On Friday, September 2, 2022 at 12:05:09 PM UTC-6, Louis Krupp wrote:

> Burroughs Extended ALGOL borrowed formatting from FORTRAN, changing
> FORMAT statements to FORMAT declarations. It seemed natural and obvious
> at the time; it would be surprising -- and unfortunate -- if no other
> ALGOL implementations did something similar.

I think that Algol W also borrowed formatting from FORTRAN, although perhaps
not in the same way that Burroughs Extended ALGOL did that.

_Some_ implementations of ALGOL, including IBM's implementation for the
System/360, annoyed potential users by not using reserved words (as Algol W
did) but instead requiring that all keywords in the language be enclosed in
single quotes. I think that may be underrated as a cause for the failure of the
language to be widely adopted.

It _also_ didn't help that the publication language, at least, had a character set
which included a lot of characters most machines didn't have. Some Russian
machines, the IBM STRETCH, and some British machines were among the very
few exceptions.

John Savard

Quadibloc <jsavard@ecn.ab.ca> wrote:
> On Friday, September 2, 2022 at 12:05:09 PM UTC-6, Louis Krupp wrote:
>
>> Burroughs Extended ALGOL borrowed formatting from FORTRAN, changing
>> FORMAT statements to FORMAT declarations. It seemed natural and obvious
>> at the time; it would be surprising -- and unfortunate -- if no other
>> ALGOL implementations did something similar.
>
> I think that Algol W also borrowed formatting from FORTRAN, although perhaps
> not in the same way that Burroughs Extended ALGOL did that.
>
> _Some_ implementations of ALGOL, including IBM's implementation for the
> System/360, annoyed potential users by not using reserved words (as Algol W
> did) but instead requiring that all keywords in the language be enclosed in
> single quotes. I think that may be underrated as a cause for the failure of the
> language to be widely adopted.

I think it would be hard to underrate that as a factor, at least as far as
IBM’s ALGOL is concerned. ;-)

>
> It _also_ didn't help that the publication language, at least, had a character set
> which included a lot of characters most machines didn't have. Some Russian
> machines, the IBM STRETCH, and some British machines were among the very
> few exceptions.
>
> John Savard
>

--
Pete

On Monday, September 5, 2022 at 9:50:00 PM UTC+10, Quadibloc wrote:
> On Friday, September 2, 2022 at 12:05:09 PM UTC-6, Louis Krupp wrote:
>
> > Burroughs Extended ALGOL borrowed formatting from FORTRAN, changing
> > FORMAT statements to FORMAT declarations. It seemed natural and obvious
> > at the time; it would be surprising -- and unfortunate -- if no other
> > ALGOL implementations did something similar.
> I think that Algol W also borrowed formatting from FORTRAN, although perhaps
> not in the same way that Burroughs Extended ALGOL did that.
>
> _Some_ implementations of ALGOL, including IBM's implementation for the
> System/360, annoyed potential users by not using reserved words (as Algol W
> did) but instead requiring that all keywords in the language be enclosed in
> single quotes. I think that may be underrated as a cause for the failure of the
> language to be widely adopted.
..
I think that most implementations used apostrophes to delimit keywords.
..
> It _also_ didn't help that the publication language, at least, had a character set
> which included a lot of characters most machines didn't have.
..
The publication language was just that: a publication language.
It used upper and lower case, with boldface keywords, and special characters.
In 1960, most line printers offered upper case only, digits, and a limited
range of special characters.
That didn't change much for a couple of decades.
Printing terminals did offer upper and lower case, including
the ASR-38, Memorex, GE Terminet, Diablo, and the Friden flexowriter,
the latter going a lot closer to the publication language.
..
> Some Russian
> machines, the IBM STRETCH, and some British machines were among the very
> few exceptions.

