Why didn't PL/I take off?
Paul Tomblin
>The list of PL/I users is a short one, but those people like it very
>much. However, the rest of the world never warmed to PL/I, sticking
>with COBOL and Fortran.
>
>Anyone know why?
Bloat. The PLI compiler I used at my last mainframe job (1982-1987,
thankfully far behind me) was slow, used tons of memory, and produced programs
that had the same faults. If what you wanted to do could be done in
FORTRAN-Hext instead, we used that. I wrote a *lot* of FORTRAN back then, and
only used PLI for one wierd program that used a bunch of SPF screens for data
entry.
--
Paul Tomblin (ptom...@xcski.com, formerly ptom...@canoe.com)
<a href="http://www.servtech.com/public/ptomblin/">My home page</a>
"The superior pilot uses his superior judgement to avoid situations in which
he has to demonstrate his superior skill" - anon.
Gene Wirchenko
>The list of PL/I users is a short one, but those people like it very
>much. However, the rest of the world never warmed to PL/I, sticking
>with COBOL and Fortran.
>Anyone know why?
Let's see. It stole features from FORTRAN, ALGOL, and COBOL
[Fine] and combined them into an awkward, complex mess [Not fine].
Just try writing a program with robust error handling.
Sincerely,
Gene Wirchenko
C Pronunciation Guide:
y=x++; "wye equals ex plus plus semicolon"
x=x++; "ex equals ex doublecross semicolon"
TUrner-7
John A. Limpert
One major problem was the lack of compilers that implemented the
full language. I have an old PL/I book that devotes many pages
to describing the subsets of the language supported by various
compilers. A well defined subset was standardized later but it
never caught on. I've been told that it was very difficult to
write a good PL/I compiler.
Tim Shoppa, x6446
>The list of PL/I users is a short one, but those people like it very
>much. However, the rest of the world never warmed to PL/I, sticking
>with COBOL and Fortran.
>
>Anyone know why?
PL/I did make some progress into "the rest of the world" (which I assume
means outside IBM). In particular, it inspired PL/M, which
was one of the first implementations of
a high-level-like language for microcomputers (cross-compilers existed
for both the 8008 and 8080) which made a lasting impact as many parts
of CP/M were written in it. (I have a cross-compiler for the 8080
written in Fortran IV, of all things. PL/M really isn't even
a subset of PL/I, it is substantially different in many respects,
particularly data types. But it does show that PL/I had some
lasting influence.) Several minicomputer implementations
were also available and moderately popular (at least some of the VMS
system utilities are still in PL/I, I'm told.)
But the main reason PL/I never became really popular was creeping
featurism. Well, in the case of PL/I, it wasn't creeping, it was
dominant. Subsets of PL/I were defined (I seem to recall "PL/G" as
being the one many minicomputer installations had) that were reasonably
trimmed down to what most people would need, but everybody had just
enough useful extensions past the bare-minimum subset that very few
programs were actually transportable across platforms.
Tim Shoppa (sho...@triumf.ca)
Joe Morris
>tur...@pacsibm.org (TUrner-7) wrote:
>>The list of PL/I users is a short one, but those people like it very
>>much. However, the rest of the world never warmed to PL/I, sticking
>>with COBOL and Fortran.
>>Anyone know why?
> Let's see. It stole features from FORTRAN, ALGOL, and COBOL
>[Fine] and combined them into an awkward, complex mess [Not fine].
>Just try writing a program with robust error handling.
It was also a resource hog, although the comment needs to be taken
in the context of the day. Don't forget that in the late 60s when NPL
(excuse me, "PL/1". I'm showing my age) was introduced, your mainframe
users were squabbling over memory allocations in units of 2KB; a middle-
sized machine might have total memory of 256 KB (512 KB if the football
team could be persuaded to give up an assistant waterboy's salary) and
only the biggest of the big sites had the money to even dream of a
1 MB mainframe. Don't forget that OS/360 ran only real memory, so
what you bought was what you used.
And systems were incredibly sluggish by today's standards. One of my
lecture stories was having to buy memory for a 360/65: 256 KB of 600 nsec
memory cost me about $330,000, and if you really had to have more
memory but didn't have that type of money you had to settle for
the 8 usec (?) Large Capacity Storage or LCS box.
Today, the resources used by a typical PL/1 program wouldn't be
noticably different from applications written for current-vogue
compilers.
Oh yes...don't forget the character set. In the early days of the S/360
your workhorse 1403 printer used an AN (business) or HN (scientific)
character train, offering for the most part only the basic FORTRAN
charcter set or its business equivalent. If you wanted to print a
PL/1 program that used the (for the time) significantly larger character
set, you had to arrange for the operator to change the train, or the
installation had to routinely use the print train with the larger
character set...and suffer the reduced print throughput because there
were fewer instances of some characters. (A 1416 print train for the
1403 printer had 240 character positions. A print line was complete and
paper motion started as soon as all positions with characters had been
struck, so the more instances of a given character the more likely that
you could print the entire line in less time than was required for
the train to complete one cycle.)
Joe Morris / MITRE
Lawrence E. McKnight
>tur...@pacsibm.org (TUrner-7) wrote:
>
>>The list of PL/I users is a short one, but those people like it very
>>much. However, the rest of the world never warmed to PL/I, sticking
>>with COBOL and Fortran.
>
>>Anyone know why?
>
> Let's see. It stole features from FORTRAN, ALGOL, and COBOL
>[Fine] and combined them into an awkward, complex mess [Not fine].
>Just try writing a program with robust error handling.
Gee, the first time I ever actually saw any PL/I code (my sister was
working on converting the programs to COBOL), I commented 'gee, it looks
like it has all the good features of FORTRAN and COBOL'. My sister
replied 'No, it has all the bad features of them'.
>
>Sincerely,
>
>Gene Wirchenko
>
>C Pronunciation Guide:
> y=x++; "wye equals ex plus plus semicolon"
> x=x++; "ex equals ex doublecross semicolon"
>
---------------
Larry McKnight
(this space unintentionally left blank.....
Nick Spalding
>in the context of the day. Don't forget that in the late 60s when NPL
>(excuse me, "PL/1". I'm showing my age) was introduced, your mainframe
Nice to see NPL again! Prior to 360's official announcement both it
and PL/I were known internally as NPL, New Product Line and New
Programming Language.
Nick
Powwow/Email at spal...@iol.ie
The total of human misery is more or less constant.
The best we can hope to do is redistribute it a little.
Anyone know the source of this (mis)quote?
Heinz W. Wiggeshoff
>
> The total of human misery is more or less constant.
> The best we can hope to do is redistribute it a little.
>
> Anyone know the source of this (mis)quote?
"The poor will always be with you."
Jesus Christ, circa 33.
I interpret that to be "The stupid rulez" ... and that's why
PL/1 (and APL) didn't take off!
Dale DePriest
in a catalog for OS/2 products.
--
_ _ Dale DePriest San Jose, California
/`) _ // da...@Cadence.COM voice: (408) 428-5249
o/_/ (_(_X_(` fax: (408) 894-3457
Charles Richmond
>In article <4iqg1e$9...@netaxs.com>, tur...@pacsibm.org (TUrner-7) writes...
>>much. However, the rest of the world never warmed to PL/I, sticking
>>with COBOL and Fortran.
>>
>>Anyone know why?
>
>means outside IBM).
>
science used PL/I except those courses which were designed to teach other languages specifically, such as assembly language, etc. PL=
/I had it's quirks, but it *did* offer some things that FORTRAM and COBOL did *not* have at the time. Specifically it had a kind of =
variable which could be allocated dynamically at runtime. Strangely, it would *stack* the allocations, so when you freed one, the pr=
evious allocation was revealed for use. It also supported a pointer data type in kind of a Pascal way. It allowed cross-section assi=
gnments of arrays with single assignment statements. This feature most closely resembled APL.
I liked PL/I, even with the massive amount of capabilities it offered. One really needed only a subset of the language for any given=
program. It also did implicit conversions with abandon. I new a girl that had function calls in a PL/I program nested five or six d=
eep. She could *not* figure out why she was getting a real off-the-wall answer. Turns out that PL/I was converting what she thought =
was an integer result into a *bit* string when one of the functions returned. Since the next function required an integer as an argu=
ment, it converted the first *bit* to an integer and used that!
From 1989 to 1991, I worked at DSC Communications. They had a VAX cluster with 10 VAX 8600's in it as their main computing resource.=
And yes, the VAX had a PL/I compiler on it. I never used it and don't know who did. But there it was on a DEC machine in the 90's.
Also, American Airlines did use (and may still use) PL/I to do their financial computing in their financil headquarters in Tulsa, OK=
Vladimir Zakharov
: The list of PL/I users is a short one, but those people like it very
: much. However, the rest of the world never warmed to PL/I, sticking
: with COBOL and Fortran.
