Wanted: 1802 CP/M (or, PL/M 1802)
Bill
formerly ~ $70,000(?) RCA Cosmac Development System.
Maybe somebody's already built this wheel?
Obviously with PL/M-1802 the CCP and BDOS can
be compiled; then just need to write a CBIOS for some
version of COSMAC system. I do find a CDP18S651
floppy controller in the 'rack'. The main deviation from
'normal' would be the 700+ Kbytes of CMOS memory
available, and how best to use it. (imagine, in 1977!)
Any interest or help out there?
Bill
Bill
wrote:
>Trying to locate some tools to allow for CP/M 1802 on a
>formerly ~ $70,000(?) RCA Cosmac Development System.
Just to correct something.....
Probably due to the cost, I'm pretty sure not many have
ever even seen one of these systems. It would have been
unusual in the extreme to find it classed with other personal
computers.
Nevertheless, the 1802 chip was a common device for the
home hobbiest/experimenter.
Since folks generally understand that Intel really built the
first 4004, 4040, 8008, 8080 etc COMPUTERS, whatever
they called them, development systems, whatever, well in
this case the Development System I'm talking about is
exactly that. Not, what you MIGHT get if you added this
piece (TTY interface) or that (Pertect floppy drives, the
EXACT same box also sold by MITS for their Altair).
This was a complete ready to run computer system.
Unlike the Altair.
Trouble in that day was, it cost more than most houses!
It uses sort of odd 22 pins x 2 card edge connnectors,
just exactly like the one on the hobbiest Cosmac boards.
RCA called them Microboards. The RCA Microsystems
manual SSD-270 pretty much enumerates all of them,
complete with jumper settings and schematics. Nearly
600 pages.
Also details available software. No CP/M, though.
The racks for the cards came in 5-card and 22 card sizes.
There were plain front panels, and more elaborate ones
that allowed inputting data, single-stepping, and numeric
display of program pointer, Op codes, and data.
I see there was a Concurrent Pascal for multitasking!
Pretty amazing, but right in line with what Ed Roberts
has been saying for thirty years now - not much new.
This was THE FIRST of the modern line of CMOS
based microcomputers. Bar none. All the others
were N-Mos or P-Mos, both totally obsolete now.
Curiously, I seem to recall RCA had to license some of
the Complementary technology from Texas Instruments.
Guess I'll have to post photos one of these days.
Bill
Bill
wrote:
>Far as I know PL/M was never done on 1802.
Maybe if instead of telling me what you don't know, you simply
asked to see what I have that even suggests its existence
we might all get someplace further along.
> CP/M is not a good fit for it either.
Well there's really only one way to find out, isn't there?
>By Time the 1802 development system cme out there wer plenty of
>packaged CP/N systems and others (like NS* Horizon).
Kentucky Fried Computers? Surely you jest. Hobbiest toys.
KFC merely slapped purchased parts together. Like anybody else.
RCA designed, developed, INVENTED a lot of their own stuff.
Try this on for size:
http://www.decodesystems.com/cosmac/case-history.html
7-1/2 inch disk drives? $5 media? 1.4Mb capacity? UARTs?
DMA? Memory Mapped Video? Boot Sectors? a DOS? ??
Huh? NINETEENSEVENTYFOUR????? And, it WORKED?????
I 'spose, unless you were around then, that article will go in one ear
and out the other. Keep reading it until it sinks in. That was WAY
ahead of anything going on in the hobbiest community at the time.
Anyway, didn't Kildall write PL/M? How could RCA come up with
PL/M 1802 and he not know anything about it? Illogical. We also
know Tom Pittman did a LOT of work on 1802 stuff. Kildall often
described him as a friend, as well as co-worker. Those guys HAD
to have had something to do with it.
>Their OS was not like CP/M but they had a DOS.
And, they had a Concurrent Multitasking one! (well, okay,
calling Pascal an ''operating system'' is sorta like calling
StandAloneDIsk Basic one - there's one inside there
somewhere, and you don't need a separate one, but .....)
I think we (okay, maybe only the Frenchman and me) would
really like to find out more about this elusive PL/M 1802.
Bill
Tarkin
> On Mon, 19 Mar 2007 22:36:55 GMT, no.s...@no.uce.bellatlantic.net
Careful, Bill. She *was* around then, and has been building
her own CP/M systems for quite some time.[1]
I would read up on her past postings here, before you
reject her opinions and try and rebuff her.
I would guess she's forgotten more about CP/M, the Z80,
and S-100 systems than you or I will ever know.
I ca also personally vouch for the fact the she has
been extraordinarily helpful here and in other groups;
and I would suggest to you that your comments can be
taken as less than flattering, and possibly even insulting.
[1] I believe she's been doing so since the first time
you saw The Empire Strikes Back, and probably longer
HTH & HAND,
Tarkin
John Crane
"Bill" <Bi...@SunSouthWest.com> wrote in message
news:hha103d0h64avdt21...@4ax.com...
> Huh? NINETEENSEVENTYFOUR????? And, it WORKED?????
>
> I 'spose, unless you were around then, that article will go in one ear
> and out the other. Keep reading it until it sinks in. That was WAY
> ahead of anything going on in the hobbiest community at the time.
>
Please bear in mind that there really wasn't much "invented" in the early
hobbiest computer industry anyway. Those guys just made it cheaper for us
normal Joe's to afford. And they took steps copying the work of the "big
guys" - so they were always a few steps behind. One only has to look at the
DEC PDP machines of the late sixties to see some of the more obvious
influences.
And BTW, I still have my old Netronics ELF II running RPN Basic. I learned
a lot tinkering with it back in high school.
-John
Bill
wrote:
>Curious, do you know just how much 4Kx8 of ram would cost in 1974?
No, do you?
>Hint, 2102 1uS part ran for $6.00 in Quantities over 1000. The 8008
>system I was doing then used 4k of ram for a mere $12 in small
>quantities and we were getting a good price but then again the 8008
>was $180 a copy.
I think I'll look it up ....hmmm how about YOU place your price dot
on this chart and tell us what your cost was again?
http://www.jcmit.com/mem2006.htm
I DO know the Cosmac System I have includes two 20 - card racks
each card holding 16Kbytes of CMOS RAM. That's 320K Bytes, times
two, or a total sixHUNDREDforty K bytes PLUS a few more memory
cards in the main cpu rack putting available RAM over 700Kbytes.
Each of the 16K cards listed in the hundreds of dollars, you do the
math. At say $300 each, that's twelve thousand just for the 40 cards.