On 9/1/22 19:58, Quadibloc wrote:
> On Thursday, September 1, 2022 at 4:13:14 PM UTC-6, Peter Flass wrote:
>> ALGOL was later sunk by the
>> development of ALGOL-68, which (AIUI) was more different from ALGOL-60 than
>> C++ is from C.
>
> I have to agree that Algol-68 sank Algol. I had thought that the reason for that,
> aside from silly things like case ... esac, was that implementing Algol-68 was
> somewhat beyond the state of the art at the time.
>
> That doesn't mean that there were _no_ implementations, just that many who
> might have implemented an update to Algol did not feel themselves qualified
> to produce an adequate implementation.
>
> Of course, the world could have just ignored Algol-68, and stuck with Algol-60,
> perhaps with slight tweaks and improvements. What closed off _that_ line of
> development was...
>
> Pascal.
>
> John Savard

Pascal is still widely used. Tiobe Index is _not_ a very accurate
indicator of its market share. If I had to guesstimate I would say that
Pascal, Delphi, et al. are usually 2-3 slots higher on the index in reality.

The highly-paid engineers who use Pascal, Ada, and other Wirthian
languages do not frequent Internet forums promoting their design
choices. They are too busy making money to care what nerd wanks and
stack monkeys think.

I have been building Pascal libraries that I expect will still be usable
in 15-20 years. Can we say that of the 'popular' nerd languages? How
long will it be before Python breaks again? I have Pascal code bases
that are 15+ years old that still compile and run bug-free on modern
hardware.

The only thing missing from modern Pascal compilers is facilities for
directly accessing all hardware interfaces, like in C. If that facility
were baked in a lot of C programmers would switch. I would love to be
able to gin up bootloaders and microkernels with a standard Pascal-like
syntax.

Wirth's vision of Algol wins. Anyway that's my nickle's wirth.

--

Bravosi

On 2022-09-06, Bravosi <syrio@for.ell> wrote:

> I have been building Pascal libraries that I expect will still be usable
> in 15-20 years. Can we say that of the 'popular' nerd languages? How
> long will it be before Python breaks again? I have Pascal code bases
> that are 15+ years old that still compile and run bug-free on modern
> hardware.

I'm still maintaining the C code base that I wrote 30 years ago.
It compiles and runs on both Windows and Linux. I haven't seen
an SCO system for some time but it should still run there too.

> The only thing missing from modern Pascal compilers is facilities for
> directly accessing all hardware interfaces, like in C. If that facility
> were baked in a lot of C programmers would switch. I would love to be
> able to gin up bootloaders and microkernels with a standard Pascal-like
> syntax.

Sounds like a worthwhile enhancement (if it's done right).
For me it's academic - I've never liked Wirthian languages -
but it sounds long overdue.

> Wirth's vision of Algol wins. Anyway that's my nickle's wirth.

Ah, you call him by value. :-)

--
/~\ Charlie Gibbs | Microsoft is a dictatorship.
\ / <cgibbs@kltpzyxm.invalid> | Apple is a cult.
X I'm really at ac.dekanfrus | Linux is anarchy.
/ \ if you read it the right way. | Pick your poison.

Robin Vowels <robin.vowels@gmail.com> wrote:
> On Monday, September 5, 2022 at 9:50:00 PM UTC+10, Quadibloc wrote:
>> On Friday, September 2, 2022 at 12:05:09 PM UTC-6, Louis Krupp wrote:
>>
>>> Burroughs Extended ALGOL borrowed formatting from FORTRAN, changing
>>> FORMAT statements to FORMAT declarations. It seemed natural and obvious
>>> at the time; it would be surprising -- and unfortunate -- if no other
>>> ALGOL implementations did something similar.
>> I think that Algol W also borrowed formatting from FORTRAN, although perhaps
>> not in the same way that Burroughs Extended ALGOL did that.
>>
>> _Some_ implementations of ALGOL, including IBM's implementation for the
>> System/360, annoyed potential users by not using reserved words (as Algol W
>> did) but instead requiring that all keywords in the language be enclosed in
>> single quotes. I think that may be underrated as a cause for the failure of the
>> language to be widely adopted.
> .
> I think that most implementations used apostrophes to delimit keywords.

