programs it created a copy of the channel program substituting real
addresses (in each ccw) for the virtual addresses.

channel architecture precluded prefetching CCWs ... so all processing
for the previous CCW had to have completed before the next CCW could be
fetched. cp67 virtual machine i/o had an issue when the virtual channel
program specified contiguous virtual area that happened to cross page
boundary and the pages were at non-contiguous real locations. To handle
the situation, cp67 had to add one or more additional "data chained"
CCWs that specified the non-contiguous real addresses ... in place of a
single virtual CCW. There were times that the virtual operating system
could work fine on real machine ... but when running virtually, the
timing constraints for the channel program was that it couldn't fetch
the data-chained CCW with the next (non-contiguous) real address before
it had to start shipping data streaming in from the disk/drum to memory.

370 solved the problem with IDALs (indirect data access list) ... an
IDAL ccw didn't have the data memory address ... but pointed to a list
of memory addresses (for scatter/gather, non-contiguous i/o). While
channel architecture prevented channel from prefetching next CCW until
the processing of the previous CCW was complete ... the channel was
allowed to "prefetch" addresses from IDALs (eliminating the timing
problem of not having addresses reading for non-contiguous i/o
transfers).

recent post discussing huge increase in number of channels for 3090
http://www.garlic.com/~lynn/2012k.html#80 360/20, was 1132 printer history

360 green card didn't have sense data ... but the 360/67 "blue card" did
included sense data (for some devices) as well as description of
additional 360/67 features. this is q&d conversion of internal gcard
ios3270 to HTML
http://www.garlic.com/~lynn/gcard.html

I had added sense data from 360/67 blue card to the original gcard
ios3270 ... including 3380 & A220 sense data not on original blue card
http://www.garlic.com/~lynn/gcard.html#17

unit check, sense byte0, bit5 "overrun".

for other drift

http://www.garlic.com/~lynn/2012k.html#88 1132 printer history

getting dragged into escalation betwee pok channel engineers and 3880
control unit engineers over unsolicated/asynchronous unit checks.

[Joe Morris]

"Shmuel (Seymour J.) Metz" <spam...@library.lspace.org.invalid> wrote:
> Anne & Lynn Wheeler <ly...@garlic.com> said:

>>DASD is for direct access storage device ... generic term, it
>>predates hard disks dominating the technology ... dates from when
>>"drums" were more common ... drums even being used for main
>>processor storage.

> By the time IBM coined the term DASD, no new machines were using drums
> as main storage. Disks were more common by then, but drums had better
> performance.

> OTOH, nobody ever accused the noodle picker of having better
> performance :-(

Nor, at times when it displayed "noodle-stuffer" behavior, did anyone accuse
it of having *any* performance.

Speaking of the 2321 Data Cell:

Is there anyone here who lives in the Denver area, and can say if there is
still a "Traildust" steak house near Arapahoe? Many years ago SHARE held
its meeting in Denver and fifty of the attendees went to the Traildust for
dinner, not least because one of the attendees (a sysprog for United
Airlines) had a gig there. The Traildust was one of those places that cuts
off the tie of any man who wears one into the restaurant, so I donned a data
cell strip as a tie, then surrendered it to the waitress. If the place
still exists, is there a data cell strip stapled to the wall?

One gimmick at the Traildust: it had a 50-oz. cut of beef (named the
"Bullshipper"); if you ordered it and finished eating all of it in one hour
it was free...no, I didn't try.

Joe

[Dave Garland]

On 8/17/2012 4:35 PM, Charles Richmond wrote:

> Today, for around $10 US, you can buy a "scientific calculator" that
> uses less power and is better... than the scientific calculator that
> cost a couple of hundred dollars back in the 1970's.

$10? I bought one at the dollar store last year. Not because I have
any particular need for such a calculator any more (if I did have such
a problem, I'd set it up in a spreadsheet, easier to debug and fix
stupid errors) but just because I couldn't afford one back in the day.
If 1970 comes around again, I'll be ready.

I suppose I'll add it to my slide rule collection for contrast.

[Robin Vowels]

On Aug 18, 7:35 am, "Charles Richmond" <numer...@aquaporin4.com>
wrote:

> Today, for around $10 US, you can buy a "scientific calculator" that uses
> less power and is better... than the scientific calculator that cost a
> couple of hundred dollars back in the 1970's.  Back in the 1960's, if you
> told someone that there would be calculators that *not* only would multiply
> and divide... but would do logrithms and calculate trig functions like sine
> and cosine... they would be sure you were "off your rocker"!!!

Calculators like that have been available since the 1800s.
I still use one of them from time to time.

They are called slide rules.

[Robin Vowels]

On Aug 17, 2:28 pm, Ben Pfaff <b...@cs.stanford.edu> wrote:

> hanco...@bbs.cpcn.com writes:
> > On Aug 16, 6:28 pm, Ben Pfaff <b...@cs.stanford.edu> wrote:
>
> >> Adding machine, too.  I've never seen one outside of a museum.
>
> > Maybe the classic adding machines, but kids do use calculators.
>

> Calculators aren't out of place yet, but soon.

Don't count on it. [no pun intended]

The abacus is still with us!

[Walter Bushell]

In article <k0mjh0$l8u$1...@dont-email.me>,

Ok, how did they get the cards onto tape?

--
This space unintentionally left blank.

[hanc...@bbs.cpcn.com]

On Aug 17, 3:29 pm, Walter Bushell <pr...@panix.com> wrote:

> Much of the 1401 work was glorified tab shop work and many times the
> sorts were done on tab equipment.

Yes, sorts were done off-line on the sorter. Apparently a 1,000 card
a minute sorter was faster than a 16k memory. (I wish I knew what our
sorter rented for in the mid 1970s. I remember the C/E was an elderly
man.)

Indeed, I think the 1401 was designed as a natural outgrowth of the
407. RPG was developed to mimic the wiring of a plugboard.

But a full sized 1401 with disks (16k, 1311's) could handle
sophisticated processing, such as all the accounting for a medium
sized hospital. I heard that the programs called in subroutines from
disk to get around the 16k limitation.

[hanc...@bbs.cpcn.com]

On Aug 17, 2:32 pm, Walter Bushell <pr...@panix.com> wrote:

> On my first summer job I spent two weeks doing a least squares fit on
> a square root Friden calculator. The poor slobs in economics did their
> calculations on calculators that made you do multiplication digit by
> digit.