I don't know what they cost - maybe it was only $125 each. Whatever.
Didn't I say not many of these systems have turned up in the thirty
years since? Try to find one! I found a dozen Altairs in the time I've
been looking. And that's not counting those being listed on eBay.
Since I was in the Twin Cities area at the time, and as far as I know
MOST of the core memory people were scattered in that area, to
be close to 3M, I was VERY familiar with memory costs. And watched
as solid state drove magnetic core out of business. You assume much,
unfortunately you are wrong about as often as you are right.
MY estimations of memory cost in 1975 seem to fit the chart fairly
well. Do you have a chart to offer, perhaps a sales invoice,
SOMETHING? (other than an obviously failing memory)
And, just for laughs, click the link at the bottom and check out the
price for 4K Bytes of Altair board, 1975. Disagree with that, too?
Bill
Bill
wrote:
>FYI: RCA had a bad case of NIH (NOT INVENTED HERE) and
>was also the first with CMOS logic (4xxx series). The COSMAC
>archetecture was done with 4000 series logic before it was
>reduced to silicon. It's design was based on a simplified version
>of the RCA mainframe.
And your point is?
Have you any (real) idea how many CPUs were done in 2901 bit-slice
and TTL before THEY were ''committed to silicon''? So what? DEC
stopped at the bit-slice versions, and called them PDP-whatevers.
The problem, in retrospect, wasn't that something ''wasn't invented
here'' by RCA, it was that they were trying to do far too much for any
one person to mentally integrate into a single company's product
lines. They went off in a dozen directions at once.
We were early adopters of 4000 series CMOS, before they had any
protection diodes on their inputs. Dangerous stuff, 'in that day'.
Bill
Bill
wrote:
>First off there is no CP/M for 1802, it's off topic for this group.
Lemme get this straight, oh great godess of all things ancient
Who appointed YOU arbiter of what existed or didn't, and as
moderator of this newsgroup?
Kildall said he wrote CP/M ONLY to support HIS choice of
programming language for microprocessors. In addition to the
8080 Intel chip and the 2605 Signetics chip, that APPEARS to
include the RCA 1800 series, however so far I have not found
a mention of it FROM KILDALL. Doesn't mean there isn't one,
maybe he signed a non-disclosure and wouldn't talk about it.
RCA certainly published information about THEIR PLM 1800.
CP/M is the OPERATING SYSTEM of the PL/M LANGUAGE.
What part of this is off-topic in a CP/M newsgroup?
As it happens, UNIX is the operating system of the 'C' language.
So, in fact, anything having to do with 'C' is OFF TOPIC here.
Go soak your head.
Bill
Al Kossow
> List all the PDP whatevers that used 2901 bitslice. It's a very short
> list even for protos.
KS-10 (2020), KT-10 (Minnow), and 11/730
DEC used 2901s in lots of other places (FPP, CIS ,periph ctlrs), but
not in CPU data paths.
I'm not going to bother commenting on the rest of the rant.
--
Posted via a free Usenet account from http://www.teranews.com
Bill
wrote:
>On Thu, 22 Mar 2007 16:18:52 -0500, Bill <Bi...@SunSouthWest.com>
>wrote:
>>I think I'll look it up ....hmmm how about YOU place your price dot
>>on this chart and tell us what your cost was again?
>>
>>http://www.jcmit.com/mem2006.htm
>
>My number are in agreement with the chart allowing for local suppler
>pricing (schwebber electronics.).
I don't think so. Place your alleged price dot
I think you'll find your 'recollection' off by about five years.
Like this: Coordinates: price = ?? Year = 1975
You obviously didn't look. Assumptions, again?
The Altair 4k 2102 board cost $195 in October.
That's roughly $50,000 a megabyte. Or, $37,500 in
Altair memory dollars for this 3/4 megabyte COSMAC.
And while I'm sure you could buy parts and make boards
to stuff them into for less, the point of all this is ready to
run 8 bit computing, not how to lay out boards.
Bill
Bill
wrote:
>On Thu, 22 Mar 2007 16:33:21 -0500, Bill <Bi...@SunSouthWest.com>
>wrote:
>>We were early adopters of 4000 series CMOS, before they had any
>>protection diodes on their inputs. Dangerous stuff, 'in that day'.
>First system used over 100 pieces of 4000, other than latchup it was
>good stuff.
Seems to me, they referred to it as the SCR effect. Put a crowbar
across the power supply rails. Chips sometimes got a little warm;
one or two maybe blew themselves to pieces. You'd go to the spot
on the board where they'd used to be, maybe you'd find a couple
legs still in their holes, the rest of the chip? Who knows? Powder.
Bill
glen herrmannsfeldt
(snip)
> I started using the 1802 (and still do) and built an expanded
> system based on the ELF and Evaluation Kit (seem MPM203)
> back in '76 the chip (white ceramic) is date coded 7624
> (24th week 1976). Have all the docs. I even had the task
> of calling on the NJ marketing and engineering group back in
> late '70s. Signed a few NDAs back then.
I believe it was the 1802 that the ICT (Integer
Cosine Transform) was developed to run on:
http://tmo.jpl.nasa.gov/progress_report/42-119/119M.pdf
>>I think we (okay, maybe only the Frenchman and me) would
>>really like to find out more about this elusive PL/M 1802.
> Should you find it, make it public. I'll bet it was a cross complier
> that ran on PDP-10 under TOPS10, that was the fashon then. Far
> too much RCA literature and design info was lost to time.
That was popular for many years, including I believe the original
Microsoft BASIC for the Altair 8800. In later years people I
knew moved over to the LSI-11.
-- glen
Tom Lake
>> that ran on PDP-10 under TOPS10, that was the fashon then. Far
>> too much RCA literature and design info was lost to time.
>
> That was popular for many years, including I believe the original
> Microsoft BASIC for the Altair 8800. In later years people I
> knew moved over to the LSI-11.
The original Altair BASIC (MicroSoft, as it was originally called, didn't
exist then) was run on an 8080 simulator on a DEC until Gates, Allen
and Davidoff (Remember Monte? He wrote the floating-point routines
in Altair BASIC) could get access to a real Altair.