No other implementation I ever saw, admittedly only a few: mostly Burroughs
and Algol W - just looked at HP - and maybe a couple of others. That alone
would make me never want to get near the language, it seems like a recipe
for disaster.

--
Pete

Bravosi <syrio@for.ell> wrote:
> On 9/1/22 19:58, Quadibloc wrote:
>> On Thursday, September 1, 2022 at 4:13:14 PM UTC-6, Peter Flass wrote:
>>> ALGOL was later sunk by the
>>> development of ALGOL-68, which (AIUI) was more different from ALGOL-60 than
>>> C++ is from C.
>>
>> I have to agree that Algol-68 sank Algol. I had thought that the reason for that,
>> aside from silly things like case ... esac, was that implementing Algol-68 was
>> somewhat beyond the state of the art at the time.
>>
>> That doesn't mean that there were _no_ implementations, just that many who
>> might have implemented an update to Algol did not feel themselves qualified
>> to produce an adequate implementation.
>>
>> Of course, the world could have just ignored Algol-68, and stuck with Algol-60,
>> perhaps with slight tweaks and improvements. What closed off _that_ line of
>> development was...
>>
>> Pascal.
>>
>> John Savard
>
> Pascal is still widely used. Tiobe Index is _not_ a very accurate
> indicator of its market share. If I had to guesstimate I would say that
> Pascal, Delphi, et al. are usually 2-3 slots higher on the index in reality.
>
> The highly-paid engineers who use Pascal, Ada, and other Wirthian
> languages do not frequent Internet forums promoting their design
> choices. They are too busy making money to care what nerd wanks and
> stack monkeys think.
>
> I have been building Pascal libraries that I expect will still be usable
> in 15-20 years. Can we say that of the 'popular' nerd languages? How
> long will it be before Python breaks again? I have Pascal code bases
> that are 15+ years old that still compile and run bug-free on modern
> hardware.

Same with PL/I. Programs from 1964 still compile, with minimal changes, and
run.

>
> The only thing missing from modern Pascal compilers is facilities for
> directly accessing all hardware interfaces, like in C. If that facility
> were baked in a lot of C programmers would switch. I would love to be
> able to gin up bootloaders and microkernels with a standard Pascal-like
> syntax.
>

Multics PL/I was a good example of this.

--
Pete

Charlie Gibbs <cgibbs@kltpzyxm.invalid> wrote:
> On 2022-09-06, Bravosi <syrio@for.ell> wrote:
>
>> I have been building Pascal libraries that I expect will still be usable
>> in 15-20 years. Can we say that of the 'popular' nerd languages? How
>> long will it be before Python breaks again? I have Pascal code bases
>> that are 15+ years old that still compile and run bug-free on modern
>> hardware.
>
> I'm still maintaining the C code base that I wrote 30 years ago.
> It compiles and runs on both Windows and Linux. I haven't seen
> an SCO system for some time but it should still run there too.
>
>> The only thing missing from modern Pascal compilers is facilities for
>> directly accessing all hardware interfaces, like in C. If that facility
>> were baked in a lot of C programmers would switch. I would love to be
>> able to gin up bootloaders and microkernels with a standard Pascal-like
>> syntax.
>
> Sounds like a worthwhile enhancement (if it's done right).
> For me it's academic - I've never liked Wirthian languages -
> but it sounds long overdue.

PL\360 is an early example, Algol-like syntax and complete access to
hardware. Wirth used it to write Algol W. The first time I saw the source I
thought it WAS ALGOL.