By 1970 electronic desktop calcualtors were in use and were a world of
improvement over the Friden and Marchant electro-mechanical
calculators widely used in industry until then. They became fairly
sophisticated with several registers and those knowing math 'tricks'
could take shortcuts.

My college had a room full of them in the business school, but nobody
used them as pocket calculators had come out. Also, I think for some
work (eg present value) people used tables in books instead.

There was a steel mill that kept using the old machines because they
weren't depreciated yet. They had to pay for a retired serviceman to
keep coming out to fix the machines; finally he refused to come out
anymore and they went electronic.

I think the movie "The Apartment" opens with Jack Lemmon watching the
carriage of his desk calculator move along as he calculated insurance
premiums.

[Rod Speed]

Robin Vowels <robin....@gmail.com> wrote
> Ben Pfaff <b...@cs.stanford.edu> wrote
>> hanco...@bbs.cpcn.com wrote

>>> Ben Pfaff <b...@cs.stanford.edu> wrote

>>>> Adding machine, too. I've never seen one outside of a museum.

>>> Maybe the classic adding machines, but kids do use calculators.

>> Calculators aren't out of place yet, but soon.

> Don't count on it. [no pun intended]

> The abacus is still with us!

But only dinosaurs still use them.

[hanc...@bbs.cpcn.com]

On Aug 17, 5:31 pm, "Charles Richmond" <numer...@aquaporin4.com>
wrote:

> I grew up in a small town/country region.  In the general store, the owner
> had an antique cash register.  He had an electro-mechanical calculator he
> used to figure up the bill... and then just used the cash register to
> "deposit" the final amount.

FWIW, not long ago I've seen several restaurants with a 1970s style
NCR electro-mechanical cash register. The cashier added up the check
on a simple pocket calcualtor, then used the cash register to enter
the final amount, just as you described.

I don't know why they didn't use the cash register to add up the
amount, maybe they didn't know it could do it.

I was always impressed by cash registers, and wish I could have one.
I also liked the simple manual units often found at the front counter
of drugstores or luncheonettes. These did not add. The clerk merely
selected the combination of keys that came to the total amount and
pressed down; that amount appeared in the window (and presumably on a
tape.) There was no provision for sales tax, if collected, it was
deposited in a separate container.

[hanc...@bbs.cpcn.com]

On Aug 17, 4:47 pm, "Charles Richmond" <numer...@aquaporin4.com>
wrote:

>DASD, by the time I was using computers in the late 1970's...
> almost always meant hard disk storage, regardless of the broader meaning.

By the late 1970s they had a virtual disk, where storage was used to
imiitate a disk drive, and soon 8" floppies would be out. Then they
had the super tape/disk memory. All were DASD.

While the acronum 'DASD' sounds pretentious, it is accurate.

Admittedly, some other IBM phrases were a bit much, such as
"functionally stabilized".

[hanc...@bbs.cpcn.com]

Heck, some 1401s were card only--no tape or disk--and very much a tab
machine replacement. (I wonder how the rentals compared).

But I think the limitation on sorting cards was the 16k or less memory
the 1401 had. Also, it probably wasn't seen as a good use of computer
time as compared to using the cheaper sorter.

I saw a 1957 bid proposal for an IBM 705-III. Tightly integrated into
the plans were a full array of tab gear to do pre- and post-
processing offline from the computer.

[Dan Espen]

Since I started in 1964, many of my assignments were to take existing
tab machine jobs and redesign for the 14xx.

My first machines were pair, a 1401 and a 1460. Neither had a disk
drive but I recall 4 or more tape drives on each.

It was normal to expect the card input to be pre-sorted on a sorter.
I remember getting requirements that included punching various cards
for the next run (aka act like a tab machine). I did my best to
not do the job that way. Instead create a tape for subsequent
processing.

Some shops took a long time to finally get rid of all their tab
equipment.

I had one operations manager get really mad when I decided on my own
to eliminate a tab run. He even complained to the big boss.
Not that it did him any good. I just looked around for more tab runs
to kill off.

My second machine was a 1440 with no tape, but 2 1311 disk drives.
They were a lot more fun than tape, and definitely not a glorified
tab machine.

The rest of the 14xx's that I had to deal with were really S/360 30s and
40s running emulation. There we generally had tape and disk.
It took a long time for all that 14xx stuff to be re-written and some
companies were better than others at creating impossible barriers to
progress.


--
Dan Espen

[Ahem A Rivet's Shot]

On Fri, 17 Aug 2012 18:45:21 -0400

Shmuel (Seymour J.) Metz <spam...@library.lspace.org.invalid> wrote:

> In <k0mdej$fqb$1...@dont-email.me>, on 08/17/2012
> at 04:35 PM, "Charles Richmond" <nume...@aquaporin4.com> said:
>
> >Back in the 1960's, if you told someone that there would be
> >calculators that *not* only would multiply and divide... but
> >would do logrithms and calculate trig functions like sine and
> >cosine... they would be sure you were "off your rocker"!!!
>
> Make no bones about it, I had a calculator like that in the 1960's,
> and so did most of my classmates.

An enhanced set of Napier's bones no doubt. I still have mine.

--
Steve O'Hara-Smith | Directable Mirror Arrays
C:>WIN | A better way to focus the sun
The computer obeys and wins. | licences available see
You lose and Bill collects. | http://www.sohara.org/

[Michael Black]

It depends on how many functions.

I remember when someone I knew got an HP-35, there was a group buy where
he worked, and the calculator hadn't been out that long. Did it seem
fancy, things I didn't know how to do yet.

But that TI-30 I got as my first scientific calculator about 1976, it had
more functions. And that level of functions remained "basic" (unless you
went for a four function) for a long time. When I replaced the TI-30 with
an LCD calculator in the late eighties, it had a few more functions, and
maybe more important, could do conversions between
binary/hex/octal/deciaml, though no actual logic operations.

That level is still the cheap "scientific calculator" though a lot of them
do logic operations at no extra cost.

You can get those for under ten dollars, even at the regular price.

I bought a couple of calculators (the LCD one from the late eighties was
seeing wear) over the past few years, waiting till back to school supplies
put them on sale. I didn't need more functions, but did get them for
under ten dollars on sale. Those do logic operations. Plus the more
expensive of the two uses solar power, I'll never run out of battery.