But it did. I mean, the rest of the world. Well, I can't speak for the the
whole rest of the world, but in Russia PL/I was very popular on IBM clones
(and I don't mean IBM *PC* clones). Most of the accounting and other business
application running on mainframes are written in PL/I (unless, of course,
they are russified "clones"). Probably for the same reason that OS/2 is more
popular among home users in Russia than Win'95. In this case, when you buy
your OS on a 5$ Chinese CD, then the only deciding factor would be the
technical superiority of the product.
--
Vladimir Zakharov | "If only the eyes could see o O
zakh...@panix.com | Then you'd know your destiny OoOO
| The mind is an enemy OU==\
| And you just need a friend" RJD \ |/
John Morris
da...@cds9172.Cadence.COM "Dale DePriest" writes:
> Surprisingly PL/1 is still being offered as a product. I just saw it
> in a catalog for OS/2 products.
(Slight topic drift, but hopefully in the right direction...)
When I was first introduced to PL/1, I was told that IBM had also
trademarked "PL/2" through to "PL/9", for future use.
True?
--
John Morris Due to circumstances beyond my control,
Jo...@kirsta.demon.co.uk I am master of my fate and captain of my soul.
GM4ANB@GB7EDN.#77.GBR.EU - Ashleigh Brilliant
Lon Stowell
it off in a snit of anger. Allegedly some competitors were taking
IBM's pseudo-code from some of the Program Logic Manuals and coding
things that WORKED...and since IBM couldn't figure out how they were
doing it, they killed PL/1.
Michael D Shapiro
>The list of PL/I users is a short one, but those people like it very
>much. However, the rest of the world never warmed to PL/I, sticking
>with COBOL and Fortran.
>
PL/I (originally called NPL for "New Programming Language") started out
as FORTRAN VI. IBM hoped that it would be the next generation of
FORTRAN. Then someone decided to add many features of COBOL and it
just kept growing. (It's rumored that the SHARE and IBM committee that
developed the language had a favorite dinner -- pizza with everything.)
When the 360 was first announced, PL/I was the only language announced.
As it became clear that customers didn't want to redo all their legacy
software (even in the 1960s) IBM scrambled to find FORTRAN and COBOL
compilers that could be ported to their machine, making a few
entrepreneurs well-to-do.
As people tried implementing PL/I for other machines, they ran across
more and more of the 360isms built into the language and determined
that it would not be easy to fully implement on other architectures. Many
projects used the language as the starting point for a definition, with
variations named such things as PL/M, PL/S, PL/0 (that's zero, a small
subset) and PL/360. They (and other languages) had in common taking
ideas from Algol and applying them as extensions to FORTRAN, using some
variation of PL/I syntax.
We've seen this process repeatedly in the past few decades. Algol
begot Pascal, which was useful for teaching programming techniques but
not good for "real" work. In the 1970s, many of us worked to extend
Pascal to make it useful. Some kept the name "Pascal" for the language
and others invented new names. We had a wild proliferation of
"Pascal-like" languages, all incompatible.
Then "C" came along (derived from "BCPL" and "B" et al) and it became a
common goal, even though at the start many people thought that C might
not be implementable on architectures other than the DEC. By the early
1980s, many developers were moving from their Pascal-like languages to
C (although some continued to use a Pascal). The major big holdout was
the Department of Defense, with its Ada language.
Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
just like most programming languages. We learned from its good
features and laughed at its bad features. I'll never forget the wonder
of a diagnostic resulting from a keypunch error:
1.41.5 instead of 1.41,5
resulted in "Illegal Sterling constant" (left as an exercise to the
reader).
--
Michael D. Shapiro, Ph.D. Internet: msha...@nosc.mil
Code 4123, NCCOSC RDT&E Division (NRaD) San Diego CA 92152
Voice: (619) 553-4080 FAX: (619) 553-4808 DSN: 553-4080
Adam J. Thornton
It's not exactly dead.
I got a brochure for IBM Visual PL/I recently.
Honest.
Adam
--
ad...@phoenix.princeton.edu | Viva HEGGA! | Save the choad! | 64,928 | Fnord
"Double integral is also the shape of lovers curled asleep":Pynchon | Linux
Thanks for letting me rearrange the chemicals in your head. | Team OS/2
You can have my PGP passphrase when you pry it from my cold, dead brain.
Per Lindqvist
> were also available and moderately popular (at least some of the VMS
> system utilities are still in PL/I, I'm told.)
I programmed in PL/I on a CP/M machine, so it was available during
that time.
Heinz W. Wiggeshoff
Michael D Shapiro (msha...@nosc.mil) writes:
>
> As people tried implementing PL/I for other machines, they ran across
> more and more of the 360isms built into the language and determined
> that it would not be easy to fully implement on other architectures. Many
> projects used the language as the starting point for a definition, with
> variations named such things as PL/M, PL/S, PL/0 (that's zero, a small
> subset) and PL/360. They (and other languages) had in common taking
> ideas from Algol and applying them as extensions to FORTRAN, using some
> variation of PL/I syntax.
>
a /360 environment. True, the I/O part of the language was heavily
influenced by Data Management Macro manual, (eg. teleprocessing, then
VSAM, etc.).
PL/M derived from XPL (McKeeman Wortman Horning) which was strongly
influenced by PL/1. PL/360 comes from the nutty notion that one can
write assembler as though it's a high level language. RIP!
>
> Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
> just like most programming languages. We learned from its good
> features and laughed at its bad features. I'll never forget the wonder
> of a diagnostic resulting from a keypunch error:
>
> 1.41.5 instead of 1.41,5
>
> resulted in "Illegal Sterling constant" (left as an exercise to the
> reader).
Clue #1: which IBM lab developed the PL/I-F compiler?
Clue #2: it hasn't quite fallen by the wayside (see OS/2), but the
job offers are scarce!
Finally, the list of "We" who learned from its good features is woefully
short. From Dr. Dobb's Journal, June 1994, Letters:
Revisiting Variably Dimensioned Arrays in C (from T. Graham Freeman,
Australian Defense Force Academy)
"Many scientific and engineering programmers are changing from Fortran
to C or C++ ...
However, a serious weakness of C/C++ is the inability of either
language to handle arrays with two or more dimensions as arguments
to a function - if the array sizes are to be specified at run time."
Well Graham, (and all you other iron-ring impaired wannabe programmers),
this issue was solved in the 1960's for APL and PL/1. And if that's the
only serious weakness you've noticed, you'd be well advised to go back
to a calculator.
Gene Wirchenko
[snip]
>Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
>just like most programming languages. We learned from its good
>features and laughed at its bad features. I'll never forget the wonder
>of a diagnostic resulting from a keypunch error:
> 1.41.5 instead of 1.41,5
>resulted in "Illegal Sterling constant" (left as an exercise to the
>reader).
Yet another example of PL/I trying to have everything?
41 farthings is not less than a pound. Was this before the Brits
decimali{zs}ed their currency or on an old compiler?
>--
>Michael D. Shapiro, Ph.D. Internet: msha...@nosc.mil
>Code 4123, NCCOSC RDT&E Division (NRaD) San Diego CA 92152
>Voice: (619) 553-4080 FAX: (619) 553-4808 DSN: 553-4080
Sincerely,
Nick Spalding
>msha...@nosc.mil (Michael D Shapiro) wrote:
>
>[snip]
>
>>Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
>>just like most programming languages. We learned from its good
>>features and laughed at its bad features. I'll never forget the wonder
>>of a diagnostic resulting from a keypunch error:
>
>> 1.41.5 instead of 1.41,5
>
>>resulted in "Illegal Sterling constant" (left as an exercise to the
>>reader).
>
> Yet another example of PL/I trying to have everything?
> 41 farthings is not less than a pound. Was this before the Brits
>decimali{zs}ed their currency or on an old compiler?
Shillings, not farthings. 20 shillings to the pound, 12 pennies to
the shilling.
--
Nick
Powwow/Email at spal...@iol.ie
Charles Richmond
>Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
>just like most programming languages. We learned from its good
>features and laughed at its bad features. I'll never forget the wonder
>of a diagnostic resulting from a keypunch error:
>
> 1.41.5 instead of 1.41,5
>
>resulted in "Illegal Sterling constant" (left as an exercise to the
>reader).
>
British monetary system of the time. You had to deal with pounds,
shillings, and pence. I believe it was as confusing as cups, pints,
quarts, and gallons are to the non-U.S. person. Now, of course, Great
Britain has a decimal monetary system. Can someone from Britain who can
still remember those days please post the details of how the "Sterling"
constants would be written?
gber...@castle.net
>In article <4iqg1e$9...@netaxs.com>, TUrner-7 <tur...@pacsibm.org> wrote:
>The list of PL/I users is a short one, but those people like it very
>much. However, the rest of the world never warmed to PL/I, sticking
>with COBOL and Fortran.