Tom Lake
roch...@laposte.net
-----------
The famous polemist of the comp.os.cpm Newsgroup, "Wild" Bill,
sent me (why me?) 3 JPEG files holding the contents of pages
515, 516, and 517 of an unknown book. Apparently, he is unable
to retype the following few paragraphs so, for the sake of the
discussion, I retyped them (I am too kind):
Sofware -- CDP18S839
PLM-1800
High-Level-Language Compiler
The PLM-1800 High-Level Language Compiler, CDP18S839, provided
on a diskette, is a software package designed to accelerate
program development on the COSMAC DOS Development System (CDS
III) CDP18S007V1 or CDP18S007V3, and on the COSMAC Development
System IV CDP18S008V1, V3, V5, and V7. It has features similar
to those of many well-known high-level languages, such as PL/1,
ALGOL, and Pascal. Use of the PLM language encourages structured
programming and, hence, provides easy readability and
maintenance. Its scoped procedures and control structures also
support modular programming.
The PLM Compiler automatically creates code for complex
conditions and signed sixteen-bit arithmetic expressions, and it
performs systematic register and storage allocations. Because of
these features, the programmer has more time to concentrate on
the application requirements.
The PLM Compiler also supports CDP1802 features. It contains
built-in functions such as shift operations (SHL, SHR, SCL,
SCR), data conversions (LOW, HIGH), and declaration-based
information functions (ADDR, LENGTH). Other built-in functions
or predeclared variables (Q, MEMORY, DMAPTR, STACKPTR, EF1, EF2,
EF3, EF4, CARRY) allow access to CDP1802-based hardware. Data
transfers through the I/O ports are supported by INPUT and
OUTPUT. The interrupt mechanism is programmable with the
INTERRUPT attribute for procedures and the ENABLE and DISABLE
statements. A built-in procedure, TIME, allows time delays based
on the microprocessor clock.
Code written in PLM may be integrated with assembly code through
the Compiler's in-line-code feature. In addition, the Compiler
produces CDP1802 assembly code that can be combined with other
assembly-time code.
PLM operates directly with the COSMAC Development System
CDP18S008. When used with a CDP18S007 COSMAC Development System,
PLM requires 60 kilobytes of read-write memory and a data
terminal or console. Required software is the CDOS System
Diskette, version 2.2 or later (supplied with the CDP18S007),
and the PLM-1800 High-Level-Language Compiler on diskette
CDP18S839. Documentation is provided with the CDOS Development
System and with the PLM-1800 High-Level-Language Compiler
diskette.
Features of PLM
---------------
The features of the PLM-1800 High-Level-Language Compiler
CDP18S839 include the following:
Data Description:
CONSTANTS - decimal, hexadecimal, octal, binary numbers, and
strings of characters.
VARIABLES - two types; may be aggregated into arrays or
initialized.
BYTE - 8-bit value
ADDRESS - 16-bit value; may contain the address of
another variable.
EXPRESSIONS - permit the following operations:
arithmetic +, -, *, /, MOD
logical AND, OR, XOR, NOT
equality and ordering =, <>, <, >, <=, >=.
Statement Description:
ASSIGNMENT - allows replacement of variable's value by
evaluation of an expression; multiple assignments are
possible in one statement.
IF..THEN..ELSE - allows execution of a group of statements
based on a condition. IF statements may be nested.
DO..END - allows execution of a group of statements.
ITERATIVE DO - allows looping based on a iterative variable
whose increment is controllable with an optional BY
clause.
DO..WHILE - allows looping based on a condition.
PROCEDURE - contains executable instructions and local variable
declarations. Procedures may be recursive if declared
with REENTRANT attribute. Procedures may take on
function attribute.
CALL - subroutine invocation.
GO TO, GOTO - branching capability to labels within scope rules
of the language.
Compiler Features:
In-line assembly code capability
Output listing controls
Assembly code output
Operating with PLM
------------------
After a program is generated in the PLM language, the first step
for using the Compiler is to place the PLM diskette in disk
drive 0. The user then invokes compilation of the file by typing
PLM fname.ext:x
where fname.ext is the user's file name and x is the drive. If
errors occur during compilation, they are transmitted to the
development system terminal device, as well as to an output file
of PLM source code interlisted with CDP1802 assembly code.
Another output file equating assembly names and PLM names is
also generated by the Compiler. The error messages indicate the
nature of the error, the number of the line in which the error
occurred, and where in the line the error was detected.
A program development cycle using the PLM-1800
High-Level-Language Compiler is given in Figure 1. The Compiler
accepts source code written in the PLM language, and generates
the equivalent assembly code that can subsequently be assembled
into CDP1802 executable code.
+----------+
|PLMLIB ASM|
| program |
| library |
+----------+
|
+-----+ | +-------+ +----+
/ \ +--------+ V / \ +----------+ / \
| PLM |-->| CDP1802|----->|Assembler|-->| CDP1802 |-->|CDSBIN|
\compiler/ |assembly| ^ \ (ASM8) / |executable| \ /
+------+ | code | | +-------+ | code | +----+
^ +--------+ | +----------+ |
| | V
+------+ +--------+ +--------+
| PLM | | Other | |Absolute|
|source| |assembly| | binary |
| code | | code | | load |
+------+ +--------+ | module |
^ +--------+
|
+------+
/ \
| Editor |
\ /
+------+
Figure 1. Block diagram of program development cycle
using the PLM Compiler.
Sample program
--------------
A sample program using PLM is given in Figure 2. This program
will sort an array by means of a method called "bubblesort".
DO;
/* This is a bubblesort program *.
DECLARE A(10) ADDRESS INITIAL
(33,10,99,60,162,3,3,272,98,2);
DECLARE (I,SWITCHED,J) BYTE, TEMP ADDRESS;
SWITCHED = 1;
DO WHILE SWITCHED = 1;
SWITCHED = 0;
DO I = 1 TO 9;
J = 1 + 1;
IF A(I) > A(J) THEN
DO;
SWITCHED = 1;
TEMP = A(I);
A(I) = A(J);
A(J) = TEMP;
END;
END;
END; /* of WHILE */
END;
/* Now completed scan without switching */
EOF
Figure 2. PLM "bubblesort" program
Literature
----------
Further information on the PLM-1800 High-Level-Language Compiler
CDP18S839 is given in the "User Manual for the RCA COSMAC
PLM-1800 High-Level-Language Compiler", MPM-239.
Information on the RCA COSMAC Development System IV CDP18S008V1,
CDP18S008V3, CDP18S008V5, and CDP18S008V7 is given in two
manuals "Operator Manual for the RCA COSMAC Development System
IV CDP18S008", MPM-235, and "Hardware Reference Manual for the
RCA COSMAC Development System IV CDP18S008", MPM-236.