Why did I buy two? Because they were so cheap on sale, I can keep one at
the desk, another one elsewhere where I might use it. In the days of the
TI-30, I sure wouldn't have done that.,

The trouble with a lot of the cheap scientific calculators are that they
remain large. I found a TI-25X in a pile of garbage with some other
interesting stuff a couple of years ago, but some of the segments don't
display. That one is nice and small, if it had worked, I'd use that as a
calculator to carry around.

Michael

[Michael Black]

I've seen a few cash registers waiting for the garbage over the years.
I've been tempted, but they are so big and likely heavy, any real
curiosity about them was dimmed by the effort of getting one home.

Michael

[Peter Flass]

As I recall you could stacker-select the punch output using the
appropriate control characters, but not the cards being read.

--
Pete

[Walter Bushell]

In article
<3a83185b-d4ec-48d8...@k20g2000vbk.googlegroups.com>,

And you could run a shop with a 1401, printer, card I/O and a sorter
and have the same job flow, but faster.

And the sorter was a O(digits to sort*number of cards) vs. O(number of
cards*log(number of cards)) time device. With 16K characters or was
that digits? for memory, you needed a bunch of tape drives for a tape
sort.

Or do you do a radix sort using tape drives?

[Robin Vowels]

On Aug 18, 12:46 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
> Robin Vowels <robin.vow...@gmail.com> wrote

If so, the dinosaurs are young ones.
The Chinese and/or Japanese still use the abacus,
and skilled users are quicker at it than those
using electronic calculators.

I recall an abacus behind a glass window at a computer site.
It read: "In emergency break glass".

[jmfbahciv]

Charles Richmond wrote:
> "Shmuel (Seymour J.) Metz" <spam...@library.lspace.org.invalid> wrote in
> message news:502def01$2$fuzhry+tra$mr2...@news.patriot.net...
>> In <k0k0br$uak$1...@dont-email.me>, on 08/16/2012
>> at 06:39 PM, "Charles Richmond" <nume...@aquaporin4.com> said:
>>
>>>ISTM that a "dataset" is IBM-ese for a disk file
>>
>> No. See, e.g., IBlM System/360 Operating System Supervisor and Data
>> Management Services, C28-6646-0.
>>
>>>And I think "DASD" means hard disk storage.
>>
>> No. It means direct access storage device, not just disk.
>>
>>>IBM has their own vocabulary for things
>>
>> So does every other vendor.
>
> I contend that IBM had *more* of a specialized vocabulary... specific to IBM
> systems. Other vendors of hardware had their special language,

DEC chose terms which could not attract IBM's notice. Law suit avoidance
was high on our list of word usage.

<snip>

/BAH

[jmfbahciv]

Shmuel (Seymour J.) Metz wrote:

> In <k0k0br$uak$1...@dont-email.me>, on 08/16/2012
> at 06:39 PM, "Charles Richmond" <nume...@aquaporin4.com> said:
>
>>ISTM that a "dataset" is IBM-ese for a disk file
>
> No. See, e.g., IBlM System/360 Operating System Supervisor and Data
> Management Services, C28-6646-0.
>
>>And I think "DASD" means hard disk storage.
>
> No. It means direct access storage device, not just disk.
>
>>IBM has their own vocabulary for things
>
> So does every other vendor.
>

>>what most computer systems refer to as "system calls", IBM has to
>>call "supervisor calls".
>
> First, it's "system call" that is the neologism, and second, a system
> call is a software artifact, not an instruction.
>
Not necessarily. The difference between TOPS-10's CALL UUOs and
the CALLI UUO may be an example.

/BAH

[jmfbahciv]

Ahem A Rivet's Shot wrote:
> On 17 Aug 2012 13:27:35 GMT
> jmfbahciv <See....@aol.com> wrote:
>

>> Shmuel (Seymour J.) Metz wrote:

>> > In <PM0004C74...@users-ibook-g4-6.unknown.dom>, on 08/15/2012
>> > at 12:36 PM, jmfbahciv <See....@aol.com> said:
>> >
>> >>Where did you find tame peripherals?
>> >
>> > In the early days the peripheral were dogs rather than cats. Card
>> > jams, stacker jams, interlocks that didn't.
>> >
>> There was an operational art form which would diminish card jams
>> to almost none.
>

> I take it this did not include carrying decks of cards around in
> jacket pockets, or bound with a rubber band and stuffed into a school bag -
> I certainly experienced far fewer jams on my decks after I was offered a
> drawer to keep my cards in.

A good operator [I don't know how to explain this] took handful of cards,
ruffled it in several different ways, knocked it on a hard surface and
felt the sides (edges) of that deck. If there was any worn card, you
could feel it. Pull the card and try to smooth the edge with two
fingernails. If it's really worn, you could go across the hall and
dup the card. Place cards on the stack and feel the entire card stack
with fingers for smoothness. Grab another handful of cards, repeat
process. Even a 1620 could hold about a box of cards in both hoppers.

/BAH

[Gene Wirchenko]

On Fri, 17 Aug 2012 20:44:35 -0400, "Joe Morris"
<j.c.m...@verizon.net> wrote:

[snip]

>Is there anyone here who lives in the Denver area, and can say if there is
>still a "Traildust" steak house near Arapahoe? Many years ago SHARE held

I live nowhere near there, but I decided to practice my Google
Fu. HI-YAH!:
http://culinary-colorado.com/2009/12/30/trail-dust-bites-the-dust/
Apparently, not anymore.

[snip]

Sincerely,

Gene Wirchenko

[Joe Morris]

And sometimes parts of them can be recycled. Back in the mid (?) 1970s
Sears Roebuck replaced all of those monster cash registers its stores used.
Anyone who ever saw one probably remembers them: free-standing, about 5'
high, with 10 or 20 separate cash drawers, and so many buttons and levers to
be operated by the clerk that a new hire must have needed a 6-week class on
how to ring up a sale. Sears auctioned the old units to get rid of them,
probably expecting them to broken up for scrap.

A good friend owned a fixed-base operator at one of the local airports and
bought a dozen or so of the cash registers. He threw away the "cash
register" parts but kept the drawers for use as parts bins. (Sadly, he
threw out the guts of the cash registers immediately after taking
possession, and they may not have even made it back to the airport; I never
saw them anywhere in or behind his hangars.)