>
>Anyone know why?
I can't vouch for the rest of the world, but in 1968 I was asked to do
an evaluation on the 360/65 running OS/PCP for the company I worked for.
I "wrote" the following 3 line program:
READ READ READ READ
WRITE WRITE READ
GOTO READ
Although the compiler generated several pages of warning diagnostics,
it produced a working card to print utility program.
That early version of the compiler did everything it could to
produce a working program in spite of intentional or unintentional
coding errors. I, therefore, deemed the language too dangerous for
production work.
Regards,
Gary
Advanced Debugging System for CICS
http://www.castle.net/~gbergman
Nick Spalding
>> 1.41.5 instead of 1.41,5
Format was L.S.D
L is pounds up to some max I forget
S is shillings in range 0 to 19
D is pence in range 0 to 11
So the original 1.41.5 fails on the value of S
--
Nick
Powwow/Email to spal...@iol.ie
Lon Stowell
>
>It's not exactly dead.
>I got a brochure for IBM Visual PL/I recently.
>Honest.
That is almost as "quaint" to consider as HTML--3270.
Edward Rice
gber...@castle.net wrote:
> I "wrote" the following 3 line program:
>
> READ READ READ READ
> WRITE WRITE READ
> GOTO READ
>
> Although the compiler generated several pages of warning diagnostics,
> it produced a working card to print utility program.
>
> That early version of the compiler did everything it could to
> produce a working program in spite of intentional or unintentional
> coding errors. I, therefore, deemed the language too dangerous for
> production work.
Production sites only used code if the compile-level warnings were either
nonexistent or trivial. (Depending on whether it was Friday or Monday, and
what project it was.)
I worked at one IBM PL/I (and other languages) site, on Multics while at
Honeywell, and then on another IBM-oriented PL/I-emphasized site, and most
of the nastiest things people can say about PL/I really related to just one
or two compilers that IBM wrote. The F-level compiler was feature-poor and
really not much better than Fortran /or/ Cobol; the H-level compiler was
code-generatur bug-rich, slow, and cumbersome; G-level didn't generate very
good code because it combined a so-so code-gen with most of the language
features. Between those three, anybody can find a lot to complain about.
I've heard /very/ little about the PL/I that Univac brought out for the
1108 (and related machines). I'd be very interested in anybody's
first-hand experienced with that compiler and run-time -- I read the manual
with attention to detail and it looked to be very, very nice, and of course
Univac offered an /extremely/ fine run-time environment for other code.
(Erm... war stories about broken computers do not apply to the /design/ I'm
referring to.)
Eric Chomko
.....as compared to C? (I believe this is what you mean).
The poster prior to me has a nice disseration that follows, I'll keep
mine simple and short.
PL/1 cost money as it came from IBM. C was free as it came from Berkeley
from Bell Labs. If Bell Labs marketed C and Unix like IBM marketed MVS
and PL/1 Unix and C would not nearly be as popular today as they are.
If IBM had given PL/1 away to say MIT then PL/1 would be the language
today. Trust me it was the money.
Eric
Michael D Shapiro (msha...@nosc.mil) wrote:
: In article <4iqg1e$9...@netaxs.com>, TUrner-7 <tur...@pacsibm.org> wrote:
: >The list of PL/I users is a short one, but those people like it very
: >much. However, the rest of the world never warmed to PL/I, sticking
: >with COBOL and Fortran.
: >
: >Anyone know why?
: PL/I (originally called NPL for "New Programming Language") started out
: as FORTRAN VI. IBM hoped that it would be the next generation of
: FORTRAN. Then someone decided to add many features of COBOL and it
: just kept growing. (It's rumored that the SHARE and IBM committee that
: developed the language had a favorite dinner -- pizza with everything.)
: When the 360 was first announced, PL/I was the only language announced.
: As it became clear that customers didn't want to redo all their legacy
: software (even in the 1960s) IBM scrambled to find FORTRAN and COBOL
: compilers that could be ported to their machine, making a few
: entrepreneurs well-to-do.
: As people tried implementing PL/I for other machines, they ran across
: more and more of the 360isms built into the language and determined
: that it would not be easy to fully implement on other architectures. Many
: projects used the language as the starting point for a definition, with
: variations named such things as PL/M, PL/S, PL/0 (that's zero, a small
: subset) and PL/360. They (and other languages) had in common taking
: ideas from Algol and applying them as extensions to FORTRAN, using some
: variation of PL/I syntax.
: We've seen this process repeatedly in the past few decades. Algol
: begot Pascal, which was useful for teaching programming techniques but
: not good for "real" work. In the 1970s, many of us worked to extend
: Pascal to make it useful. Some kept the name "Pascal" for the language
: and others invented new names. We had a wild proliferation of
: "Pascal-like" languages, all incompatible.
: Then "C" came along (derived from "BCPL" and "B" et al) and it became a
: common goal, even though at the start many people thought that C might
: not be implementable on architectures other than the DEC. By the early
: 1980s, many developers were moving from their Pascal-like languages to
: C (although some continued to use a Pascal). The major big holdout was
: the Department of Defense, with its Ada language.
: Anyway PL/I (or PL/1 -- both variations were used) fell by the wayside
: just like most programming languages. We learned from its good
: features and laughed at its bad features. I'll never forget the wonder
: of a diagnostic resulting from a keypunch error:
: 1.41.5 instead of 1.41,5
: resulted in "Illegal Sterling constant" (left as an exercise to the
: reader).
: --
Heinz W. Wiggeshoff
Eric Chomko (cho...@haven.ios.com) writes:
> ......as compared to C? (I believe this is what you mean).
>
> The poster prior to me has a nice disseration that follows, I'll keep
> mine simple and short.
>
> PL/1 cost money as it came from IBM. C was free as it came from Berkeley
> from Bell Labs. If Bell Labs marketed C and Unix like IBM marketed MVS
> and PL/1 Unix and C would not nearly be as popular today as they are.
>
>
> If IBM had given PL/1 away to say MIT then PL/1 would be the language
> today. Trust me it was the money.
The first time I laid eyes, (but not keystrokes) upon PL/1 was at Queen's U.
in Kingston Ont. circa 1969.
In those days, the rumour mill said that universities go IBM hardware
and software at a fraction of the commercial price to induce us young'uns
to wear Blue underwear.
(Finally C is to PL/1 what walking is to the space shuttle.)
Adam J. Thornton
Nick Spalding <spal...@iol.ie> wrote:
>Format was L.S.D
>L is pounds up to some max I forget
>S is shillings in range 0 to 19
>D is pence in range 0 to 11
>So the original 1.41.5 fails on the value of S
Doesn't the PL/I Specification say that Sterling constants must end in "L"?
Thus 1.19.5 would have to be 1.19.5L ?
I'm pretty sure at least the System/360 PL/I spec (before any compilers
were actually delivered) says this.
Joe Morris
> ......as compared to C? (I believe this is what you mean).
>
> The poster prior to me has a nice disseration that follows, I'll keep
> mine simple and short.
>
> PL/1 cost money as it came from IBM. C was free as it came from Berkeley
> from Bell Labs. If Bell Labs marketed C and Unix like IBM marketed MVS
> and PL/1 Unix and C would not nearly be as popular today as they are.
> If IBM had given PL/1 away to say MIT then PL/1 would be the language
> today. Trust me it was the money.
That's interesting, since the original PL/1 compiler wasn't even
copyrighted by IBM. If you weren't an IBM hardware customer you
had to pay what amounted to handling charges if you wanted to get
a copy from IBM, but there weren't any restrictions on redistributing
either the object >>or source<< for the compiler.
See any OS/360 distribution kit, and look at the entries for the
PL/1 F compiler (360S-NL-511) or library (360S-LM-512).
Of course, after New World the PL/1 implementations were all
copyrighted program products, but in the early days it was free.
Just like the original HASP, which prompted some would-be bard to pen:
HASP belongs to everyone;
The best things of life are free...
The tasks are active, every one:
They're working for you and me...
Joe Morris / MITRE
John Hendrickx
PL/I that started off on IBM mainframes and has since been ported to a
number of platforms. An official version is included in OS/2 and AIX
4.x. Regina, a GNU freeware port, is available at
http://wwwcn.cern.ch/regina/.
SAS started out as a statistical package but has expanded into things
like EIS (executive information systems). The mainframe version (pre
version 5, which was introduced somewhere around 1986) was written
mainly in PL/I and the SAS language is heavily inspired by it. SAS
Institute refers to its product as a 4GL. They had to switch to C when
they went cross-platform in version 5 and up.
Lon Stowell
>
>Were I spending much time on an actual mainframe terminal--and they do
>still exist--I'd certainly want a web browser I could use from it.
