11 Nov
2019
11 Nov
'19
9:10 p.m.
Hi.
Doug McIlroy is probably the best person to answer this.
Looking at the V3 and V4 manuals, there is a reference to the m6 macro
processor. The man page thereof refers to
A. D. Hall, The M6 Macroprocessor, Bell Telephone Laboratories, 1969
1. Is this memo available, even in hardcopy that could be scanned?
2. What's the history of m6, was it written in assembler? C?
3. When and why was it replaced with m4 (written by DMR IIRC)?
More generally, what's the history of m6 prior to Unix?
IIRC, the macro processor in Software Tools was inspired by m4,
and in particular its immediate evaluation of its arguments during
definition.
I guess I'll also ask, how widespread was the use of macro processors
in high level languages? They were big for assembler, and PL/1 had
a macro language, but I don't know of any other contemporary languages
that had them. Were the general purpose macro processors used a lot?
E.g. with Fortran or Cobol or ...
I'm just curious. :-)
Thanks,
Arnold
11 Nov
11 Nov
10:18 p.m.
Looking at the V3 and V4 manuals, there is a reference
to the m6
macro processor. The man page thereof refers to
A. D. Hall, The M6 Macroprocessor, Bell Telephone
Laboratories, 1969
1. Is this memo available, even in hardcopy that could
be scanned?
Amen - the cited document, CSTR #2 seems to be hard to find.
De
12 Nov
12 Nov
4:01 p.m.
Dennis Boone <drb(a)msu.edu> writes:
Indeed, it took a bit to find:
https://plan9.io/cm/cs/cstr/2.pdf
(Also it's not on archive.org for the original bell-labs.com domain;
it would be nice to have it officially archived.)
Best,
--
Leah Neukirchen <leah(a)vuxu.org> https://leahneukirchen.org/
Looking at the
V3 and V4 manuals, there is a reference to the m6
macro processor. The man page thereof refers to
A. D. Hall, The M6 Macroprocessor, Bell Telephone
Laboratories, 1969
1. Is this memo available, even in hardcopy that
could be scanned?
Amen - the cited document, CSTR #2 seems to be hard to find. 
11 Nov
11 Nov
10:31 p.m.
On Mon, Nov 11, 2019 at 11:10:26PM +0200, Arnold Robbins wrote:
I guess I'll also ask, how widespread was the use
of macro processors
in high level languages? They were big for assembler, and PL/1 had
a macro language, but I don't know of any other contemporary languages
that had them. Were the general purpose macro processors used a lot?
E.g. with Fortran or Cobol or ...
Well, there's Ratfor (Rational Fortran), which was implemented as a
preprocessor in front of Fortran. I don't think it was used a lot,
though....
- Ted
10:58 p.m.
On Mon, 11 Nov 2019, Theodore Y. Ts'o wrote:
On Mon, Nov 11, 2019 at 11:10:26PM +0200, Arnold
Robbins wrote:
> I guess I'll also ask, how widespread was the use of macro processors
> in high level languages? They were big for assembler, and PL/1 had
> a macro language, but I don't know of any other contemporary languages
> that had them. Were the general purpose macro processors used a lot?
> E.g. with Fortran or Cobol or ...
Late trivia: it's PL/I, not PL/1; rumour has it that IBM trademarked them
all up to PL/C i.e. PL/100...
Well, there's Ratfor (Rational Fortran), which
was implemented as a
preprocessor in front of Fortran. I don't think it was used a lot,
though....
I think I used it once, on the principle that I'll try anything once;
ugh...
-- Dave
12 Nov
12 Nov
12:07 a.m.
On 11/11/19 17:58, Dave Horsfall wrote:
I worked for a company that wrote everything on ratfor. #6-)
On Mon, 11 Nov 2019, Theodore Y. Ts'o wrote:
Found nothing on TESS. Maybe copyright but I am not familiar with
copyright.
On Mon, Nov 11, 2019 at 11:10:26PM +0200, Arnold
Robbins wrote:
> I guess I'll also ask, how widespread was the use of macro processors
> in high level languages? They were big for assembler, and PL/1 had
> a macro language, but I don't know of any other contemporary languages
> that had them. Were the general purpose macro processors used a lot?
> E.g. with Fortran or Cobol or ...
Late trivia: it's PL/I, not PL/1; rumour has it that IBM trademarked
them all up to PL/C i.e. PL/100... Well, there's Ratfor (Rational Fortran),
which was implemented as a
preprocessor in front of Fortran. I don't think it was used a lot,
though....