Joe

[Joe Morris]

"jmfbahciv" <See....@aol.com> wrote:

> Charles Richmond wrote:

>> I contend that IBM had *more* of a specialized vocabulary... specific to
>> IBM
>> systems. Other vendors of hardware had their special language,

> DEC chose terms which could not attract IBM's notice. Law suit avoidance
> was high on our list of word usage.

IBM wasn't the only place that DEC wanted to not get interested in DEC
products: it included corporate management of its customers. That's why the
"PDP" name was chosen: the machines were "Programmed Data Processors" and
the word "computer" was never used...thus allowing groups within a
customer's organization to purchase a PDP-xxx and be able to say with a
straight face to corporate manaement that nobody had violated the company
policy that only the central computer center could purchase a "computer".

Of course, DEC was in no way the only company that did that; at my PPOE the
bookstore bought an "automated cash register system" that had a mini (make
and model I forget) as part of the architecture...but "it wasn't a
computer." It *was* a piece of crap, though, writing half-inch magnetic
tape with data records as small as one character...with the obvious
consequences whenever the tape had a read error...but the bookstore manager
insisted that there was no mandate to involve the computer center with the
purchase decision because "it wasn't a computer."

Joe

[John W Kennedy]

On 2012-08-17 21:35:11 +0000, Charles Richmond said:

> "glen herrmannsfeldt" <g...@ugcs.caltech.edu> wrote in message
> news:k0khn0$lf6$2...@speranza.aioe.org...
>> In comp.lang.pl1 Ben Pfaff <b...@cs.stanford.edu> wrote:
>>
>> (snip)
>>
>>> Calculators aren't out of place yet, but soon. My coworkers
>>> think it is very strange that I keep a calculator on my desk (an
>>> HP RPN calculator, at that, but that's not why they think it's
>>> strange).
>>
>> I noticed not so long ago that the HP28C is on the list
>> of graphing calculators allowed for the SAT.
>>
>> Old enough to be in a museum, but still allowed.

>>
>
> Today, for around $10 US, you can buy a "scientific calculator" that
> uses less power and is better... than the scientific calculator that

> cost a couple of hundred dollars back in the 1970's. Back in the

> 1960's, if you told someone that there would be calculators that *not*
> only would multiply and divide... but would do logrithms and calculate
> trig functions like sine and cosine... they would be sure you were "off
> your rocker"!!!

That's an overstatement. There were true electronic calculators by
1965, with a four-element stack and a memory, and a square-root
function. (Granted, they were the size of a typewriter and cost over
$10,000 pre-Nixon dollars, but they existed.) Transcendental functions
are certainly /imaginable/ by that point.

--
John W Kennedy
"...when you're trying to build a house of cards, the last thing you
should do is blow hard and wave your hands like a madman."
-- Rupert Goodwins

[John W Kennedy]

On 2012-08-18 03:50:27 +0000, Dan Espen said:
> My first machines were pair, a 1401 and a 1460. Neither had a disk
> drive but I recall 4 or more tape drives on each.

The capacity on either was six, and four was a minimum for general use.
(You might have fewer if you used tapes only for backup or for
transcription, but four were a minimum for sorting and for running the
Autocoder-on-Tape assembler.)

--
John W Kennedy
"The blind rulers of Logres
Nourished the land on a fallacy of rational virtue."
-- Charles Williams. "Taliessin through Logres: Prelude"

[John W Kennedy]

On 2012-08-17 23:36:35 +0000, Joe Morris said:
> Remember that the S/360 line (with the exception of the /85) had no
> block mux channels,

195, too.

--
John W Kennedy
Read the remains of Shakespeare's lost play, now annotated!
http://www.SKenSoftware.com/Double%20Falshood

[John W Kennedy]

On 2012-08-18 11:17:57 +0000, Walter Bushell said:

> In article
> <3a83185b-d4ec-48d8...@k20g2000vbk.googlegroups.com>,
> hanc...@bbs.cpcn.com wrote:
>
>> On Aug 17, 3:29 pm, Walter Bushell <pr...@panix.com> wrote:
>>
>>> Much of the 1401 work was glorified tab shop work and many times the
>>> sorts were done on tab equipment.
>>
>> Yes, sorts were done off-line on the sorter. Apparently a 1,000 card
>> a minute sorter was faster than a 16k memory. (I wish I knew what our
>> sorter rented for in the mid 1970s. I remember the C/E was an elderly
>> man.)
>>
>> Indeed, I think the 1401 was designed as a natural outgrowth of the
>> 407. RPG was developed to mimic the wiring of a plugboard.
>>
>> But a full sized 1401 with disks (16k, 1311's) could handle
>> sophisticated processing, such as all the accounting for a medium
>> sized hospital. I heard that the programs called in subroutines from
>> disk to get around the 16k limitation.
>
> And you could run a shop with a 1401, printer, card I/O and a sorter
> and have the same job flow, but faster.
>
> And the sorter was a O(digits to sort*number of cards) vs. O(number of
> cards*log(number of cards)) time device.

...but note that, in most sensible situations, digits to sort is
roughly proportional to log(number of cards).

> With 16K characters or was
> that digits? for memory, you needed a bunch of tape drives for a tape
> sort.

16,000 (actually 15,998, because positions 000 and 100 carried
undocumented volatile state information for the card reader and card
punch). The basic 1401 frame held 4,000. If you needed more than that,
you got a 1406 expansion unit.

> Or do you do a radix sort using tape drives?

There are three usual tape sorts. The basic principle of all three is
to create "strings" of sorted records in main memory and to write them
to tape, then to merge the strings together into longer and longer
strings, until a final merge can be done onto the ultimate output tape.
Disk sorts use the same idea, though they are more complicated.

--
John W Kennedy
"Information is light. Information, in itself, about anything, is light."
-- Tom Stoppard. "Night and Day"

[Ahem A Rivet's Shot]

On 18 Aug 2012 14:02:06 GMT

Yeah, I've seen it done. I never got that good with cards. It was
just about when cards were becoming completely obsolete that I saw a card
reader that could handle a card that had been crumpled and roughly
straightened. If that reader had been invented twenty (or even ten) years
earlier it would no doubt have saved thousands of hours with the card saw
and retyping.

[Rod Speed]

jmfbahciv <See....@aol.com> wrote

> Charles Richmond wrote
>> Shmuel (Seymour J.) Metz <spam...@library.lspace.org.invalid> wrote

>>> Charles Richmond <nume...@aquaporin4.com> wrote

>>>> ISTM that a "dataset" is IBM-ese for a disk file

>>> No. See, e.g., IBlM System/360 Operating System Supervisor
>>> and Data Management Services, C28-6646-0.