>Something like lynx would be quite sufficient. However, can't later-model
>3278s (or something like them) do 16-color graphics too? I seem to recall
>Rick once showing me something very much like a gif decoder on one of our
>True Blue Terminals.
Any 3270 terminal with Programmable Symbols can do rather
surprising graphics, but you won't downloading GIF's of
your favorite subject with any sort of resolution.
The 3279 S3G and 3179-G have rather impressive graphics
capabilities, a web browser based on one of those would
be an interesting exercise--in a morbid sort of way.
Eric Burch
is looking for programmers and one of the skills they're
looking for is PL/I experience. (I guess if the Microsoft
internet browser doesn't work out they could write a new one
using PL/I...maybe.)
(AOL is based just outside D.C., and I don't use AOL.)
--
Eric Burch -- Loral Federal Systems -- Gaithersburg, MD
er...@lfs.loral.com usual disclaimers apply
imagine a " :-) " after each period above
ph...@watdcs.uwaterloo.ca
J.Hen...@maw.kun.nl (John Hendrickx) wrote:
>PL/I lives on in REXX and SAS. REXX is a macro language loosely based on
>PL/I that started off on IBM mainframes and has since been ported to a
>number of platforms. An official version is included in OS/2 and AIX
>4.x. Regina, a GNU freeware port, is available at
>http://wwwcn.cern.ch/regina/.
While John's suggestion that REXX was "loosely based on PL/I" isn't
*incorrect*, it isn't really correct either. Some aspects of REXX were
inspired by the good points of PL/I; I don't consider that synonymous with
"loosely based on". Perhaps I'm nitpicking (won't be the first time).
Meanwhile, REXX is provided as part of the operating system on MVS, VM,
OS/400, AIX 4.x (as noted above), IBM PC-DOS 7.0, and the Commodore Amiga
(RIP). It's also available from one or (often) more vendors for various other
Unix platforms, Tandem, DOS (for 12 years now!), Windows, OS/2 (yes,
duplicating to some extent the built-in version), VSE, and probably a few
others I've forgotten about.
REXX's elegance, power, and simplicity have made it the macro language of
choice for most who have it available, including end-users who would never
dream of learning, say, Visual BASIC (e.g., liberal arts professors).
There's a lot more REXX information available at
<http://rexx.hursley.ibm.com/rexx>.
..phsiii (co-editor of "The REXX Handbook", McGraw-Hill, 1992)
Valdis Kletnieks
>IBM's pseudo-code from some of the Program Logic Manuals and coding
>things that WORKED...and since IBM couldn't figure out how they were
>doing it, they killed PL/1.
Well.. let's get this straight.. What was *really* happening was that
IBM was supplying PL/S (*NOT* PL/1) source for some modules that were
alledgedly "Object Code Only", and a customer took the PL/S language
spec and wrote his own compiler...
PL/S was a "systems programming" variant based on PL/1 - imagine PL/1 but
with the bit-level structure handling of C rather than the way PL/1 did it...
--
Valdis Kletnieks
Computer Systems Engineer
Virginia Tech
Joseph Chacko
> In article <4jodue$s...@info.uci.kun.nl>,
> J.Hen...@maw.kun.nl (John Hendrickx) wrote:
> >PL/I lives on in REXX and SAS. REXX is a macro language loosely based on
> >PL/I that started off on IBM mainframes and has since been ported to a
> >number of platforms. An official version is included in OS/2 and AIX
> >4.x. Regina, a GNU freeware port, is available at
> >http://wwwcn.cern.ch/regina/.
>
> While John's suggestion that REXX was "loosely based on PL/I" isn't
> *incorrect*, it isn't really correct either. Some aspects of REXX were
> inspired by the good points of PL/I; I don't consider that synonymous with
> "loosely based on". Perhaps I'm nitpicking (won't be the first time).
>
> Meanwhile, REXX is provided as part of the operating system on MVS, VM,
> OS/400, AIX 4.x (as noted above), IBM PC-DOS 7.0, and the Commodore Amiga
> (RIP). It's also available from one or (often) more vendors for various othe
> Unix platforms, Tandem, DOS (for 12 years now!), Windows, OS/2 (yes,
> duplicating to some extent the built-in version), VSE, and probably a few
> others I've forgotten about.
Rexx comes free with OS/2 warp, I think.
--
Psst
Heinz W. Wiggeshoff
In fact, it's an integral part of OS/2.
What bums me out is that PL/1 for OS/2 is $1250, plus
$199 for a toolkit (?); assuming that's US dollars I'd
be looking at $2,000 grand Canadian!
Derek Peschel
Heinz W. Wiggeshoff <ab...@FreeNet.Carleton.CA> wrote:
> What bums me out is that PL/1 for OS/2 is $1250, plus
> $199 for a toolkit (?); assuming that's US dollars I'd
> be looking at $2,000 grand Canadian!
Wow, the exchange rate must be REALLY skewed. Two million dollars is a LOT
to spend on one program!
-- Derek
George
>Joe Morris / MITRE
Unix and C were not "free" from Bell Labs unless you were part of the
Bell System. I ran a small PDP11 in-house system and we were quoted a
price of $40,000. for the sources with NO SUPPORT. We elected to go
with DEC's RSTS/E which was cheaper and supported.
Been there...Done that....
George Orth
Joseph Chacko
assume that it is the precursor to OS/2 (my keen analytical mind at work
again), but I would like to know more. Can anyone enlighten me further?
--
Psst
Nick Spalding
You could not do better than get hold of a book called 'The Mythical
Man-month' by Frederick Brooks. He was in charge of the project,
which was the original top-of-the-range operating system for System
360. It's present day descendant is MVS, hardly OS/2.
--
Nick
Powwow/Email at spal...@iol.ie
A committee is a group of people who individually can decide nothing
but collectively can decide that nothing should be done.
Adam J. Thornton
Joseph Chacko <j...@sheril.demon.co.uk> wrote:
>I recently read about the IBM OS project as an example of a failed project.
No, not really. Slow, bloated, buggy, and a year past due, but ultimately
not a failure. You've heard of MVS, right? It's the lineal descendent of
OS/360.
>I assume that it is the precursor to OS/2
Only in the sense that it also was an operating system made by IBM and it
came before OS/2.
Read Fred Brooks's marvellous and seminal book _The Mythical Man-Month_.
That's the best place to start.
Glenn Chambers
J. Thornton) wrote:
> Read Fred Brooks's marvellous and seminal book _The Mythical Man-Month_.
> That's the best place to start.
This book is now out in a new edition, with additional chapters and a
new forward. I don't have it handy, or I'd include the ISBN & price data.
My copy is a trade-format paperback.
--
Glenn Chambers
gcha...@bright.net
Toledo, OH
John Hendrickx
>Joseph Chacko <j...@sheril.demon.co.uk> wrote:
>>I recently read about the IBM OS project as an example of a failed project.
>
>No, not really. Slow, bloated, buggy, and a year past due, but ultimately
>not a failure. You've heard of MVS, right? It's the lineal descendent of
>OS/360.
>
descendent of OS/360 - MVS? Does anyone know when it originated? I'm curious
about it because it was my first computer system, but our VM mainframe gets
the axe on January 1 :-(. It's a really primitive system in many ways, but
seeing it go makes me sentimental.
Chuck Greenwald
CMS started as the Cambridge (Mass., USA) Monitor System. It was
very heavily influenced by Multics and may be viewed as one of
Unix's many brothers.
-------------------------------------------------------------------------
Chuck Greenwald | All opinions are | chu...@meaddata.com
Lexis/Nexis | my own. | (513) 865-1020
kiy...@ibm.net
Given that the Cambridge Monitor System was running a decade before Unix, it
would be more accurate to call it Unix's rich uncle, Kindel O'Hara.
The MVS precursor, OS/360, was a failure if you consider late, big, slow,
buggy, feature poor, overpriced, memory and CPU demanding, a failure. However,
compared to PC OS's, OS/360 was an incredible success. It ran well and multi-tasked
in 512 K on a .5 MIPS box. On a 1.5 MIPS box, it'll support hundreds of users and
run the compiles and production work for a small company, 1,000 persons or so.
PC and Unix-heads can't appreciate what a full house 3031 or S/370 168 can do.
It is sad to see a real computer like a VM/CMS box get shut down.
OS/2 Software Store http://www.kiyoinc.com
Kiyo Design Inc Stop in for your free ColorWorks CD and poster.
11 Annapolis St. Free ColorWorks CD with each Web order,
Annapolis, Md 21401 just type "CW CD" in the comments field.
gme...@carroll.com
Yes! I did my undergrad projects on an IBM 360/50, MVT, 512K memory and it was
fast!
>
>PC and Unix-heads can't appreciate what a full house 3031 or S/370 168 can do.