I think I used it once, on the principle that I'll try anything once;
ugh...
-- Dave
12:39 a.m.
On Mon, 11 Nov 2019, Nemo Nusquam wrote:
Could be copyright; I was told this in the early 70s when I was doing
Computer Science. Of course, it could also be apocryphal :-(
Late trivia:
it's PL/I, not PL/1; rumour has it that IBM trademarked
them all up to PL/C i.e. PL/100...
Found nothing on TESS. Maybe copyright but I am not familiar with
copyright. I worked for a company that wrote everything on
ratfor. #6-)
Urk...
-- Dave
12:30 a.m.
Not so fast there my friend. Ratfor was used a great deal. For instance The PGI compiler
folks wrote a parallel optimizing compiler for the FPS-164 in it. And FPS used it for
there OS. Plus Apollo wrote all their original utilities in it.
Sent from my PDP-7 Running UNIX V0 expect things to be almost but not quite.
On Nov 11, 2019, at 5:31 PM, Theodore Y. Ts'o
<tytso(a)mit.edu> wrote:
On Mon, Nov 11, 2019 at 11:10:26PM +0200, Arnold
Robbins wrote:
I guess I'll also ask, how widespread was the use of macro processors
in high level languages? They were big for assembler, and PL/1 had
a macro language, but I don't know of any other contemporary languages
that had them. Were the general purpose macro processors used a lot?
E.g. with Fortran or Cobol or ...
Well, there's Ratfor (Rational Fortran), which was implemented as a
preprocessor in front of Fortran. I don't think it was used a lot,
though....
- Ted
12:42 a.m.
Not so fast there my friend. Ratfor was used a great
deal. For
instance The PGI compiler folks wrote a parallel optimizing compiler
for the FPS-164 in it. And FPS used it for there OS. Plus Apollo
wrote all their original utilities in it.
The Software Tools stuff too, including the STVOS stuff Berkeley
managed. If you pick an old platform and count RATFOR implementations,
all of them seem to have enough of them you can't swing a dead cat.
RATFOR may not be enjoyable in modern terms, but compared to building in
native F66...
De
16 Nov
16 Nov
4:27 p.m.
El mar., 12 nov. 2019 1:31, Clem cole <clemc(a)ccc.com> escribió:
Not so fast there my friend. Ratfor was used a great
deal. For instance
The PGI compiler folks wrote a parallel optimizing compiler for the FPS-164
in it. And FPS used it for there OS. Plus Apollo wrote all their original
utilities in it.
My first FORTRAN textbook was titled "FORTRAN with WATFOR and WATFIV". It
had a long print run as well.
Cordiales saludos / Best Regards / Salutations / Freundliche Grüße
-----
Sergio Pedraja
17 Nov
17 Nov
5:30 a.m.
On Sat, 16 Nov 2019, SPC wrote:
My first FORTRAN textbook was titled "FORTRAN
with WATFOR and WATFIV".
It had a long print run as well.
Now *that* brings back memories (not necessarily pleasant). WATFOR was as
ugly as sin, but WATFIV was much better i.e. it actually knew about stuff
such as labelled common etc (I could be wrong here).
And I still feel like taking a shower after discussing FORTRAN :-)
Then we got ALGOLW, and it was heavenly. PL/360 had a good run too.
-- Dave
5:50 a.m.
On Sun, Nov 17, 2019 at 04:30:15PM +1100, Dave Horsfall wrote:
On Sat, 16 Nov 2019, SPC wrote:
I'm pretty sure that was the Fortran I learned. Yeah, it was not C. But
it was math. I spent a bunch of time learning accumulated errors and
more time on floating point numbers. My dad was a theoretical physics
guy so I did some coding for him. I respected Fortran for what it could
do but I developed a hate for floating point. In my mind, floating
point numbers meant you couldn't handle the world you were working in.
It just felt like you could shift the domain you were working in so
integers could work. If you couldn't do that, you were admitting that
you were not accurate.
My first FORTRAN textbook was titled "FORTRAN
with WATFOR and WATFIV". It
had a long print run as well.
Now *that* brings back memories (not necessarily pleasant). WATFOR was as
ugly as sin 
6:12 p.m.
On Sat, 16 Nov 2019 21:50:58 -0800 Larry McVoy <lm(a)mcvoy.com> wrote:
Larry McVoy writes:
On Sun, Nov 17, 2019 at 04:30:15PM +1100, Dave
Horsfall wrote:
Many numbers can't be represented perfectly using integers or
rationals (a pair of integers) but can be computed using a
series expansion to arbitrary precision. I thought FP numbers
were a clever & practical compromise that worked quite well.