>>>> And I think "DASD" means hard disk storage.

>>> No. It means direct access storage device, not just disk.

>>>> IBM has their own vocabulary for things

>>> So does every other vendor.

>> I contend that IBM had *more* of a specialized vocabulary... specific to
>> IBM systems. Other vendors of hardware had their special language,

> DEC chose terms which could not attract IBM's notice.
> Law suit avoidance was high on our list of word usage.

Have fun listing even a SINGLE example of a law suit over the use of a word
like that.

[Rod Speed]

jmfbahciv <See....@aol.com> wrote

> Ahem A Rivet's Shot wrote

>> jmfbahciv <See....@aol.com> wrote
>>> Shmuel (Seymour J.) Metz wrote

>>>> jmfbahciv <See....@aol.com> wrote

>>>>> Where did you find tame peripherals?

>>>> In the early days the peripheral were dogs rather than cats.
>>>> Card jams, stacker jams, interlocks that didn't.

>>> There was an operational art form which would diminish card jams to
>>> almost none.

We didn’t bother and still got almost no card jams at all.

>> I take it this did not include carrying decks of cards around in
>> jacket pockets, or bound with a rubber band and stuffed into
>> a school bag - I certainly experienced far fewer jams on my
>> decks after I was offered a drawer to keep my cards in.

> A good operator [I don't know how to explain this] took handful of
> cards, ruffled it in several different ways, knocked it on a hard surface

That wasn’t necessary with decks manually punched on an 026 or 029

[Ahem A Rivet's Shot]

On Sat, 18 Aug 2012 06:27:19 -0700 (PDT)
Robin Vowels <robin....@gmail.com> wrote:

> On Aug 18, 12:46 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
> > Robin Vowels <robin.vow...@gmail.com> wrote
> >
> > > Ben Pfaff <b...@cs.stanford.edu> wrote
> > >> hanco...@bbs.cpcn.com wrote
> > >>> Ben Pfaff <b...@cs.stanford.edu> wrote
> > >>>> Adding machine, too.  I've never seen one outside of a museum.
> > >>> Maybe the classic adding machines, but kids do use calculators.
> > >> Calculators aren't out of place yet, but soon.
> > > Don't count on it. [no pun intended]
> > > The abacus is still with us!
> >
> > But only dinosaurs still use them.
>
> If so, the dinosaurs are young ones.
> The Chinese and/or Japanese still use the abacus,
> and skilled users are quicker at it than those
> using electronic calculators.

A slide rule in skilled hands is also faster and more accurate than
a calculator.

[Rod Speed]

Robin Vowels <robin....@gmail.com> wrote

> Rod Speed <rod.speed....@gmail.com> wrote
>> Robin Vowels <robin.vow...@gmail.com> wrote
>>> Ben Pfaff <b...@cs.stanford.edu> wrote
>>>> hanco...@bbs.cpcn.com wrote
>>>>> Ben Pfaff <b...@cs.stanford.edu> wrote

>>>>>> Adding machine, too. I've never seen one outside of a museum.

>>>>> Maybe the classic adding machines, but kids do use calculators.

>>>> Calculators aren't out of place yet, but soon.

>>> Don't count on it. [no pun intended]
>>> The abacus is still with us!

>> But only dinosaurs still use them.

> If so, the dinosaurs are young ones.

Nope, no young ones still use them.

> The Chinese and/or Japanese still use the abacus,

Only the dinosaurs.

> and skilled users are quicker at it than
> those using electronic calculators.

That's just plain wrong.

> I recall an abacus behind a glass window at a computer site.
> It read: "In emergency break glass".

Sure, but that's just a joke and no one bothered to do that.

[Rod Speed]

Ahem A Rivet's Shot <ste...@eircom.net> wrote
> Robin Vowels <robin....@gmail.com> wrote

>> Rod Speed <rod.speed....@gmail.com> wrote
>>> Robin Vowels <robin.vow...@gmail.com> wrote
>>>> Ben Pfaff <b...@cs.stanford.edu> wrote
>>>>> hanco...@bbs.cpcn.com wrote
>>>>>> Ben Pfaff <b...@cs.stanford.edu> wrote

>>>>>>> Adding machine, too. I've never seen one outside of a museum.

>>>>>> Maybe the classic adding machines, but kids do use calculators.

>>>>> Calculators aren't out of place yet, but soon.

>>>> Don't count on it. [no pun intended]
>>>> The abacus is still with us!

>>> But only dinosaurs still use them.

>> If so, the dinosaurs are young ones.
>> The Chinese and/or Japanese still use the abacus,
>> and skilled users are quicker at it than those
>> using electronic calculators.

> A slide rule in skilled hands is also faster
> and more accurate than a calculator.

Oh bullshit.

[Dan Espen]

My experience with card readers is that they weren't all that sensitive
and jammed rarely. Any card handler worth his salt would joggle the
cards so they were all aligned before reaching the picker knives. No
big deal.

I recall we had to sort some bent/stapled and
generally mangled cards. Our IBM 083 wasn't going to do it.

Fortunately our parent company had an IBM 084 card sorter that could sort almost
anything. It ran at twice the speed I'd ever seen before, 2000 CPM.
Very impressive.

--
Dan Espen

[Joe Morris]

"Ahem A Rivet's Shot" <ste...@eircom.net> wrote:

> A slide rule in skilled hands is also faster and more accurate than
> a calculator.

The meaning of "accurate" depends on several criteria, including how many
decimal places are required.

However...while it may give you a faster respone delivering the digit
string, the slide rule tells you nothing about the location of the decimal
point in the answer. Rightly or wrongly, from the 1930s through the
mid-1960s, my father prohibited his students from using a slide rule for any
of their work. He did have a reason: he was teaching nuclear physics, and
(perhaps from a radiation accident he suffered around 1930 and his work on
the Manhattan Project) he gave little or no credit for an answer that had
the right steps and the right digit string, but the wrong decimal point
location.

Having said that...back when I was maybe in 7th grade I bought a 10"
circular slide rule. That instrument, with a C scale of about 30", carried
me through grad school - although I'll admit that in my last year the pocket
calculators (such as the HP35) started to make an appearance.