You're 100% right!
>
>It is sad to see a real computer like a VM/CMS box get shut down.
Cory, I remember my days of struggle at Columbia U learning how to use a...
Teletypewriter terminal!!!! and the generic Televideo terminals hooked up to
DEC-20's. I never learned how to use the darn thing and hated the DEC-20.
I just had to comtemplate my friends at CUNY (City University of New York) using
VM/SP and MVS/JES3 and how easy it was for them, form their IBM 3278 terminals,
to send jobs to the MVS machine and get it back on their virtual "RDRs", or how
they did it interactively on VM/CMS full screen . And I was envious!! What a beauty
to use XEDIT full screen to write your programs!!. I thought I was not getting my
money's worth at Columbia.
The following semester I took a Comp Sci course at CCNY just to get an account
at CUNYVM. Man what a difference!
What a difference to use REXX to compose "Exec files" (we now call them script)
which would create the JCLs needed to spool my PL/I and BAL programs to the
MVS machine or to do it locally on the VM machine! I never went back to the
Columbia terminal room.
When my profs asked me why I was using another University computer to do
my projects, I told them right away that the Columbia DEC system was crap!!
Since them I've been in love with IBM VM/CMS and would not work on anything
different!
I wished there was a VM/CMS MVS/JES emulator on the Intel plattform. I'd buy it
just to have fun with it. :)
Gilbert Mella
PL/I developer on IBM VM/CMS, MVS, and OS/2
S/390 BAL
kiy...@ibm.net
>Date: Fri, 19 Apr 1996 05:32:48 GMT
There are lots of ways to get VM and MVS into the home. The cheapest way is the
1985 XT and AT/370 which runs single user CMS. You can run PL/I and the VS-Assembler
on it. I have the hardware and the VM-PC/CMS software running.
The next step up is the FSI OPEN/370. It runs all S/370 mode software including MVS/370,
JES2, VTAM and all the S/370 compilers. It's a small 4381. Info at http://www.funsoft.com
Modern VLSI CMOS hardware, 3.5 PC disks.
You could also get a 9370 or rack mounted ES/9000s. Some of those are being given away,
especially the under 5 MIPS systems but realisticly it would take a few thousand dollars to
get one. 4000 series systems are true give-aways but they're so big and the DASD is
huge.
Then there's the P/390 from IBM, fifty thousand bucks, but it's a full house S/390 mainframe
in a PC file server. Runs everything. You'd need a heck of a consulting income stream or
mainframe product to support this thing. Rex, in Missouri (I think), has one of these but
his product sales and consulting supports the beast.
If you're ever in the Washington DC area, I'll show you my personal VM/CMS and MVS systems.
Cory Hamasaki
Ben Weiner
:In <4l62co$1h...@news-s01.ny.us.ibm.net>, kiy...@ibm.net writes:
:>The MVS precursor, OS/360, was a failure if you consider late, big, slow,
:>buggy, feature poor, overpriced, memory and CPU demanding, a failure.
:>However, compared to PC OS's, OS/360 was an incredible success. It ran well
:>and multi-tasked in 512 K on a .5 MIPS box. On a 1.5 MIPS box, it'll
:>support hundreds of users and run the compiles and production work for a
:>small company, 1,000 persons or so.
:Yes! I did my undergrad projects on an IBM 360/50, MVT, 512K memory and it
:was fast!
:>PC and Unix-heads can't appreciate what a full house 3031 or S/370 168 can do.
:You're 100% right!
:>It is sad to see a real computer like a VM/CMS box get shut down.
[OK, this I can agree with, even if I don't want to use it myself.]
:Cory, I remember my days of struggle at Columbia U learning how to use a...
:Teletypewriter terminal!!!! and the generic Televideo terminals hooked up to
:DEC-20's. I never learned how to use the darn thing and hated the DEC-20.
:I just had to comtemplate my friends at CUNY (City University of New York) using
:VM/SP and MVS/JES3 and how easy it was for them, form their IBM 3278 terminals,
:to send jobs to the MVS machine and get it back on their virtual "RDRs", or how
:they did it interactively on VM/CMS full screen . And I was envious!! What a beauty
:to use XEDIT full screen to write your programs!!. I thought I was not getting my
:money's worth at Columbia.
:The following semester I took a Comp Sci course at CCNY just to get an account
:at CUNYVM. Man what a difference!
:What a difference to use REXX to compose "Exec files" (we now call them script)
:which would create the JCLs needed to spool my PL/I and BAL programs to the
:MVS machine or to do it locally on the VM machine! I never went back to the
:Columbia terminal room.
:When my profs asked me why I was using another University computer to do
:my projects, I told them right away that the Columbia DEC system was crap!!
:Since them I've been in love with IBM VM/CMS and would not work on anything
:different!
AAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHH !!!11!!1!!11!!!1!!
aaaaarrrrrrrrgggggghhhhh!!!!1!!11!!!
gagaaaaaaaaaaahhhh
burble
glorp
You have just been eaten by a grue
:I wished there was a VM/CMS MVS/JES emulator on the Intel plattform. I'd buy it
Keith Pepin/DECwolf
(Ben Weiner) wrote:
-> :Since them I've been in love with IBM VM/CMS and would not work on anything
-> :different!
->
-> AAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHH !!!11!!1!!11!!!1!!
->
-> You have just been eaten by a grue
Erm. :-)
[BELCH]
/YOU/, on the other hand, have just been snacked on by a werewolf :-)
--
Keith Pepin/DECwolf
Obsessed with DEC, lover of lupines
----****####****----
Want to save money on your next phone bill? E-mail dec...@earthlink.net and tell me YOU WANT TO SAVE MONEY!
----****####****----
Today, LOA stands for 'Logged Off Again'
LOA- Southern New England's _WORST_ Internet Service Provider
[The above statement is NOT endorsed by Log On America, Inc--
hence, it is probably true.]
----****####****----
$ SHUTDOWN
%SYSTEM-W-FAL, NUMBR5:: ALIVE!!
Daniel Roganti
Will Scratchley
> I wished there was a VM/CMS MVS/JES emulator on the Intel plattform. I'd buy it
> just to have fun with it. :)
All you need is a P/390 card and you can run VM (or MVS or VSE) on your PC.
Check out http://vmdev.gpl.ibm.com/
Will Scratchley
VM Bigot
Gene Wirchenko
>In <4l5mio$f...@mailgate.lexis-nexis.com>, chu...@meaddata.com (Chuck Greenwald) writes:
>>In article <4l4qqf$g...@info.uci.kun.nl>, J.Hen...@maw.kun.nl (John Hendrickx) writes:
>>|> In article <4kv46a$4...@cnn.Princeton.EDU>,
>>|> ad...@tucson.princeton.edu (Adam J. Thornton) wrote:
>>|> >In article <829586...@sheril.demon.co.uk>,
>>|> >Joseph Chacko <j...@sheril.demon.co.uk> wrote:
>>|> >>I recently read about the IBM OS project as an example of a failed project.
>>|> >
>>|> >No, not really. Slow, bloated, buggy, and a year past due, but ultimately
>>|> >not a failure. You've heard of MVS, right? It's the lineal descendent of
>>|> >OS/360.
>>|> >
>>|> I've used MVS once or twice, but otherwise I've used CMS. Is it also a
>>|> descendent of OS/360 - MVS? Does anyone know when it originated? I'm curious
>>|> about it because it was my first computer system, but our VM mainframe gets
>>|> the axe on January 1 :-(. It's a really primitive system in many ways, but
>>|> seeing it go makes me sentimental.
>>
>>
>>CMS started as the Cambridge (Mass., USA) Monitor System. It was
>>very heavily influenced by Multics and may be viewed as one of
>>Unix's many brothers.
>>
>>-------------------------------------------------------------------------
>>Chuck Greenwald | All opinions are | chu...@meaddata.com
>>Lexis/Nexis | my own. | (513) 865-1020
>>
>Given that the Cambridge Monitor System was running a decade before Unix, it
>would be more accurate to call it Unix's rich uncle, Kindel O'Hara.
>The MVS precursor, OS/360, was a failure if you consider late, big, slow,
>buggy, feature poor, overpriced, memory and CPU demanding, a failure. However,
>compared to PC OS's, OS/360 was an incredible success. It ran well and multi-tasked
>in 512 K on a .5 MIPS box. On a 1.5 MIPS box, it'll support hundreds of users and
>run the compiles and production work for a small company, 1,000 persons or so.
Oh, please! Apples and oranges!
Please give some thought to the differences involved in using
text (almost only) a la mainframe vs. heavy graphics use a la PC.
There are other differences.
I do grant that mainframes have/had better buses, but that's
hardware not software, so OS/360 can hardly take credit for that.
>PC and Unix-heads can't appreciate what a full house 3031 or S/370 168 can do.