Information on the RCA COSMAC DOS Development System CDP18S007V1
and CDP18S007V3 is given in the two manuals "Operator Manual for
the RCA COSMAC DOS Development System (CDS III) CDP18S007",
MPM-232, and "Hardware Reference Manual for the RCA COSMAC DOS
Development System (CDS III) CDP18S007", MPM-233.
General information on the RCA 1800 Microprocessor Series,
including software, programming techniques, and architecture, is
given in the "User Manual for the CDP1802 COSMAC
Microprocessor", MPM-201.
Binary arithmetic software packages on disk are also available
for use on the COSMAC DOS Development System CDS III. The COSMAC
Microprocessor Fixed-Point Binary Arithmetic Subroutines
(CDP18S826) are described in Product Description PD-6, and the
COSMAC Microprocessor Floating-Point Arithmetic Subroutines
(CDP18S827) are described in Product Description PD-7.
Additional information on these arithmetic diskettes is given in
the manuals "Fixed-Point Binary Arithmetic Subroutines for RCA
COSMAC Microprocessors", MPM-206, and "Floating-Point Arithmetic
Subroutines for RCA COSMAC Microprocessors", MPM-207.
Two application notes are available for PLM. They are ICAN-6928
"Interfacing PLM Code to CDOS System Functions" and "ICAN-6918
"A Methodology for Programming COSMAC 1802 Applications Using
Higher-Level Languages".
Conclusion
----------
"When used with a CDP18S007 COSMAC Development System, PLM
requires 60 kilobytes of read-write memory (...)" 60 KILOBYTES
of RAM!!! I have no idea when the RCA COSMAC was made, but it
was definitely not one of the pioneers of microprocessors. Now,
re-open your old 1976-vintage CP/M 1.4 manuals, and re-read the
minimum size of RAM needed to run it: 16K (24K recommended). No
wonder that the RCA COSMAC disappeared without descendents. Only
a rich administration could afford to make such a "White
Elephant".
As for "Wild" Bill idea that the above advert proves that Gary
Kildall ported Intel's high-level system language PL/M to the
RCA COSMAC, I object the following:
1) Nowhere is Intel's PL/M mentioned (hence, it does not claim
to be compatible with it, the "standard" for microprocessors).
2) The name used is "PLM", not "PL/M".
3) Among other things, it is me who mentioned "PLM/X" on the
comp.os.cpm Newsgroup. This one was advertised in microcomputer
magazines, and is a subset of Mainframe PL/M that used to run
under CP/M. (The only 8-bit version of PL/M was running under
ISIS-II only, a Unix-like OS).
For all those reasons, I simply don't believe that this advert
proves the opinion that "Wild" Bill want us to share.
EOF
Bill
>Conclusion
>3) Among other things, it is me who mentioned "PLM/X" on the
>comp.os.cpm Newsgroup. This one was advertised in microcomputer
>magazines, and is a subset of Mainframe PL/M that used to run
>under CP/M. (The only 8-bit version of PL/M was running under
>ISIS-II only, a Unix-like OS).
You are saying the only 8-bit PL/M compiler(s) (you are aware of)
were NOT running on any 8 bit microprocessors?
Then the RCA PLM 1800 could be unique indeed, could it not?
Not only might you write PL/M programs on an 8 bit Cosmac, you
could also compile them? Or, was PLM/X an 8 bit compiler?
Yes, I see the PLM 1800 compiler takes more memory to run in
than 8 bit CP/M (on an 8080 anyway). That assumes the compiler
is running on CDOS, though, and who can say how that compares
to CP/M? We don't (yet) know how big CP/M 1800 might be.
Kildall writes that the earliest CP/M was written entirely in PL/M,
although he does not mention disk drives nor BIOS in that section.
I believe this included the NCR cash registers, the fortune telling
machines, and Omron data terminals.
Need more input....mnmnmnmn...input.....
And, thanks for the typing. Small job for you, perhaps.....
Bill
Herb Johnson
argument-bating, or acid commentary on a number of unrelated subjects.
I'm only sorry to see my good friends and colleagues Allison and
Emmanuel Roche taken in. If "Bill" has some facts and documents to put
on the table, he can post them when he chooses to.
Any additional comments from me on "argumentive" posting can be found
at my on-line document which covers these sort of situations:
http://www.retrotechnology.com/herbs_stuff/offtopic.html
I'll now respond to "French Luser's" post of documents about
"PLM-1800" for the RCA COSMAC 1802. Roche was kind enough to post a
few pages from an RCA manual, which he mysteriously recieved at some
point. Exerpts of Roche's post follow my comments.
I concurr with "Luser" that it's UNLIKELY that Kildall had anything to
do with this "PLM-1800". Simply put, there was no need for him to be
involved, as any number of people and companies developed variations
of IBM's PL/I language during the 1970's. PL/I in the 1970's was a
very popular language on a very popular series of IBM mainframe and
minicomputer products. Cross-compliers were also very popular during
the period. A Web search of "PL/I" will find sites which list DOZENS
of more-or-less derivative language implemenations that can be traced
by *inspiration* to IBM's PL/I, including of course Kildall's PL/M.
That said, Dr. Kildall wrote on the subject of a PL/I derivative for a
8080 microprocessor and the PACE from National: his work could have
"inspired" RCA's, that's how things work from the academic community.
Bottom line: show documentary EVIDENCE linking Kildall with RCA's "PLM
1800". No one can prove a negative, that Kildall is NOT associated,
one can only guess without positive evidence.
I have some additional information about RCA COSMAC development, which
may be informative, and even on-topic!
I had some experience with an RCA 1802 COSMAC development system, from
RCA. It was an S-100 bus based system, with a Z80 CPU card, full of
mostly proprietary cards and running proprietary software which, based
on my recollections of my work at the time, was NOT CP/M. The floppy
controller was a Micromotion (?) FDC card which COULD and DID run CP/M
on other systems, however. There was a in-circuit emulator card with
"pod" for the 1802. When "Bill" first posted, I thought he had THIS
system, but "Bill's" system has 44-pin cards and (apparently) used an
1802 CPU card. RCA offered a number of 44-pin boards with their COSMAC
chip sets on them: small 44-pin cards were a popular choice of the
era, including the STDbus 44-pin bus and card standard.
The document "Luser" posted SEEMS to reference TWO RCA development
systems. Namely, "COSMAC DOS Development System (CDS III)
CDP18S007V1 or CDP18S007V3", and "COSMAC Development System IV
CDP18S008V1, V3, V5, and V7". I don't know the RCA designation for the
S-100 based system I've described; and I won't sort through
"Bill's" ...postings... to guess which of these refer to the 44-pin
bussed system he has. The "007" references below specifically mention
floppies for the compiler and the OS for the "007" model.