Joe

[Peter Flass]

The IBM OS/360 sort was a wonder, with lots of choices about sorting
technique. Possibly you could do a tape sort with as few as three
drives. One of the optimizations used was to write as long a string as
possible to a drive forward, and then read it back to merge it to avoid
having to rewind.

--
Pete

[Peter Flass]

On 8/18/2012 5:46 PM, Rod Speed wrote:
> jmfbahciv <See....@aol.com> wrote
>> Ahem A Rivet's Shot wrote
>>> jmfbahciv <See....@aol.com> wrote
>>>> Shmuel (Seymour J.) Metz wrote
>>>>> jmfbahciv <See....@aol.com> wrote
>
>>>>>> Where did you find tame peripherals?
>
>>>>> In the early days the peripheral were dogs rather than cats.
>>>>> Card jams, stacker jams, interlocks that didn't.
>
>>>> There was an operational art form which would diminish card jams to
>>>> almost none.
>
> We didn’t bother and still got almost no card jams at all.
>
>>> I take it this did not include carrying decks of cards around in
>>> jacket pockets, or bound with a rubber band and stuffed into
>>> a school bag - I certainly experienced far fewer jams on my
>>> decks after I was offered a drawer to keep my cards in.
>
>> A good operator [I don't know how to explain this] took handful of
>> cards, ruffled it in several different ways, knocked it on a hard surface
>
> That wasn’t necessary with decks manually punched on an 026 or 029

Maybe not the first time they were read, but card decks were often read
over and over, and gradually things tended to happen.

--
Pete

[Joe Pfeiffer]

Ahem A Rivet's Shot <ste...@eircom.net> writes:

> On Sat, 18 Aug 2012 06:27:19 -0700 (PDT)
> Robin Vowels <robin....@gmail.com> wrote:
>
>> On Aug 18, 12:46 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
>> > Robin Vowels <robin.vow...@gmail.com> wrote
>> >
>> > > Ben Pfaff <b...@cs.stanford.edu> wrote
>> > >> hanco...@bbs.cpcn.com wrote
>> > >>> Ben Pfaff <b...@cs.stanford.edu> wrote
>> > >>>> Adding machine, too.  I've never seen one outside of a museum.
>> > >>> Maybe the classic adding machines, but kids do use calculators.
>> > >> Calculators aren't out of place yet, but soon.
>> > > Don't count on it. [no pun intended]
>> > > The abacus is still with us!
>> >
>> > But only dinosaurs still use them.
>>
>> If so, the dinosaurs are young ones.
>> The Chinese and/or Japanese still use the abacus,
>> and skilled users are quicker at it than those
>> using electronic calculators.
>
> A slide rule in skilled hands is also faster and more accurate than
> a calculator.

I can believe "faster" -- but how can it possibly be more accurate?

[Rod Speed]

Dan Espen <des...@verizon.net> wrote
> Ahem A Rivet's Shot <ste...@eircom.net> wrote

>> jmfbahciv <See....@aol.com> wrote
>>> Ahem A Rivet's Shot wrote

>>>> jmfbahciv <See....@aol.com> wrote
>>>>> Shmuel (Seymour J.) Metz wrote

>>>>>> jmfbahciv <See....@aol.com> wrote

Yeah, me too. And I would have been the one unjamming the
reader if it had ever jammed and I might have done it just once
in two decades.

> Any card handler worth his salt would joggle the cards so they
> were all aligned before reaching the picker knives. No big deal.

The users mostly loaded the card reader themselves, and we
still go no card jams.

> I recall we had to sort some bent/stapled and generally mangled cards.

We never did use them for turnaround documents in the
field etc so didn't see the worst of that sort of thing.

[John W Kennedy]

And I've seen Victorian slide rules of ten feet or more in museums.
They were usually helical.

--
John W Kennedy
"The grand art mastered the thudding hammer of Thor
And the heart of our lord Taliessin determined the war."
-- Charles Williams. "Mount Badon"

[Rod Speed]

Peter Flass <Peter...@Yahoo.com> wrote

Sure, I didn’t say that very clearly. I meant that JUST that bit I separated
out deliberately from what Barb said wasn’t necessary with decks manually
punched on 026s and 029s. The other stuff that followed that comment
of mine is certainly something that can be needed over time with the
deck that had been read over and over.

[John W Kennedy]

Operator reliability, I suppose.

Of course, in real-world engineering, three decimal digits of precision
may be all you've got to begin with. (It is for the same reason that
the new version of IEEE floating point includes a 16-bit option, which
is implemented in the z/Architecture.)

--
John W Kennedy
"Give up vows and dogmas, and fixed things, and you may grow like That.
...you may come to think a blow bad, because it hurts, and not because
it humiliates. You may come to think murder wrong, because it is
violent, and not because it is unjust."
-- G. K. Chesterton. "The Ball and the Cross"

[John W Kennedy]

And later versions did every known funky disk optimization. (The IBM
sort got an undeserved bad rep from its rivals; when SyncSort was
pushing hard to sell us, I did some studies and discovered that
SyncSort's ballyhooed superiority was simply a question of better I/O
routines for SORTIN and SORTOUT. I told my bosses that it would be much
easier for IBM to fix that than for SyncSort to improve their actual
sorting algorithm. For once, they listened, and I was right; IBM
improved SORTIN and SORTOUT, and also did another redesign of the sort
algorithm; it /screamed/.)

[Shmuel Metz]

In <proto-CC8B0B....@news.panix.com>, on 08/18/2012

at 07:17 AM, Walter Bushell <pr...@panix.com> said:

>And the sorter was a O(digits to sort*number of cards) vs.
>O(number of cards*log(number of cards)) time device. With 16K
>characters or was that digits?

Characters: 6 bits plus wordmark plus parity.

>Or do you do a radix sort using tape drives?

I hope not.

--
Shmuel (Seymour J.) Metz, SysProg and JOAT <http://patriot.net/~shmuel>

Unsolicited bulk E-mail subject to legal action. I reserve the
right to publicly post or ridicule any abusive E-mail. Reply to
domain Patriot dot net user shmuel+news to contact me. Do not
reply to spam...@library.lspace.org

[Shmuel Metz]

In <k0mj6k$82c$4...@dont-email.me>, on 08/17/2012
at 07:13 PM, Peter Flass <Peter...@Yahoo.com> said:

>I was going to say I didn't think this was possible, but the reader
>and punch were independent devices - 00C and 00D as I recall.