Oh, but I can. I used to spend my allowance money on computer
time at the University of British Columbia's IBM 370/168 system.
While I had a lot of fun, I had even more fun some years later on my
TRS-80 Model I as IT WAS MORE FLEXIBLE.
>It is sad to see a real computer like a VM/CMS box get shut down.
True, nostalgia. I remember UBC's CC bragging about the big
upgrade from 4 to 8MB of core! I like to keep in mind how other
systems were implemented. If nothing else, they can serve as bad
examples <G>. The reality is that examining them is often very
instructive.
I am glad that I got involved with computers before micros and
especially before the IBM pc. As a result of my beginnings, I have a
perspective that few seem to have these days.
> OS/2 Software Store http://www.kiyoinc.com
> Kiyo Design Inc Stop in for your free ColorWorks CD and poster.
> 11 Annapolis St. Free ColorWorks CD with each Web order,
> Annapolis, Md 21401 just type "CW CD" in the comments field.
Sincerely,
Gene Wirchenko
C Pronunciation Guide:
y=x++; "wye equals ex plus plus semicolon"
x=x++; "ex equals ex doublecross semicolon"
Dave
>There are lots of ways to get VM and MVS into the home. The cheapest way is the
>1985 XT and AT/370 which runs single user CMS. You can run PL/I and the VS-Assembler
>on it. I have the hardware and the VM-PC/CMS software running.
>The next step up is the FSI OPEN/370. It runs all S/370 mode software including MVS/370,
>JES2, VTAM and all the S/370 compilers. It's a small 4381. Info at http://www.funsoft.com
>Modern VLSI CMOS hardware, 3.5 PC disks.
>You could also get a 9370 or rack mounted ES/9000s. Some of those are being given away,
>especially the under 5 MIPS systems but realisticly it would take a few thousand dollars to
>get one. 4000 series systems are true give-aways but they're so big and the DASD is
>huge.
>Then there's the P/390 from IBM, fifty thousand bucks, but it's a full house S/390 mainframe
>in a PC file server. Runs everything. You'd need a heck of a consulting income stream or
>mainframe product to support this thing. Rex, in Missouri (I think), has one of these but
>his product sales and consulting supports the beast.
>If you're ever in the Washington DC area, I'll show you my personal VM/CMS and MVS systems.
>Cory Hamasaki
You might also keep your eyes open for either a IBM 7437 box (about
the size of a IBM PS/2-model 80), which is a S/370 in a box that
attaches to a PS/2. They run about 750 kips (very rough estimate).
You might also keep your eyes open for an IBM P/370 card, which was
the precursor to the P/390. The P/370 implements the S/370 on a card
which plugs into a PS/2. They run about 3.5-4 MIPs (Again, a very
rough estimate).
You can run VM/ESA-370 Feature on either of these (but can't run
VM/ESA since these are only S/370 systems).
The XT/370 is not a full S/370 implementation (non-priviledged
instructions only), and requires a special VM/PC operating system.
Additionally, it's even slower at about 50 KIPs.
Dave
P.S. Yes, I own an XT/370, and I have used the 7437 and P/370s (and
would dearly love to get my hands on a P/390!).
JMFBAH
<:Cory, I remember my days of struggle at Columbia U learning how to use
a...
<:Teletypewriter terminal!!!! and the generic Televideo terminals hooked
up to
<:DEC-20's.
Are you sure it was a DEC-20? I don't think we ever sold a DEC-20 with
teletypes. B.T.W. My definition of teletypes is an ASR-33. My memories
of those terminals are of frustration. Field Service had me banned from
using them because, if you typed faster than 30-35 words/minute, they
broke. So I had to use ASR-35s, a better terminal, in my opinion. When
there wasn't any -35s available, and I had to use the -33s, I usually
used LINED instead of TECO because it forced me to type slower. The best
video terminals were the VT06s. Remember them, guys? I think JMF had the
very last one (he had all the broken ones in his office to be used for
spare parts). I do believe he was close to tears when that one finally
died and couldn't be fixed.
/BAH
/BAH
Heinz W. Wiggeshoff
Daniel Roganti (daniel....@tip.nl) writes:
> ab...@FreeNet.Carleton.CA (Heinz W. Wiggeshoff) wrote:
>
>
>
> I MISS PL/I
working in
(and APL, REXX ...)
Peter Orbaek
: In article <4kv46a$4...@cnn.Princeton.EDU>, ad...@tucson.princeton.edu (Adam
: J. Thornton) wrote:
: > Read Fred Brooks's marvellous and seminal book _The Mythical Man-Month_.
: > That's the best place to start.
: This book is now out in a new edition, with additional chapters and a
: new forward. I don't have it handy, or I'd include the ISBN & price data.
: My copy is a trade-format paperback.
The ISBN (according to our library system) 0201835959.
It's from Addison-Wesley, 1995.
Happy reading,
- Peter.
Keith Pepin/DECwolf
-> Somebody wrote:
-> <:Cory, I remember my days of struggle at Columbia U learning how to use
-> a...
-> <:Teletypewriter terminal!!!! and the generic Televideo terminals hooked
-> up to
-> <:DEC-20's.
->
-> Are you sure it was a DEC-20? I don't think we ever sold a DEC-20 with
-> teletypes. B.T.W. My definition of teletypes is an ASR-33. My memories
-> of those terminals are of frustration. Field Service had me banned from
-> using them because, if you typed faster than 30-35 words/minute, they
-> broke. So I had to use ASR-35s, a better terminal, in my opinion. When
-> there wasn't any -35s available, and I had to use the -33s, I usually
Heh, I have, well, less than fond memories of an ASR-33... I worked with
one on a pdp-12 in a friend's basement. Well, I type faster than 35 WPM.
My friend was, well, less than happy...
Gilbert Saint-flour
>>See any OS/360 distribution kit, and look at the entries for the
>>PL/1 F compiler (360S-NL-511) or library (360S-LM-512).
You can find the S/360 PLI (F) compiler on the CBT tape, it works in ESA.
You can order the documentation from IBM. For around $50 you get the
language reference, programmer's guide, and reference card.
Just in case anyone is interested.
John Everett
>
>
>Are you sure it was a DEC-20? I don't think we ever sold a DEC-20 with
>video terminals were the VT06s. Remember them, guys? I think JMF had the
>very last one (he had all the broken ones in his office to be used for
>spare parts). I do believe he was close to tears when that one finally
>died and couldn't be fixed.
>
>/BAH
>
Yes, Barb, the -35s were vastly superior to the -33s. I'd forgotten that
there was an ASR-35, most of the ones we had at DEC were KSRs. I liked the
-35 because of the tactile feedback you got from the keys when you typed, a
feature missing from the -33. Remember the -37, the last (I think) of the
Teletype brand terminals offered by DEC; 150 baud, upper/lower case? I don't
believe many of them were delivered before DEC started making DECwriters,
sometimes refered to as "Tom's Terrific Terminal" after Tom Stockebrand.
Jim's VT06 reminded me of the carpenter who had been using the same hammer
for 35 years. Of course he had replaced the handle many times and the head a
time or two. When I returned to DEC in 1977 to port TOPS-10 to the KS-10 I
couldn't find a free VT06 and had to settle for a VT05. I recall they had an
interesting problem that scrambled the screen from time to time. The
"approved" fix was to lift the front of the terminal precisely one inch off
the desk and let it drop. Cleared up the screen every time.
--
jeve...@wwa.com (John V. Everett)
Nick Spalding
>Jim's VT06 reminded me of the carpenter who had been using the same hammer
>for 35 years. Of course he had replaced the handle many times and the head a
>time or two.
When I heard that one it was George Washington's axe!
Keith Pepin/DECwolf
-> When I returned to DEC in 1977 to port TOPS-10 to the KS-10 I
-> couldn't find a free VT06 and had to settle for a VT05. I recall they had an
-> interesting problem that scrambled the screen from time to time. The
-> "approved" fix was to lift the front of the terminal precisely one inch off
-> the desk and let it drop. Cleared up the screen every time.
Remember the Apple ///? My school got one, it didn't work. I called Apple
tech support [I was responsible for all computer equipment, and I was in
4th grade. Not too damned shabby, eh?] The dialog:
"Ok, Keith, here's what you do. Take the monitor off, put it to the side."
"Done. Next?"
"Pick it up 4 inches."
"Ok, this is heavy. What now?"
"Let go..."
Worked! Apparently, that was to reseat the chips. We ended up doing it
every 2-3 days. Remember the Mac Plus? A friend had one, and I was working
on it with him. He said, "Damn, there goes the screen again" as it faded
off. I picked it up 4 inches, and dropped it... Worked fine. When I had a
pdp-8, I had trouble with the programmer's console. I replaced half the
chips, and socketed the replacements. Two weeks later, six friends and I
are lifting the cabinet four inches and another friend of mine, who
actually bothers to put 3.5" disks in the cases, came in just in time for
the ritual drop....