A Google search found only a few discussion group/Usenet posts but
they also suggest the "007" model as floppy-based. But I found a post
which had a Web link to an active site:
http://bellsouthpwp.net/d/o/dogas/cosmac/dev4.htm
..which describes a "008" system, apparently a "System IV", with (44-
pin) Microboards. It has a floppy drive. The manuals listed include
something on BASIC; development systems IV or II. No Pascal or PLM,
but that's merely a lack of evidence. I'd speculate that one of the
Microborards had parallel ports to a SASI single-board floppy
controller; as no Microboard is called out as a floppy controller.
In any event, there are many BETTER FORUMS for 1802 discussions than
comp.os.cpm; especially for products which did not run CP/M! A Web
search will find active Web sites (like the one above), discussion
groups, sites with discussions. In the 1970's and still today, the RCA
1802 is an interesting processor and chip set; old kits and new are
still sold and assembled today, uses and projects still in discussion
and still active.
Herb Johnson
Exerpts from "French Luser" post:
------------------------------------
[RCA document title:?] "PLM-1800 High-Level-Language Compiler"
The PLM-1800 High-Level Language Compiler, CDP18S839, provided
on a diskette, is a software package designed to accelerate
program development on the COSMAC DOS Development System (CDS
III) CDP18S007V1 or CDP18S007V3, and on the COSMAC Development
System IV CDP18S008V1, V3, V5, and V7. It has features similar
to those of many well-known high-level languages, such as PL/1,
ALGOL, and Pascal.
PLM operates directly with the COSMAC Development System
CDP18S008. When used with a CDP18S007 COSMAC Development System,
PLM requires 60 kilobytes of read-write memory and a data
terminal or console. Required software is the CDOS System
Diskette, version 2.2 or later (supplied with the CDP18S007),
and the PLM-1800 High-Level-Language Compiler on diskette
CDP18S839. Documentation is provided with the CDOS Development
System and with the PLM-1800 High-Level-Language Compiler
diskette.
Herbert R. Johnson, New Jersey USA
<a href="http://www.retrotechnology.com/herbs_stuff/"> web site</a>
<a href="http://www.retrotechnology.net/herbs_stuff/"> domain mirror</
a>
my email address: hjohnson AAT retrotechnology DOTT com
if no reply, try in a few days: herbjohnson ATT comcast DOTT net
"Herb's Stuff": old Mac, SGI, 8-inch floppy drives
S-100 IMSAI Altair computers, docs, by "Dr. S-100"
Tarkin
Herb, wasn't/isn't the acronym PL* ubiquitous for
'P_rogramming L_anguage *_'? It seems to me
the PL/M associated with Killdall and CP/M is
always _correctly_ referred to as PL/M. In this
day and age of SMS-inspired lack of attention
to punctuation and detail, I can see how some
people *might*, *maybe*, confuse PL/M and
PLM. But the / would have made mounds of
difference in the era we're discussing, IMHO.
RCA *may* have even been attempting to
capitalize on the success of the name 'PL/M'
in order to sell their 'white elephant' (wild
specualtion on my part).
OT-
Thanks also for pointing out that the poster
was Wild Bill...because I follow the group
from Google, I don't often see the full headers,
and thought this was a different Bill...quite
confusing to say the least...
TTFN,
Tarkin
Bill
wrote:
> If "Bill" has some facts and documents to put
> on the table, he can post them when he chooses to.
Seems to me this started out when I merely ASKED if anyone
had the PLM 1800 or 1802 compiler, and speculated that with it,
a CP/M 1800 might be almost 2/3 done: If the CCP and BDOS
would compile correctly; If PL/M and PLM are reasonably close.
Some things I learn are not up to me to post. I can not post
or give away copies of Kildall's 'Memoirs', for example. But
there's nothing stopping me referring to them where something
Kildall had to say explains what we know from public sources.
As far as RCA products, I'm sure everyone remembers Kildall
was at one time an Intel employee, and continued to develop
software for the Intel processors for many years after.
While he would be unlikely to have had any direct contact
with RCA, he does say he developed some version of
PL-something-or-other for Signetics and their 2650 line
of processors, so who can say for sure?
There is a curious glitch in his statements about virtually
owning the microcomputing operating system business,
accusing gates and IBM of plotting together to steal it,
yet not having his operating system running on most of
the microprocessors being made and sold at the time.
(Atari, Commodore, Apple, even Radio Shack Z-80s)
Or maybe I missed something.
Bill
Tarkin
> On 23 Mar 2007 09:52:52 -0700, "Herb Johnson" <herbrjohn...@gmail.com>
Here's what you missed:
you:
> Maybe somebody's already built this wheel?
allison:
> Far as I know PL/M was never done on 1802.
you:
> Maybe if instead of telling me what you don't know, you simply
> asked to see what I have that even suggests its existence
> we might all get someplace further along.
She was informing you that she did know of a CP/M
variant in her experience for the 1802. She didn't
say there wasn't one. You then blasted her, with some
cryptic reference to some esoteric documentation
which supports your claim for finding this 'unicorn'.
After blasting her, she then definitively states that
your 'unicorn' doesn't exist, and explains why.
Then comes an exchange relating to the prices
of RAM chips. What this has to do with your unicorn,
I'll never know.
At some point after, you sent Luser some images
from some manual you have, this being the esoteric
literature referenced earlier. It seems the basis for the
existence of your unicorn is founded on the fact that
the acronym PLM superficially resembles another acronym,
PL/M, the language Kildall used to develop CP/M. What you
are either unaware of or are blatantly disregarding here is
the proliferation of PL* or PL/* languages in that era.
Root cause? PL was taken to mean 'Programming Language'.
Not every PL* or PL/* variant was used to develop (a) CP/M.
It's sort of like GT or X- in automobiles today. PL* != PL/M .
This was the conclusion reached by Luser, and reinforced
by Herb. Even then, they went to pains to be polite, saying
that it was 'unlikely', or that they were 'unaware'. If you have
a 'smoking gun', then by all means, please show us.
Now, you seem to be acting hurt and shocked. And you also say:
> there's nothing stopping me referring to them where something
> Kildall had to say explains what we know from public sources.