>
>> Wirth's vision of Algol wins. Anyway that's my nickle's wirth.
>
> Ah, you call him by value. :-)
>

--
Pete

On Mon, 5 Sep 2022 18:44:37 -0700 (PDT)
Robin Vowels <robin.vowels@gmail.com> wrote:

> I think that most implementations used apostrophes to delimit keywords.

With Algol 68 I came across stropping by quotes, case or underline,
that was where I started to think there was a design problem.

--
Steve O'Hara-Smith
Odds and Ends at http://www.sohara.org/

On 2022-09-06, Peter Flass <peter_flass@yahoo.com> wrote:

> Robin Vowels <robin.vowels@gmail.com> wrote:
>
>> On Monday, September 5, 2022 at 9:50:00 PM UTC+10, Quadibloc wrote:
>>
>>> On Friday, September 2, 2022 at 12:05:09 PM UTC-6, Louis Krupp wrote:
>>>
>>>> Burroughs Extended ALGOL borrowed formatting from FORTRAN, changing
>>>> FORMAT statements to FORMAT declarations. It seemed natural and obvious
>>>> at the time; it would be surprising -- and unfortunate -- if no other
>>>> ALGOL implementations did something similar.
>>>
>>> I think that Algol W also borrowed formatting from FORTRAN, although
>>> perhaps not in the same way that Burroughs Extended ALGOL did that.
>>>
>>> _Some_ implementations of ALGOL, including IBM's implementation for the
>>> System/360, annoyed potential users by not using reserved words (as Algol W
>>> did) but instead requiring that all keywords in the language be enclosed in
>>> single quotes. I think that may be underrated as a cause for the failure of
>>> the language to be widely adopted.
>>
>> I think that most implementations used apostrophes to delimit keywords.
>
> No other implementation I ever saw, admittedly only a few: mostly Burroughs
> and Algol W - just looked at HP - and maybe a couple of others. That alone
> would make me never want to get near the language, it seems like a recipe
> for disaster.

I agree. At university we were exposed to Algol 60, Algol 68, Algol W,
and pl360 (which I liked to refer to as the misbegotten bastard child
of Algol and assembly language). Putting apostrophes around every keyword
seemed like a particularly clumsy thing to do. I never did get an Algol
program to work; I gratefully retreated to assembly language (plus a
real-world mix of RPG and COBOL), where I happily stayed until C came
along. I've never liked Wirthian "bondage and discipline" languages.

--
/~\ Charlie Gibbs | Microsoft is a dictatorship.
\ / <cgibbs@kltpzyxm.invalid> | Apple is a cult.
X I'm really at ac.dekanfrus | Linux is anarchy.
/ \ if you read it the right way. | Pick your poison.

On Tuesday, September 6, 2022 at 6:21:53 AM UTC-6, Bravosi wrote:
> On 9/1/22 19:58, Quadibloc wrote:

> > Of course, the world could have just ignored Algol-68, and stuck with Algol-60,
> > perhaps with slight tweaks and improvements. What closed off _that_ line of
> > development was...
> >
> > Pascal.

> Pascal is still widely used.

What I had written was not intended to imply otherwise.

Precisely because Pascal became very popular, and offered everything
that Algol-60 did, with important additions, the alternative of sticking with
Algol-60 instead, if one didn't like Algol-68, wasn't a common choice.

John Savard

On Tuesday, September 6, 2022 at 12:31:42 PM UTC-6, Peter Flass wrote:
> Robin Vowels <robin....@gmail.com> wrote:
> > On Monday, September 5, 2022 at 9:50:00 PM UTC+10, Quadibloc wrote:

> >> _Some_ implementations of ALGOL, including IBM's implementation for the
> >> System/360, annoyed potential users by not using reserved words (as Algol W
> >> did) but instead requiring that all keywords in the language be enclosed in
> >> single quotes. I think that may be underrated as a cause for the failure of the
> >> language to be widely adopted.

> > I think that most implementations used apostrophes to delimit keywords.