I didn't hear the end of THAT for about a year!!
Joe Morris
> B.T.W. My definition of teletypes is an ASR-33. My memories
>of those terminals are of frustration. Field Service had me banned from
>using them because, if you typed faster than 30-35 words/minute, they
>broke. So I had to use ASR-35s, a better terminal, in my opinion.
And in the opinion of Teletype Corporation as well. I don't recall the
exact numbers, but the TTY33 wasn't designed for a high duty cycle,
whereas the TTY35 never seemed to have any problems when used continuously -
and the price difference between the two reflected this. Given typical
budgets (especially in the academic world) this meant that the vast
majority of the Teletype boxes used for time sharing were TTY33s.
Obnitpick: "Teletype" is a trademark of the Teletype Corporation; the generic
term would be "teleprinter."
Joe Morris / MITRE
Nick Spalding
>Obnitpick: "Teletype" is a trademark of the Teletype Corporation; the generic
>term would be "teleprinter."
I believe Teleprinter may also be a tradename, of the English firm of
Creed. The true generic is possibly Teletypewriter!
JMFBAH
<Obnitpick: "Teletype" is a trademark of the Teletype Corporation; the
generic
<term would be "teleprinter."
Hmmmm....I don't think I ever heard that term (teleprinter) except maybe
in using a hardcopy terminal as an alternative LPT. Maybe Teletype became
the generic term like Kleenix did for paper tissues. TOPS10 designated
terminal devices as TTYs. I always pronounced this word as "tittys".
This reminds me of a risque story. Is the telling of risque stories
allowed in this newsgroup?
I apologize for X-posting; AOLs software doesn't allow me to edit the
"send to" list.
/BAH
Ross Alexander
>Heh, I have, well, less than fond memories of an ASR-33... I worked with
>one on a pdp-12 in a friend's basement. Well, I type faster than 35 WPM.
>My friend was, well, less than happy...
Just got two, in fair shape, along with the full exploded drawings and
ten years worth of service bulletings, in the original Teletype grey
bindings. Have to go hack up a current loop for the unix box
sometime. The real plan is to build an ascii to baudot convertor and
get one on air as part of my amateur RTTY station.
BTW, the one in `better' condition is clearly labeled "IMP console".
So I own a piece of history, much to my surprise.
regards,
Ross
--
Ross Alexander, ve6pdq -- (403) 675 6311 -- r...@cs.athabascau.ca
Robert Bonomi
>On 22 Apr 96, Joe Morris /MITRE wrote:
><Obnitpick: "Teletype" is a trademark of the Teletype Corporation; the
>generic
><term would be "teleprinter."
>
>Hmmmm....I don't think I ever heard that term (teleprinter) except maybe
>in using a hardcopy terminal as an alternative LPT. Maybe Teletype became
>the generic term like Kleenix did for paper tissues. TOPS10 designated
"Tele-Type" is (_still_) a trademark of Tele-Type Corp.
'tele-typewriter' is the generic name for a SR (send-receive) device.
If it had a paper-tape(!) reader/punch on it, it was an ASR (A for automatic).
If the only input device was a keyboard, it was a KSR.
a 'tele-printer' is the generic name for a RO (receive only) unit. No keyboard.
Wulf Losee
>
<snip>
>However,
>compared to PC OS's, OS/360 was an incredible success. It ran well and
>multi-tasked in 512 K on a .5 MIPS box. On a 1.5 MIPS box, it'll support
>hundreds of users and run the compiles and production work for a small
>company, 1,000 persons or so.
>
>
>
I agree that it's sad to seem them go. I cut my teeth on VM/CMS.
However, Carl Amdahl, Gene Amdahl's son, claimed that the ability of the 360
architecture to support huge numbers of users was less to do with the OS than
with the bus architecture (which off-loaded much of the I/O from the CPU).
Excuse me if I've garbled his message somewhat. I saw the spiel several years
ago, and I don't remember the details of his statement. But Carl Amdahl has
made a successful business out of his NetFrame superservers which adapt a
mainframe-style bus architecture to an Intel CPU.
We'll soon have a 1000 users running on a NetWare superserver at our site, but
it will be handling mostly file and print sharing.
--Wulf
******************************************
^-^ Wulf Losee
/ = Network Analyst
, / | J. Paul Getty Trust
(( / } \\ Vox: 310.656.5817
=/ = \\ Email: wlo...@getty.edu
******************************************
Keith Pepin/DECwolf
Bonomi) wrote:
-> "Tele-Type" is (_still_) a trademark of Tele-Type Corp.
->
-> 'tele-typewriter' is the generic name for a SR (send-receive) device.
-> If it had a paper-tape(!) reader/punch on it, it was an ASR (A for
automatic).
-> If the only input device was a keyboard, it was a KSR.
->
-> a 'tele-printer' is the generic name for a RO (receive only) unit. No
keyboard.
The best was a friend of mine, who insisted that ASR was 'asynchronous
serial receive'. He also thought that the IBM AS/400 got it's name because
it had 400 asynchronous serial ports. After a practical but rather dumb
joke I pulled, he still refers to the 'pdp' as 'pretty dumb processor'...
OTOH, thanks to my sense of humor, he thinks SCSI stands for $hitty crap
through $hitty interface. And has referred to it as such at computer
stores, while in my presence. Needless to say, I don't go anywhere with
him anymore...
Keith Pepin/DECwolf
Alexander) wrote:
-> dec...@earthlink.com (Keith Pepin/DECwolf) writes:
-> >Heh, I have, well, less than fond memories of an ASR-33... I worked with
-> >one on a pdp-12 in a friend's basement. Well, I type faster than 35 WPM.
-> >My friend was, well, less than happy...
->
-> Just got two, in fair shape, along with the full exploded drawings and
-> ten years worth of service bulletings, in the original Teletype grey
-> bindings. Have to go hack up a current loop for the unix box
-> sometime. The real plan is to build an ascii to baudot convertor and
-> get one on air as part of my amateur RTTY station.
->
-> BTW, the one in `better' condition is clearly labeled "IMP console".
-> So I own a piece of history, much to my surprise.
Congratulations, that IS a heck of an acquirement. I tell you, I'd liked
to have gotten the wooden RK05 from the Boston Computer Museum. The jerks
tossed it out, as I understand. They didn't even put it up for sale, just
tossed it...
BTW, if anyone is interested in purchasing or selling veteran DEC
equipment [like spare parts for that ancient and precious controller, or
even another controller], e-mail me with what you have to sell or want to
buy. I can arrange the transactions [I now have a few connections in this
area].
Cripes, if I had the room to store a pdp-1.... :-)
Mark Freid
Pepin/DECwolf) wrote:
>Remember the Apple ///? My school got one, it didn't work. I called Apple
>tech support [I was responsible for all computer equipment, and I was in
>4th grade. Not too damned shabby, eh?] The dialog:
>"Ok, Keith, here's what you do. Take the monitor off, put it to the side."
>"Done. Next?"
>"Pick it up 4 inches."
>"Ok, this is heavy. What now?"
>"Let go..."
>Worked! Apparently, that was to reseat the chips. We ended up doing it
>every 2-3 days. Remember the Mac Plus? A friend had one, and I was working
>on it with him. He said, "Damn, there goes the screen again" as it faded
>off. I picked it up 4 inches, and dropped it... Worked fine. When I had a
>pdp-8, I had trouble with the programmer's console. I replaced half the
>chips, and socketed the replacements. Two weeks later, six friends and I
>are lifting the cabinet four inches and another friend of mine, who
>actually bothers to put 3.5" disks in the cases, came in just in time for
>the ritual drop....
>
>I didn't hear the end of THAT for about a year!!
Heheh, I'm sort of a small-time collector of "classic" video consoles
& computers. from what I understand, that is a documented fix for the
first release Apple ///. Basically, the system was poor altogether,
especially in the field mentioned, as well as backward compatibility
and other factors. Apple eventually got around to fixing the problems,
and began remanufacturing "fixed" ones, but the reputaion and sales
were so bad, that the project was scrapped for the more popular Apple
//. I was hunting for old systems at thrift shops last month, and one
near my home actually has an Apple ///! Too bad I don't know if its a
"old" or "fixed" version. :) They also want 70 bucks for it. too much
(for now...)
-Mark
=============================================
Mark Freid ("Canuss") Wolf, Cartoonist, Loony
=============================================
Web: http://www.av.qnet.com/~canuss
Email: can...@qnet.com
wo...@worldnet.att.net
"Canuss" on GEnie's Beastie Board & FurryMUCK!