And there's nothing stopping you from posting links to the relevant
passages you are referring to. It's certainly no harder than emailing
images of some manual. With the abundance of material on even the
most esoteric, ancient, and even insignificant of technologica of
yesteryear, I find it difficult to believe that you cannot accept the
fact
that maybe *there just never was a CP/M for the 1802*. Your case for
it's existence rests on the similarity of three letters, and some as-
yet
unrevealed interpretations of someone's memoirs. A quick and
judicious use of Occam's Razor should tell you that it doesn't exist.
Which, provides the answer to your original question: that wheel
hasn't been invented yet.
Thus, now that it has been reasonably ascertained (pending your
smoking gun) that such a creature doesn't, in fact, exist, you may
be motivated to create a CP/M for the 1802. What is most
unfortunate is that you have managed to irritate a good deal of
the people who would have valuable insight into the details of
such an endeavor.
I believe that constitutes the greater protion of what you missed.
HAND,
Tarkin
Tarkin
>
> Thanks.
>
> My arguement is that while we can do it, does it make sense? To copy
> CP/M archetecture to a CPU where page 0 is reserved or where program
> sizes that fit the 8080/z80 model are too big for other CPU models is
> a project for the exercise rather than developing on OS that exploits
> a new CPUs capabilities. Soon as you go from CP/M80 to some CP/M-xx
> that is not 8080 the discussion revolves around file transfer not
> running the basic 8080 code on something that would crash if it tried.
>
> Could CP/M be written for a 1802, sure. I know of someone that
> took the 68000 version (written in C) and ported it to a VAX. So
> yes it could be done..
>
> What does CP/M do: It can:
>
> TYPE a file,
> ERAse it,
> show a disk DIRectory,
> REName a file,
> SAVE a snapshot of ram starting at TPA start to disk file,
> and execute a program that is 8080/z80 based starting at nominal TPA
> start address (usually 100H for 8080/z80).
>
> Anything beyond this is an application program that must load and run
> at TPA start address (nominal cp/m versions 100H). This includes:
>
> DDT, ASM, LOAD, ED, STAT, PIP and about 20,000 programs written for
> CP/M on 8080/z80.
>
> The 8086 version is similar but it will not RUN a system floppy from
> my kaypro 4/84. the library is also far smaller for CP/M-86, and much
> of it is source ported from CP/M-80 that takes advantage of the 8086
> being very similar to 8080. But a 8086 cannot run DDT from a CP/M 80
> system disk (without an emulator or V20 cpu).
>
> One thing I'd learned early on (pre-1979) is that CP/M has meaning
> ONLY on 8080 and derivitive processors. To be explicit, it can be
> ported to anything but, only 8080 and later derived CPUs can actually
> run the library of CP/M program and third party application binaries
> (ignoring emulators/SIMs). So the only reason to "port" CP/M to any
> other non-8080 family CPU is to have access to the filesystem to
> transfer NON-EXECUTABLE sources, data files or text. That is useful
> and tools like RTCPM are out there for that reason. It's far easier
> to do something that runs natively on a given CP/U and OS that allows
> access to the CP/M file system for transfer. This was widely done.
> After all once you have application source code, the API (CP/M BDOS
> calls and BIOS calls) and a filesystem emulation the rest of CP/M is
> logically unimportant and the source code level. This also has been
> done many times as both widely known examples like various SIMs and
> uniquely done one off projects.
>
> There are few OSs that actually span a different CPU archetectures
> and the examples that come to mind are:
>
> CP/M 8080/z80/Z180/NSC800 family, 68000 and Z8000 also 8086(and later)
> were developed by DRI offically. I'd point out the 68k and Z8000
> versions are almost unknown but they exist.
>
> VMS runs on both VAX (32bit) and Alpha(64bit).
>
> Unix or unix flavored OSs run on VAX, Alpha, PDP-11, PDP-7, PCs, MIPS,
> ARM, 68k, powerPC and many others. Unix or it's clones are by far
> the most ported to anything.
>
> I'm sure there are others but they are notable by there obscurity.
>
> Allison
> Thanks.
No need to thank me for doing or saying what is right,
but you're welcome.
> Could CP/M be written for a 1802, sure. I know of someone that
> took the 68000 version (written in C) and ported it to a VAX. So
> yes it could be done..
and
> I'd point out the 68k and Z8000
> versions are almost unknown but they exist.
I agree with, and tells me that it's at least feasible.
> ...is a project for the exercise
Which is what I thought he was driving at, at first.
I've observed people bringing ethernet connectivity
to C=64's, file transfer capability to the TS1000/ZX81,
and build CPU cores from wire-wrapped TTL boards.
Building a project on technical merits alone is a worthwhile
endeavor to most, I think we could all agree.
Which is what disheartened me as I watched Bill's
posts devolve. Part of the reason I read this group
and others is edutainment value; I learn a lot while
satisfying my technical itch. The porting of CP/M to
a different processor like the 1802 might have an
enjoyable thread worth keeping up with, and would
have provided an opportunity to learn more about
systems programming, cross-platform development,
the 1802 architecture, and the RCA Cosmac development
system.
Thankfully, neither this nor any other group is likely
to be 'broken' by a single thread.
TTFN,
Tarkin
Al Kossow
>> I'd point out the 68k and Z8000
>> versions are almost unknown but they exist.
"almost unknown" by whom? They exist on Gaby's site and
have been there for years. Godbout, Stride and others sold the
68K version. It appears in magazine ads of the time. A friend
of mine bought a copy in the early 80's' for a 68K system he
and I designed, which was described in DDJ.
Olivetti was the only company that sold the Z8K version AFAIK, and
paid for its development.
> The porting of CP/M to
> a different processor like the 1802 might have an
> enjoyable thread worth keeping up with
This was NOT what he was saying. He claimed there was
a CP/M and PL/M for the 1802. There is no evidence that
this existed. Having been in contact with people in the
past associated with DR, I have never heard them mention
anything about the 1802 (or 2650, for that matter).
Tarkin
Re-read my post and the thread (from the beginning).
It's my fault I snipped the attribution lines, though.
Al K. :
> "almost unknown" by whom?
By people like me, who are somewhat familiar
with CP/M but not the finer nuances of historical
detail(s). I've heard of both the Z8000 and 68K
versions, thanks to sites like Gaby's, but am much,
much, much more aware of the 8080/Z80 versions,
variants, hardware, etc. I was born in '74, and thus
sorely lack a certain historical perspective; in the
(early) 80's, I drooled over the TS1000 and the Atari ST,
etc. I had never even seen a 'serious' business machine
until '89 or so. My first brush with CP/M occured when
my parents purchased a C= 128 for me.
Al K. :
> This was NOT what he was saying.
It was strongly hinted at in his first post.