The addresses were arbitrary, although those were the most common.
Shops tended to configure peripherals to addresses consistent with the
starter system for their OS. There were no self-identifying
peripherals in those days; IBM had not yet invented the idea from
Burroughs.

>I don't know what HASP would have done with this, but then you
>couldn't stacker-select input cards with HASP (I believe).

ASP, HASP and OS spooling allowed you to select the output stacker;
none of them selected the stacker on input. If you wanted to select
the stacker on input then you had to directly allocate the 2540.

//READER DD UNIT=00C
//PUNCH DD UNIT=00D

[Shmuel Metz]

In <Pine.LNX.4.64.12...@darkstar.example.net>, on
08/18/2012
at 01:11 AM, Michael Black <et...@ncf.ca> said:

>But that TI-30 I got as my first scientific calculator about 1976, it
>had more functions. And that level of functions remained "basic"
>(unless you went for a four function) for a long time. When I
>replaced the TI-30 with an LCD calculator in the late eighties, it
>had a few more functions, and maybe more important, could do
>conversions between binary/hex/octal/deciaml, though no actual logic
>operations.

TI already had a hex calculator in the 1970's. I don't recall the
exact date, but it was before 1976.

[Shmuel Metz]

In <k0mki...@news1.newsguy.com>, on 08/17/2012

at 07:36 PM, "Joe Morris" <j.c.m...@verizon.net> said:

>Remember that the S/360 line (with the exception of the /85) had
>no block mux channels,

What is the 360/195, chopped liver?

[Shmuel Metz]

In <PM0004C78...@ac8409ec.ipt.aol.com>, on 08/18/2012
at 02:02 PM, jmfbahciv <See....@aol.com> said:

>Not necessarily. The difference between TOPS-10's CALL UUOs and the
>CALLI UUO may be an example.

TOPS-10 is software.

[Shmuel Metz]

In <20120818224313....@eircom.net>, on 08/18/2012

at 10:43 PM, Ahem A Rivet's Shot <ste...@eircom.net> said:

>A slide rule in skilled hands is also faster and more accurate than a
>calculator.

No. For simple calculations it is faster, but it has a built-in
imprecision and has no memory. I would expect a skilled user to get
X^Y faster on a modern calculator than the slide rule, although I
would expect the slide rule to be faster for, e.g., a straight
multiply, log.

[Joe Morris]

"Peter Flass" <Peter...@Yahoo.com> wrote:

> Maybe not the first time they were read, but card decks were often read
> over and over, and gradually things tended to happen.

...which is why console decks (the decks that sat atop the reader, and were
repeatedly read in - for example, the FORTRAN compiler for the 1620 for
systems with no disk storage) were frequently created using edge-coated
cardstock that resisted fraying. I have no recollection of the additional
cost for edge-coated vs. plain cardstock other than it was large enough that
nobody even dreamed about using the edge-coated stock for everything, but
for cards that would be read over and over and over, I don't recall hearing
anyone complaining about the decision to spend the money to get edge-coated
cards.

Joe

[Ahem A Rivet's Shot]

On Sat, 18 Aug 2012 18:14:53 -0400
Dan Espen <des...@verizon.net> wrote:

> My experience with card readers is that they weren't all that sensitive
> and jammed rarely. Any card handler worth his salt would joggle the
> cards so they were all aligned before reaching the picker knives. No
> big deal.

IME all it took was a kink on the leading edge and that card would
split on the gate in the 1442 and the next two or three cards would split on
it, by the time the machine stopped bits of those three cards would be all
through the mechanism and it was card saw time.

[Ahem A Rivet's Shot]

accurate != precise. It's easier to make a mistake on a
calculator, one wrong digit and the errors tend to get nasty.

[Ahem A Rivet's Shot]

On Sat, 18 Aug 2012 18:48:47 -0400
"Joe Morris" <j.c.m...@verizon.net> wrote:

> "Ahem A Rivet's Shot" <ste...@eircom.net> wrote:
>
> > A slide rule in skilled hands is also faster and more accurate than
> > a calculator.
>
> The meaning of "accurate" depends on several criteria, including how many
> decimal places are required.

The slide rule gave less precision and fewer errors, based on our
daily slide rule drills, in which for the last couple of years calculators
were allowed. Nobody ever won with a calculator and when the answers were
wrong they were hideously wrong.

> However...while it may give you a faster respone delivering the digit
> string, the slide rule tells you nothing about the location of the
> decimal point in the answer.

Yep, keeping track of the decimal point was part of the skill.

[Ahem A Rivet's Shot]

On Sun, 19 Aug 2012 00:40:15 -0400
Shmuel (Seymour J.) Metz <spam...@library.lspace.org.invalid> wrote:

> In <20120818224313....@eircom.net>, on 08/18/2012
> at 10:43 PM, Ahem A Rivet's Shot <ste...@eircom.net> said:
>
> >A slide rule in skilled hands is also faster and more accurate than a
> >calculator.
>
> No. For simple calculations it is faster, but it has a built-in
> imprecision and has no memory. I would expect a skilled user to get
> X^Y faster on a modern calculator than the slide rule, although I
> would expect the slide rule to be faster for, e.g., a straight
> multiply, log.

Our slide rule drills were usually multi-step calculations and
would sometimes involve using the exponential scales to get powers. On
precision the calculator wins hands down - on speed it fails it's quicker
to set a slide than type the digits. - on accuracy it fails it's too easy
to mistype a digit and not notice.

[Rod Speed]

Ahem A Rivet's Shot <ste...@eircom.net> wrote

> Dan Espen <des...@verizon.net> wrote

>> My experience with card readers is that they weren't all that sensitive
>> and jammed rarely. Any card handler worth his salt would joggle the
>> cards so they were all aligned before reaching the picker knives. No
>> big deal.

> IME all it took was a kink on the leading edge and that card would
> split on the gate in the 1442 and the next two or three cards would
> split on it, by the time the machine stopped bits of those three
> cards would be all through the mechanism and it was card saw time.

Never say that happen with any of the card readers I ever used.

Must have been a lousy design in that area.