==============================================
Bob Manners
In reply to jcmo...@mwunix.mitre.org's message:
| And in the opinion of Teletype Corporation as well. I don't recall the
| exact numbers, but the TTY33 wasn't designed for a high duty cycle,
| whereas the TTY35 never seemed to have any problems when used continuously
| and the price difference between the two reflected this. Given typical
| budgets (especially in the academic world) this meant that the vast
| majority of the Teletype boxes used for time sharing were TTY33s.
Any idea what the price difference was? Or indeed what the 33 cost?
These old 33's seem to last. I have an ASR33 in my flat which is still
running OK, having been rescued from exile in a garage! They seem
pretty well engineered, although I'd image that servicing neglect is
rewarded by dramatic failures.
I've also got a KSR43, but that is much less interesting :)
Also, whatever happened to the Teletype corp? Were they bought out by
some monolithic organisation or what?
--
----------------------------------------------------------------------------
Robert Manners Osney Laboratory
r...@swift.eng.ox.ac.uk Dept of Engineering Science
University of Oxford
01865 288762
Computer Museum: http://swift.eng.ox.ac.uk/rjm/museum.html
Also: http://swift.eng.ox.ac.uk
"... The sound of happy couples, coupling happily in the dark.
Whilst you and I sat down to tea, I remember you said to me,
that no amount of poetry could mend this broken heart,
but you can put the hoover round if you want to make a start" Billy Bragg
----------------------------------------------------------------------------
Carl Lowenstein
|
| In reply to jcmo...@mwunix.mitre.org's message:
|
| | And in the opinion of Teletype Corporation as well. I don't recall the
| | exact numbers, but the TTY33 wasn't designed for a high duty cycle,
| | whereas the TTY35 never seemed to have any problems when used
continuously
| | and the price difference between the two reflected this. Given typical
| | budgets (especially in the academic world) this meant that the vast
| | majority of the Teletype boxes used for time sharing were TTY33s.
|
| Any idea what the price difference was? Or indeed what the 33 cost?
| These old 33's seem to last. I have an ASR33 in my flat which is still
| running OK, having been rescued from exile in a garage! They seem
| pretty well engineered, although I'd image that servicing neglect is
| rewarded by dramatic failures.
From my memory, the price that DEC charged its PDP-8 customers for an ASR33
was about $1600. A KSR35 (heavy-duty, no reader/punch) listed at perhaps
twice that. Of course, in those days your basic PDP-8 with 4k of core
went for about $18000, so the ASR33 was only about 10% of the total price.
carl
--
carl lowenstein marine physical lab u.c. san diego
clowe...@ucsd.edu
Jim Stewart
These prices are accurate to the best of my memory. The 35 was at least
twice the cost of the 33. Despite the high cost of the processor, I saw
one 35 for every 20 33's here in the Bay area. The only trick to keeping
the 33's running was to keep them well oiled. I can recall carrying
clutches and mainshafts in my tool kit to service them.
-Jim Stewart
Eric Smith
In article <RJM.96Ap...@swift.eng.ox.ac.uk> r...@swift.eng.ox.ac.uk (Bob Manners) writes:
> Also, whatever happened to the Teletype corp? Were they bought out by
> some monolithic organisation or what?
They were bought out by THE monolithic organization, AT&T. At least it
was back then (pre-divestiture).
For a while they marketed things like the AT&T Teletype 5620, but before
too long the Teletype name was dropped.
Eric
J. Chris Hausler
Jim Stewart <jkm...@dsp.com> writes:
>These prices are accurate to the best of my memory. The 35 was at least
>twice the cost of the 33. Despite the high cost of the processor, I saw
>one 35 for every 20 33's here in the Bay area. The only trick to keeping
>the 33's running was to keep them well oiled. I can recall carrying
>clutches and mainshafts in my tool kit to service them.
"well oiled", I tell people the only reason I keep my two 33's
running (one KSR, one ASR) is that computing just would not be
the same without that unique smell of hot oil you get from a
teletype. It somehow helps one think........;-> Chris
Patrick Klos
In article <RJM.96Ap...@swift.eng.ox.ac.uk> you write:
>Any idea what the price difference was? Or indeed what the 33 cost?
>These old 33's seem to last. I have an ASR33 in my flat which is still
>running OK, having been rescued from exile in a garage! They seem
>pretty well engineered, although I'd image that servicing neglect is
>rewarded by dramatic failures.
Sorry, I don't know anything about prices. I have an ASR33 myself, and
was wondering if you know of anyplace to buy the rolls or paper or paper
tape for the beast? The paper itself probably isn't too hard to find
a substitute for, but I'm sure it's harder to find a replacement for
the paper tape rolls.
>I've also got a KSR43, but that is much less interesting :)
>
>Also, whatever happened to the Teletype corp? Were they bought out by
>some monolithic organisation or what?
I'd be interested to hear what happened to them also.
Patrick
============================================================================
Patrick Klos Email: kl...@klos.com
Klos Technologies, Inc. Voice: (603) 424-8300
604 Daniel Webster Highway FAX: (603) 424-9300
Merrimack, New Hampshire 03054 Web: http://www.klos.com/
============================================================================
Ross Alexander
er...@goonsquad.spies.com (Eric Smith) writes:
Yes; I've got a 5310 sitting lonely and unloved in the back room.
It's actually a nice little piece of work, but the setup interface is
from hell - far more complicated than I can imagine ever being useful.
BTW, further to those two doc sets for x33's, what I exactly have is
Bulletin 1184b changelevel 15, 1972:
_33 Page Printer Set (ASR, KSR, and RO) Parts_
About 90 pages of exploded drawings and parts lists. Straight double
sided printing, no fold outs.
Bulletin 310b Vol 1, changelevel 2, 1972:
_Technical Manual, 33 Teletypewriter Sets RO, KSR, ASR_
The jackpot. 250 pages of drawings, specs, procedures. waveforms,
schematics, photographs, you name it and it's in there somewhere.
Almost all double sided, one fold-out page.
I estimate that to copy 310b would set me back around C$10, plus
postage. The scale in the bio lab says it weighs 985 grams. 1184b
would probably come in for about C$4 + postage. I can't guarantee
turnaround times beyond `when I get a chance', and I'll probably
wait till I have at least 5 in the queue.
JMFBAH
On 25 Apr 96, J. Chris Hausler <jcha...@delphi.com> wrote:
< that computing just would not be the same without that unique
< smell of hot oil you get from a
<teletype. It somehow helps one think........;-> Chris
Ahhh. I had forgotten about the smell...Thanks for the reminder. Which
reminds me...
My group, Tape Prep, had the job of editing the TOPS10 monitor every
Tuesday. We would traipse upstairs from 12-1 to 3-5 to edit the marks
that the members of the TOPS10 Monitor Group had made during the last
week. Then two guys would take the edited sources and make all the
various flavours of monitors that were needed to run for development,
testing, etc.
The terminals we worked on were 35s. There was a monitor guy who would
take out his frustration on the poor 35s with the direct result of the
windows breaking. One always knew if this man had been in the room by the
presence of the glassless 35s. Field Service finally got so fed up with
the hopeless job replacing these windows, that they stopped. As a result,
whenever one of these 35s got serviced by a Field Service guy who was
liberal with his oilcan, one could get a waterless oil bath. Just
another hazard of the mill in those good ol' days.
/BAH
Dick Maliska
jeve...@wwa.com (John Everett) mentioned:
>Jim's VT06 reminded me of the carpenter who had been using the same hammer
>for 35 years. Of course he had replaced the handle many times and the head a
>time or two. When I returned to DEC in 1977 to port TOPS-10 to the KS-10 I
>couldn't find a free VT06 and had to settle for a VT05. I recall they had an
>interesting problem that scrambled the screen from time to time. The
>"approved" fix was to lift the front of the terminal precisely one inch off
>the desk and let it drop. Cleared up the screen every time.
Yes, and another item about the VT05 was the annoying beep you would
get whenever a character was displayed a position 72 on the line.
Presumably this was for the touch typist to be reminded that the end
of the line was near. Larry Rogers and I then created the NO BELL
prize which provided a cut etch on the circuit board to prevent the
"bell" from ringing when one was not typing. (We still needed the
bell for other things.)
Regards,
Dick Maliska (Rmal...@cris.com)
COUNTRYSIDE SOFTWARE SERVICES. (508) 692-3124
Hardware/Software Support, Networking and Application Development.
"These are the days of miracles and wonder. This is the long distance call."
- Paul Simon
Barry O'Grady
dec...@earthlink.com (Keith Pepin/DECwolf) wrote:
:OTOH, thanks to my sense of humor, he thinks SCSI stands for $hitty crap
:through $hitty interface. And has referred to it as such at computer
:stores, while in my presence. Needless to say, I don't go anywhere with
:him anymore...
I take it you mean humour, but how do words starting with $ relate to
an acronim without $ signs?