TTFN,
Tarkin
s_dub...@yahoo.com
> On 27 Mar 2007 19:53:25 -0700, "Tarkin" <Tarkin...@gmail.com> wrote:
>
> versions are almost unknown but they exist.
>
>
> Unix or unix flavored OSs run on VAX, Alpha, PDP-11, PDP-7, PCs, MIPS,
> ARM, 68k, powerPC and many others. Unix or it's clones are by far
> the most ported to anything.
>
> I'm sure there are others but they are notable by there obscurity.
>
>
> - Show quoted text -
Very well put.
-ot-
I don't know why you have 'request not to be archived' in your message
header, Allison. I value your posts as I imagine those who come
later, by searching, will also. I've cheated usenet by quoting it in
its entirety here.
-e ot-
I'm struck, in reading the CP/M 1.4 docs, that some emphasis is made
to the point that the CCP is considered a separate 'program', implying
a possible separation from the FDOS and for later versions there are
replacement CCP's, as you well know. Yet the built-in functions you
list; TYPE..SAVE are the elements of the CCP, without which CP/M loses
its 'look and feel'. I'm restating the obvious I guess. I would also
guess that you would feel those other cp/m's with replacement CCPs are
no longer CP/M.
Steve
s_dub...@yahoo.com
>
> > -ot-
> >I don't know why you have 'request not to be archived' in your message
> >header, Allison. I value your posts as I imagine those who come
> >later, by searching, will also. I've cheated usenet by quoting it in
> >its entirety here.
> > -e ot-
>
> turned on does help with spammers that use address harvesting it to
> keep their lists up to date.
>
I use yahoo mail, its bulk folder catches 99% of the spam, and it is
easy to retrieve mail no matter where I am.
>
>
> >I'm struck, in reading the CP/M 1.4 docs, that some emphasis is made
> >to the point that the CCP is considered a separate 'program', implying
> >a possible separation from the FDOS and for later versions there are
> >replacement CCP's, as you well know. Yet the built-in functions you
> >list; TYPE..SAVE are the elements of the CCP, without which CP/M loses
> >its 'look and feel'. I'm restating the obvious I guess. I would also
> >guess that you would feel those other cp/m's with replacement CCPs are
> >no longer CP/M.
>
> The CCP is a seperate module and can be replaced. It's difference
> from all other CP/M programs is it loads in the 2k space below the
> BDOS rather than the start of the TPA (100h). Nominally it's
> overwrtten by larger programs and reloaded by the warmboot
> code in the BIOS. It is the user interface, the shell in more modern
> terms.
>
That's an important detail for someone new to programming for CP/M-80
to keep in mind. SID and DDT relocate to below the BDOS to get out of
the way.
[regarding replacement CCPs]
> Not at all. I use ZCPR and it retains most if not all of the look and
> feel. There are otehr shells that give cpm a more unix like look and
> feel. I was illustrating that outside of the CCP built in functions
> everything else depends on transient programs most of which
> would also have to be ported to a new cpu.
>
Right, I get your point.
> For example we create a new version of CP/M for the XYZ24 24bit cpu.
> What would the ASM assembler for that new system put out XYZ20 or
> 8080 code? It's details like this rather than the unser looks and
> feel that are meaningful.
>
It helped that most of the transient utilities are written in PL/M,
and regarding the port to the 8086, there was a PL/M-86 for it. ED,
STAT,...and it helped that CP/M-86 kept v2.2 file structure
compatibility.
> FYI:OS/8, RT-11 and VMS at the command line level
> are remarkably similar in look and feel to CP/M. The DCL
> [digital command line] user interface dates back to OS/8 (pdp-8) which
> predates CP/M. OS/8 had other similarities (loadable drivers), user
> interface that was overlaid to save core (ram) and a fairly modular
> design for its day. it could be said the TOPS-10 (also similar
> interface) system used by Killdall for cross development may have
> influenced CP/M user interface development. In the 80s I took
> advanatge of that and had my CP/M, PDP11 and VAX systems all
> running a suite of native programs that made them look alike and
> similar to the CP/M system just so I'd not have to remember three
> very different machines to do routine editing and other tasks.
>
> Allison-
Interesting.
Steve
CBFalconer
>
... snip ...
>
> I'm struck, in reading the CP/M 1.4 docs, that some emphasis is
> made to the point that the CCP is considered a separate 'program',
> implying a possible separation from the FDOS and for later
> versions there are replacement CCP's, as you well know. Yet the
> built-in functions you list; TYPE..SAVE are the elements of the
> CCP, without which CP/M loses its 'look and feel'. I'm restating
> the obvious I guess. I would also guess that you would feel those
> other cp/m's with replacement CCPs are no longer CP/M.
Because it is. Alternatives include DDT, DDTZ, CCPLUS, ZCPR. The
DDT* have a much different feel than CCP. One unique point about
CCP is that it is included in the cold boot load (as can CCPLUS or
ZCPR). All of those run in the high memory area, leaving the TPA
available.
--
Chuck F (cbfalconer at maineline dot net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net>
s_dub...@yahoo.com
> s_dubrov...@yahoo.com wrote:
>
> ... snip ...
>
> > I'm struck, in reading the CP/M 1.4 docs, that some emphasis is
> > made to the point that the CCP is considered a separate 'program',
> > implying a possible separation from the FDOS and for later
> > versions there are replacement CCP's, as you well know. Yet the
> > built-in functions you list; TYPE..SAVE are the elements of the
> > CCP, without which CP/M loses its 'look and feel'. I'm restating
> > the obvious I guess. I would also guess that you would feel those
> > other cp/m's with replacement CCPs are no longer CP/M.
>
> Because it is. Alternatives include DDT, DDTZ, CCPLUS, ZCPR. The
> DDT* have a much different feel than CCP. One unique point about
> CCP is that it is included in the cold boot load (as can CCPLUS or
> ZCPR). All of those run in the high memory area, leaving the TPA
> available.
>
much promoted. And by the time of cp/m-86, with patches in place,
this is even less so.
<ot> on on off thread question to you:
Was there ever a small-c lint type program around for the micros?
I've been messing around with Cain's small-c this weekend and it is
laughable how bad I can misbehave with the syntax.
thxs,
Steve
s_dub...@yahoo.com
[snip]
>
> Herb, wasn't/isn't the acronym PL* ubiquitous for
> 'P_rogramming L_anguage *_'? It seems to me
> the PL/M associated with Killdall and CP/M is
> always _correctly_ referred to as PL/M. In this
> day and age of SMS-inspired lack of attention
> to punctuation and detail, I can see how some
> people *might*, *maybe*, confuse PL/M and
> PLM. But the / would have made mounds of
> difference in the era we're discussing, IMHO.