> No other implementation I ever saw, admittedly only a few: mostly Burroughs
> and Algol W - just looked at HP - and maybe a couple of others. That alone
> would make me never want to get near the language, it seems like a recipe
> for disaster.

I'm not too familiar with the other examples, but I think some early ones for British
computers were like that.

John Savard

On Wednesday, September 7, 2022 at 5:30:03 AM UTC+10, Ahem A Rivet's Shot wrote:
> On Mon, 5 Sep 2022 18:44:37 -0700 (PDT)
> Robin Vowels <robin....@gmail.com> wrote:
>
> > I think that most implementations used apostrophes to delimit keywords.
..
I was referring to punch card versions.
We had an Elliott paper tape version and an ICL card version.
..
> With Algol 68 I came across stropping by quotes, case or underline,
> that was where I started to think there was a design problem.

On Wednesday, September 7, 2022 at 12:54:16 PM UTC+10, Quadibloc wrote:
> On Tuesday, September 6, 2022 at 6:21:53 AM UTC-6, Bravosi wrote:
> > On 9/1/22 19:58, Quadibloc wrote:
>
> > > Of course, the world could have just ignored Algol-68, and stuck with Algol-60,
> > > perhaps with slight tweaks and improvements. What closed off _that_ line of
> > > development was...
> > >
> > > Pascal.
> > Pascal is still widely used.
> What I had written was not intended to imply otherwise.
>
> Precisely because Pascal became very popular, and offered everything
> that Algol-60 did,
..
It didn't really. It didn't have exponentiation operator,
and some of the I/O operations never worked properly
on cards.
..
> with important additions, the alternative of sticking with
> Algol-60 instead, if one didn't like Algol-68, wasn't a common choice.

On Sunday, September 4, 2022 at 10:42:05 AM UTC+10, Rich Alderson wrote:
> Robin Vowels <robin....@gmail.com> writes:
>
>
> > PL/C was a pretty good implementation, and fast .
> > Other non-IBM PL/I compilers included those for
> > CDC, Univac, Burroughs, and on micros DR PL/I, Q1/Lite PL/I.
> I'm surprised that no one has mentioned the GE/Honeywell PL/I compiler which
> was used as the implementation language for much of Multics...
..
OK, other implementations of PL/I language included
IBM Series/1, Burroughs B 6700 / B 7700, CDC Cyber 70, 170, 6000,
Data General Eclipse, DEC System 10 and 20, Honeywell 66/6000,
Q1 Corporation Q1/LMC and Q1/Lite, PLUM, SP/k compilers,
PLAGO, Student PL/I, IBM 1130, CIMS PL/I, Olivetti A7, DR PL/I,
Wang PL/I.
(In addition to the above GE/Honeywell PL/I, of course).
..
IBM produced the PL/I-F compiler for OS/360, and the PL/I-D compiler for smaller
machines using DOS, and the PL/I Checkout and Optimising compilers,
plus a number of others since.

According to Quadibloc <jsavard@ecn.ab.ca>:
>_Some_ implementations of ALGOL, including IBM's implementation for the
>System/360, annoyed potential users by not using reserved words (as Algol W
>did) but instead requiring that all keywords in the language be enclosed in
>single quotes. I think that may be underrated as a cause for the failure of the
>language to be widely adopted.

IBM Algol F seemed to me to be a product that existed purely so they could
check the box yes, we have Algol.

As distributed by IBM, every time your program entered a block it would do
a GETMAIN system call and a FREEMAIN when it left the block. Needless to
say that was a disaster for performance. Someone at Princeton patched it
to allocate and free in larger more sensible chunks.

The quoted keywords were the same sort of thing, technically correct and
horrible in practice.