[Dan Espen]

Ahem A Rivet's Shot <ste...@eircom.net> writes:

> On Sat, 18 Aug 2012 18:14:53 -0400
> Dan Espen <des...@verizon.net> wrote:
>
>> My experience with card readers is that they weren't all that sensitive
>> and jammed rarely. Any card handler worth his salt would joggle the
>> cards so they were all aligned before reaching the picker knives. No
>> big deal.
>
> IME all it took was a kink on the leading edge and that card would
> split on the gate in the 1442 and the next two or three cards would split on
> it, by the time the machine stopped bits of those three cards would be all
> through the mechanism and it was card saw time.

Never saw a 1442 tear up a card.

Worst I saw is a card hang up in the card path and a couple of cards jam
into it before feeding stopped. You could just pull out the cards, dupe
them on the key punch and restart the job.

I suppose it could jam up worse than that, just never saw it.
(Or never had to card saw it.)

--
Dan Espen

[jmfbahciv]

Shmuel (Seymour J.) Metz wrote:
> In <PM0004C78...@ac8409ec.ipt.aol.com>, on 08/18/2012
> at 02:02 PM, jmfbahciv <See....@aol.com> said:
>
>>Not necessarily. The difference between TOPS-10's CALL UUOs and the
>>CALLI UUO may be an example.
>
> TOPS-10 is software.

Doing hardware instructions. Think about the device and controller codes
to make them go.

The System Reference cards have those opcodes.

/BAH

[Robin Vowels]

On Aug 19, 12:02 am, jmfbahciv <See.ab...@aol.com> wrote:

> A good operator [I don't know how to explain this] took handful of cards,

> ruffled it in several different ways, knocked it on a hard surface and

> felt the sides (edges) of that deck.

That was standard practice.
Our card reader had an L-shaped chromed plate on which a card deck
was knocked sideways and up and down so as to get the cards
perfectly lined up, prior to placing them in the reader chute.

[Robin Vowels]

On Aug 19, 7:50 am, "Rod Speed" <rod.speed....@gmail.com> wrote:
> Robin Vowels <robin.vow...@gmail.com> wrote
>
> > Rod Speed <rod.speed....@gmail.com> wrote
> >> Robin Vowels <robin.vow...@gmail.com> wrote
> >>> Ben Pfaff <b...@cs.stanford.edu> wrote
> >>>> hanco...@bbs.cpcn.com wrote
> >>>>> Ben Pfaff <b...@cs.stanford.edu> wrote
> >>>>>> Adding machine, too.  I've never seen one outside of a museum.
> >>>>> Maybe the classic adding machines, but kids do use calculators.
> >>>> Calculators aren't out of place yet, but soon.
> >>> Don't count on it. [no pun intended]
> >>> The abacus is still with us!
> >> But only dinosaurs still use them.
> > If so, the dinosaurs are young ones.
>
> > The Chinese and/or Japanese still use the abacus,
>
> > and skilled users are quicker at it than
> > those using electronic calculators.
>

> That's just plain wrong.

I'm afraid that it's true.

> > I recall an abacus behind a glass window at a computer site.
> > It read: "In emergency break glass".
>
> Sure, but that's just a joke and no one bothered to do that.

Of course it's a joke.

[Robin Vowels]

On Aug 19, 8:48 am, "Joe Morris" <j.c.mor...@verizon.net> wrote:
> "Ahem A Rivet's Shot" <ste...@eircom.net> wrote:
>
> > A slide rule in skilled hands is also faster and more accurate than
> > a calculator.
>
> The meaning of "accurate" depends on several criteria, including how many
> decimal places are required.
>
> However...while it may give you a faster respone delivering the digit
> string, the slide rule tells you nothing about the location of the decimal
> point in the answer.  Rightly or wrongly, from the 1930s through the
> mid-1960s, my father prohibited his students from using a slide rule for any
> of their work.

A slide rule was an essential tool of engineers until electronic
hand calculators came along.
As well as the usual mult/div functions, trig functions etc were
provided. For more accurate work, log/sin tables were used.

[Michael Black]

On Sat, 18 Aug 2012, Shmuel (Seymour J.) Metz wrote:

> In <Pine.LNX.4.64.12...@darkstar.example.net>, on
> 08/18/2012
> at 01:11 AM, Michael Black <et...@ncf.ca> said:
>
>> But that TI-30 I got as my first scientific calculator about 1976, it
>> had more functions. And that level of functions remained "basic"
>> (unless you went for a four function) for a long time. When I
>> replaced the TI-30 with an LCD calculator in the late eighties, it
>> had a few more functions, and maybe more important, could do
>> conversions between binary/hex/octal/deciaml, though no actual logic
>> operations.
>
> TI already had a hex calculator in the 1970's. I don't recall the
> exact date, but it was before 1976.
>

Yes, I think it was '76. It wasn't that much price-wise (not like the HP
calculators), but more than I could justify. Oddly, it was a time when I
really could have used it, hand assemblying code starting in 1979 on my
KIM-1. I just kept lookup charts handy, decimal to hex, an ASCII chart,
the card of op-codes that came with the KIM-1.

Michael

[Robin Vowels]

After about 20 to 50 passes thru the card reader, is was necessary to
take the deck to the card reproducer and make a new copy.

[Robin Vowels]

On Aug 19, 7:43 am, Ahem A Rivet's Shot <ste...@eircom.net> wrote:
> On Sat, 18 Aug 2012 06:27:19 -0700 (PDT)
>
>
>
>
>
>
>
>
>

> Robin Vowels <robin.vow...@gmail.com> wrote:

> > On Aug 18, 12:46 pm, "Rod Speed" <rod.speed....@gmail.com> wrote:
> > > Robin Vowels <robin.vow...@gmail.com> wrote
>
> > > > Ben Pfaff <b...@cs.stanford.edu> wrote
> > > >> hanco...@bbs.cpcn.com wrote
> > > >>> Ben Pfaff <b...@cs.stanford.edu> wrote
> > > >>>> Adding machine, too.  I've never seen one outside of a museum.
> > > >>> Maybe the classic adding machines, but kids do use calculators.
> > > >> Calculators aren't out of place yet, but soon.
> > > > Don't count on it. [no pun intended]
> > > > The abacus is still with us!
>
> > > But only dinosaurs still use them.
>
> > If so, the dinosaurs are young ones.
> > The Chinese and/or Japanese still use the abacus,
> > and skilled users are quicker at it than those
> > using electronic calculators.
>

>         A slide rule in skilled hands is also faster and more accurate than
> a calculator.

And quieter too, specially for multiplication.