> RCA *may* have even been attempting to
> capitalize on the success of the name 'PL/M'
> in order to sell their 'white elephant' (wild
> specualtion on my part).
>
> TTFN,
> Tarkin-
As a follow up to 'PL' there's this languages list:
PL/C - Slight subset of PL/I, aimed at student use. "User's Guide to
PL/C", S. Worona et al, Cornell, June 1974. "PL/C - A High
Performance
Compiler" H.L. Morgan et al, Proc SJCC, AFIPS 38:503-510 (1971).
Implementated on IBM 370.
PL/I - Programming Language I. George Radin, 1964. Originally named
NPL.
An attempt to combine the best features of FORTRAN, COBOL and ALGOL
60.
Result is large but elegant. One of the first languages to have a
formal
semantic definition, using the Vienna Definition Language. EPL, a
dialect
of PL/I, was used to write almost all of the Multics OS. PL/I has no
reserved words. Types are fixed, float, complex, character strings
with
max length, bit strings, and label variables, no user-defined types.
Dynamic arrays. Summation, multi-level structures, structure
assignment.
Untyped pointers, side effects, aliasing. Control flow goto, do-end
groups, do-to-by-while-end loops, external procedures, nested
procedures
and blocks. Procedures may be declared recursive, or grouped into
generic
families. Controlled variables. Many implementations support
concurrency
('call task' and 'wait(event)' amount to fork/join) and compile-time
statements. Exception handling. "A Structural View of PL/I", D.
Beech,
Computing Surveys, 2,1 33-64 (1970). ANS X3.53-1976. PL/I is still
widely
used, at IBM and elsewhere. The CFS file system used at Argonne to
manage
terabytes of data is written in PL/I.
list: P...@UIUCVMD.BITNET
ftp://wuarchive.wustl.edu/mirrors/msdos/pli/runpli1a.arc, PL/I
interpreter
version: LPI for PC's and Unix, Liant <r...@lpi.liant.com>
(508)626-0006
PL/I SUBSET - Early 70's version of PL/I for minis.
PL/I Subset G - ("General Purpose") The commercial PL/I subset (i.e.,
what
was actually implemented by most vendors). ANS X3.74-1981.
PL/I-FORMAC - Variant of FORMAC. "The PL/I-FORMAC Interpreter", J.
Xenakis, Proc 2nd Symp Symbolic and Algebraic Manip, ACM (Mar 1971).
Sammet 1969, p.486.
Plisp - Pattern LISP. 1990. A pattern-matching rewrite-rule
language,
optimized for describing syntax translation rules. (See LISP70).
PLITS - Programming Language In The Sky. A computational model for
concurrency with communication via asynchronous message-passing.
"High
Level Programming for Distributed Computing", J.A. Feldman, CACM 22(6):
353-
368 (Jun 1979).
PL/M - Programming Language/Microcomputers. Gary Kildall, MAA (later
Digital Research) for Intel, 1972. A very low level language
incorporating
ideas from PL/I, ALGOL and XPL. Integrated macro processor. CP/M
was
written to support development of the PL/M compiler (not the other
way
around!). "PL/M-80 Programming Manual", Doc 98-268B, Intel 1976. "PL/
M
Programmer's Guide", Doc 452161-003, Intel. "A Guide to PL/M
Programming
for Microcomputer Applications", D. McCracken, A-W 1978. Versions: PL/
M-
80, PL/M-86, PL/M-286, PL/M-386.
ftp://iecc.com/pub/file/plm.shar.gz parser for PL/M-386
//locke.ccil.org/pub/retro/plm2c.tar.gz PL/M-286 to C
translator
PL/MP - C.J. Tan, IBM TJWRC, 1978. A microprogramming language
resembling
a subset of PL/I. "Code Optimization for Microcode Compilers", C.J.
Tan,
AFIPS Conf Proc 47:649-655 (1978).
PL/P - Programming Language, Prime. Russ Barbour, PRIME Computer,
late
70's. Subset of PL/I used internally for implementation of PRIMOS.
(See
SPL[4]).
PL/PROPHET - PL/I-like language for the PROPHET system, used by
pharmacologists. "The Implementation of the PROPHET System", P.A.
Castleman et al, NCC 43, AFIPS (1974).
PL/S - Programming Language/Systems. IBM late 60's, for the IBM 360
and
370. A machine-oriented language derived from PL/I, permiting inline
assembly code and control over register usage. Much of IBM 360
OS/MFT/MVT/SVS/MVS was written in it. Used internally, never released
to
the public. Documented by various IBM internal ZZ-? publications.
"PL/S,
Programming Language/Systems", W.R. Brittenham, Proc GUIDE Intl, GUIDE
34,
May 14, 1972, pp.540-556. Versions: PLS1, PLSII.
PL/Seq - Programming Language for Sequences. A DSP language. "A
General
High Level Language for Signal Processors", J. Skytta & O. Hyvarinen,
Digital Signal Processing 84, Proc Intl Conf, Fiorence, Italy, Sep
1984,
pp.217-221.
PLZ - Zilog. A high level language for programming microprocessors.
A
minimal block structured language, goto-less and only DO-OD loops
with
exit, repeat [from] or continue [from] placed anywhere in the loop.
Record
structures. Used by Zilog, and by Olivetti in their S6000 series.
"Introduction to Microprocessor Programming Using PLZ", Richard Conway
et
al, Winthrop Pub 1979. Available as PLZ/SYS. "Report on the
Programming
Language PLZ/SYS", Tod Snook et al, Springer 1978.
PLZ/ASM - Similar to PLZ, but with assembler instructions instead of
statements.
PLUM - U. Maryland. Compiler for a substantial subset of PL/I for
the
Univac 1100. "PL/I Programming with PLUM", M.V. Zelkowitz, Paladin
House,
1978.
glen herrmannsfeldt
(snip)
> of PL/I, was used to write almost all of the Multics OS. PL/I has no
> reserved words. Types are fixed, float, complex, character strings
Attributes are either fixed or float, either real or complex, either
binary or decimal. That is, fixed decimal complex is a legal type.
DCL X FIXED DECIMAL(10,3) COMPLEX;
X=0;
DO IMAG(X)=0.1 TO 999 BY 0.13;
PUT SKIP LIST (SQRT(X),X**X);
END;
-- glen