--
Regards,
John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly

On Tue, 06 Sep 2022 19:58:34 -0700, Quadibloc wrote:

> On Tuesday, September 6, 2022 at 12:31:42 PM UTC-6, Peter Flass wrote:
>> Robin Vowels <robin....@gmail.com> wrote:
>> > On Monday, September 5, 2022 at 9:50:00 PM UTC+10, Quadibloc wrote:
>
>> >> _Some_ implementations of ALGOL, including IBM's implementation for
>> >> the System/360, annoyed potential users by not using reserved words
>> >> (as Algol W did) but instead requiring that all keywords in the
>> >> language be enclosed in single quotes. I think that may be
>> >> underrated as a cause for the failure of the language to be widely
>> >> adopted.
>
>> > I think that most implementations used apostrophes to delimit
>> > keywords.
>
>> No other implementation I ever saw, admittedly only a few: mostly
>> Burroughs and Algol W - just looked at HP - and maybe a couple of
>> others. That alone would make me never want to get near the language,
>> it seems like a recipe for disaster.
>
> I'm not too familiar with the other examples, but I think some early
> ones for British computers were like that.

The IMP language [1] (nothing to do with ARPAnet) developed at the
University of Edinburgh had a different stropping convention. This worked
well with the line reconstruction algorithm (inherited, sort of, from the
Compiler Compiler).

A reserved word started with a '%' character and ended with whitespace.
For example:

%begin ... %end

You could split words as long as the '%' character was introdiced, so:

%const %integer and %constinteger

were equivalent.

I know this because I have just written an IMP compiler.

[1] https://en.wikipedia.org/wiki/Edinburgh_IMP

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

On 07/09/2022 11:56, Bob Eager wrote:
> You could split words as long as the '%' character was introdiced, so:
> %const %integer and %constinteger
> were equivalent.

The Walgol* compiler went further and tried to be helpful by
completing words if it could. The trouble was that many of our students
couldn't spell, and "proceedure" turned into "procedure e do until real"
which was far more confusing than simply giving an error message in the
first place. It was a great relief when Algol 68 arrived, and the word
was spelled "proc", which even the wonkiest spellers could manage.
____
* I think it was Walgol, which we used for batch processing of student
jobs. But it could have been Egdon, which we also used at some stage.

--
Andy Walker, Nottingham.
Andy's music pages: www.cuboid.me.uk/andy/Music
Composer of the day: www.cuboid.me.uk/andy/Music/Composers/Couperin

On 7 Sep 2022, Andy Walker wrote
(in article <tfa0ou$gq$1@gioia.aioe.org>):

> The Walgol* compiler went further and tried to be helpful by
> completing words if it could. The trouble was that many of our students
> couldn't spell, and "proceedure" turned into "procedure e do until real"
> which was far more confusing than simply giving an error message in the
> first place. It was a great relief when Algol 68 arrived, and the word
> was spelled "proc", which even the wonkiest spellers could manage.
> ____
> * I think it was Walgol, which we used for batch processing of student
> jobs. But it could have been Egdon, which we also used at some stage.

Walgol? I think not, see:

> /Users/wf/KDF9/emulation/Testing: cat Whetstone/WALKER.a60
> WALKER|
> _b_e_g_i_n
> _p_r_o_c_e_e_d_u_r_e eh; ;
> _e_n_d
> |

> /Users/wf/KDF9/emulation/Testing: walgol WALKER
> This is ee9 9.0p, compiled by GNAT 12.1.0 on 2022-09-06.
> Running the KDF9 problem program Adjuncts/KMW0201--UPU in fast mode.
> ____________________________________________________________________
> [q]
> OUT;N.|
> ee9: OUT 5: requested a #02 type and got TR1 in Latin-1 mode.
> [m] WALKER|
> [m] FAIL
> [m] RAN/EL/000M01S/000M03S SIZE 2
> ____________________________________________________________________
> TP0: ===
> FAIL 126 LINE 2 DELIMITER ;
> RAN/EL/000M01S/000M03S SIZE 2
> ===
> /Users/wf/KDF9/emulation/Testing:

FAIL 126 is "illegal underlined word".

--
Bill Findlay

Robin Vowels <robin.vowels@gmail.com> writes:

> OK, other implementations of PL/I language included
[ snip ]
> DEC System 10 and 20
[ snip ]

Details?

I went from being a PL/I programmer on OS/360 to SVS to MVS, and latterly moved
to TOPS-20 on a DEC-20. The rest of my career (~40 years) was spent on the
latter architecture, and I've NEVER HEARD OF PL/I FOR THE PDP-10.

Had I know it existed, I'd have badgered management anywhere I worked for it.
Relentlessly.

--
Rich Alderson news@alderson.users.panix.com
Audendum est, et veritas investiganda; quam etiamsi non assequamur,
omnino tamen proprius, quam nunc sumus, ad eam perveniemus.
--Galen

On Thursday, September 8, 2022 at 5:49:08 AM UTC+10, Rich Alderson wrote:
> Robin Vowels <robin....@gmail.com> writes:
>
> > OK, other implementations of PL/I language included
> [ snip ]
> > DEC System 10 and 20
> [ snip ]
>
> Details?
>
> I went from being a PL/I programmer on OS/360 to SVS to MVS, and latterly moved
> to TOPS-20 on a DEC-20. The rest of my career (~40 years) was spent on the
> latter architecture, and I've NEVER HEARD OF PL/I FOR THE PDP-10.
>
> Had I know it existed, I'd have badgered management anywhere I worked for it.
> Relentlessly.
..
You needed to ask for CPL.
The publication is decsystem10 Conversational Programming Language
User's Manual, DEC-10-LCPUA-A-D, 1975.

Rich Alderson <news@alderson.users.panix.com> wrote:
> Robin Vowels <robin.vowels@gmail.com> writes:
>
>> OK, other implementations of PL/I language included
> [ snip ]
>> DEC System 10 and 20
> [ snip ]
>
> Details?
>
> I went from being a PL/I programmer on OS/360 to SVS to MVS, and latterly moved
> to TOPS-20 on a DEC-20. The rest of my career (~40 years) was spent on the
> latter architecture, and I've NEVER HEARD OF PL/I FOR THE PDP-10.
>
> Had I know it existed, I'd have badgered management anywhere I worked for it.
> Relentlessly.
>

I think it was under development, but I don’t think it was ever released.
Robin mentioned CPL, which was a conversational PL/I subset, like INM’s
CPL. I believe at least one of the LCM systems has it. I’d like to get the
source, although, being MACRO-10 it’s fairly useless, but I’ve never had
the time or energy to try to extract it.

--
Pete

Quadibloc <jsavard@ecn.ab.ca> writes:
> Precisely because Pascal became very popular, and offered everything
> that Algol-60 did, with important additions, the alternative of sticking with
> Algol-60 instead, if one didn't like Algol-68, wasn't a common choice.

IBM Los Gatos VLSI lab was using Metaware's TWS for various projects
.... some ref by Metaware's founders:
https://dl.acm.org/doi/pdf/10.1145/69622.357187

then the lab used it for ibm mainframe (370) pascal implementation that
they used for developing VLSI tools ... eventually it becomes VS/Pascal
and was used to implement original mainframe tcp/ip stack. I also used
it for a number of other internal applications and projects.

The IBM communication group was fiercely fighting off client/server and
distributed computing trying to preserve their dumb terminal paradigm
and install base ... and were also trying to block release of mainframe
tcp/ip ... when they lost that battle, they claimed that since the
communication group had corporate responsibility for everything that
crossed datacenter walls ... it had to be released through them. What
shipped got 44kbytes/sec aggregate using nearly whole 3090 processor.
I then did the "enhancements" to support RFC1044 and in some tuning
tests at cray research between ibm 4341 and cray, got 1mbyte aggregate
sustained using only modest amount of 4341 processor (something like
500 times improvement in bytes moved per instruction executed).

IBM Palo Alto was then doing a port of USC BSD to IBM mainframe and
needed a C compiler ... one of the LSG VLSI people responsible for
VS/Pascal was working on C language frontend for the VS/Pascal backend
... but had recently left for Metaware ... and I got Palo Alto to get
mainframe C compiler from Metaware. Palo Alto was then directed to shift
and do the BSD port to IBM's PC/RT workstation ... and they got Metaware
to do a 801/RISC (ROMP) backend for the C compiler.

a couple past (a.f.c.) posts mentioning metaware
http://www.garlic.com/~lynn/2016c.html#62 Which Books Can You Recommend For Learning Computer Programming?
http://www.garlic.com/~lynn/2004d.html#71 What terminology reflects the "first" computer language ?

Then in the IBM troubles in the 1st half of the 90s, IBM was
transferring lots of VLSI tools to industry tool vendors (but they all
had to be ported to SUN since that was the industry standard
platform). I had already left IBM, but get an (IBM) contract to port
50,000 pascal statement VLSI physical layout app to SUN. In retrospect,
it would have been simpler to have rewritten in C ... I'm not sure that
SUN pascal had been used for anything other that entry computer
classes. While SUN hdqtrs was just down the road and easy to drop into
.... SUN had outsourced their pascal support to an operation 12 time
zones away ... and everything took at least overnight.

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

On Wed, 07 Sep 2022 12:54:06 +0100, Andy Walker wrote:

> On 07/09/2022 11:56, Bob Eager wrote:
>> You could split words as long as the '%' character was introdiced, so:
>> %const %integer and %constinteger
>> were equivalent.
>
> The Walgol* compiler went further and tried to be helpful by
> completing words if it could. The trouble was that many of our students
> couldn't spell, and "proceedure" turned into "procedure e do until real"
> which was far more confusing than simply giving an error message in the
> first place. It was a great relief when Algol 68 arrived, and the word
> was spelled "proc", which even the wonkiest spellers could manage.

I learned Algol on an Elliott/ICL 4130. That had a 24 bit word and 6 bit
character code. The compiler only looked at the first four letters of a
(stropped) word. My dyslexic fellow postgrad always used 'UNTILL'.

Then we moved to a new machine and the excellent Edinburgh Algol. He had
a lot of editing to do.

--
Using UNIX since v6 (1975)...

Use the BIG mirror service in the UK:
http://www.mirrorservice.org

On Wednesday, September 7, 2022 at 3:10:01 PM UTC+10, Robin Vowels wrote:
> On Sunday, September 4, 2022 at 10:42:05 AM UTC+10, Rich Alderson wrote:
> > Robin Vowels <robin....@gmail.com> writes:
> >
> >
> > > PL/C was a pretty good implementation, and fast .
> > > Other non-IBM PL/I compilers included those for
> > > CDC, Univac, Burroughs, and on micros DR PL/I, Q1/Lite PL/I.
> > I'm surprised that no one has mentioned the GE/Honeywell PL/I compiler which
> > was used as the implementation language for much of Multics...
> .
> OK, other implementations of PL/I language included
> IBM Series/1, Burroughs B 6700 / B 7700, CDC Cyber 70, 170, 6000,
> Data General Eclipse, DEC System 10 and 20, Honeywell 66/6000,
> Q1 Corporation Q1/LMC and Q1/Lite, PLUM, SP/k compilers,
> PLAGO, Student PL/I, IBM 1130, CIMS PL/I, Olivetti A7, DR PL/I,
> Wang PL/I.
..
and Univac 1100 series, and 1106, 1108, and 1110.
..
> (In addition to the above GE/Honeywell PL/I, of course).
> .
> IBM produced the PL/I-F compiler for OS/360, and the PL/I-D compiler for smaller
> machines using DOS, and the PL/I Checkout and Optimising compilers,
> plus a number of others since.