David Goldberg's "What every computer scientist should know
about floating-point" is worth reading.
https://www.itu.dk/~sestoft/bachelor/IEEE754_article.pdf
Earlier I remember reading the "Numerical Recipes" books by
Press, Teukolsky, Vetterling & Flannery. IIRC, the original
version used Fortran. They also had versions using Pascal and
C (I finally bought the C version in '80s though never used it).
Note that Scheme & CL have a full complement of numeric types:
big nums, rationals, reals and complex numbers. At least some
versions of CL have arbitrary precision FP numbers.
What I really want is a programming language with support for
symbolic manipulation of formulas!
On Sat, 16 Nov 2019, SPC wrote:
I'm pretty sure that was the Fortran I learned. Yeah, it was not C. But
it was math. I spent a bunch of time learning accumulated errors and
more time on floating point numbers. My dad was a theoretical physics
guy so I did some coding for him. I respected Fortran for what it could
do but I developed a hate for floating point. In my mind, floating
point numbers meant you couldn't handle the world you were working in.
It just felt like you could shift the domain you were working in so
integers could work. If you couldn't do that, you were admitting that
you were not accurate. My first FORTRAN textbook was titled "FORTRAN
with WATFOR and WATFIV". It
had a long print run as well.
Now *that* brings back memories (not necessarily pleasant). WATFOR was as
ugly as sin 
6:23 p.m.
On 17 Nov 2019 10:12 -0800, from bakul(a)bitblocks.com (Bakul Shah):
What I really want is a programming language with
support for
symbolic manipulation of formulas!
Do you mean something like HAL/S?
https://en.wikipedia.org/wiki/HAL/S#Syntax
Though quite frankly, at a glance that multi-line syntax looks
somewhat awkward for little benefit over a single-line format,
especially if the code is already annotated (as certainly I would
expect for in-flight avionics software, never mind aboard a
spacecraft).
--
Michael Kjörling • https://michael.kjorling.se • michael(a)kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”
6:56 p.m.
On Sun, 17 Nov 2019 18:23:05 +0000 Michael =?utf-8?B?S2rDtnJsaW5n?=
<michael(a)kjorling.se> wrote:
Michael Kjörling writes:
On 17 Nov 2019 10:12 -0800, from bakul(a)bitblocks.com
(Bakul Shah):
Syntax is a separate issue. Something like what Macsyma (or
Maxima now) or even LaTeX would be fine. Many word processors
and browsers can render the latter properly. Syntax is the
easy part (though it can be controversial). The hard part is
having to build in a tremendous amount of math knowledge. Even
if the compiler/interpreter were to outsource the symbolic
algebra manipulation part to maxima, mathematica or some such,
smooth integration with other language features would not be easy.
What I really want is a programming language with
support for
symbolic manipulation of formulas!
Do you mean something like HAL/S?
https://en.wikipedia.org/wiki/HAL/S#Syntax
Though quite frankly, at a glance that multi-line syntax looks
somewhat awkward for little benefit over a single-line format,
especially if the code is already annotated (as certainly I would
expect for in-flight avionics software, never mind aboard a
spacecraft). 
10:46 p.m.
Just a note:
(I am new to this list and find the history revealed fascinating, so
thank you all for insights.)
There is an interesting paper on symbolic formula manipulation:
https://pdfs.semanticscholar.org/0367/25f0abfd88879dd88d77e3a5e51915db5f1b.…
There is also a symbolic Python library: https://www.sympy.org/en/index.html
On 11/17/2019 10:12 AM, Bakul Shah wrote:
On Sat, 16 Nov 2019 21:50:58 -0800 Larry McVoy
<lm(a)mcvoy.com> wrote:
Larry McVoy writes:
On Sun, Nov 17, 2019 at 04:30:15PM +1100, Dave
Horsfall wrote:
Many numbers can't be represented perfectly using
integers or
rationals (a pair of integers) but can be computed using a
series expansion to arbitrary precision. I thought FP numbers
were a clever & practical compromise that worked quite well.
David Goldberg's "What every computer scientist should know
about floating-point" is worth reading.
https://www.itu.dk/~sestoft/bachelor/IEEE754_article.pdf
Earlier I remember reading the "Numerical Recipes" books by
Press, Teukolsky, Vetterling & Flannery. IIRC, the original
version used Fortran. They also had versions using Pascal and
C (I finally bought the C version in '80s though never used it).
Note that Scheme & CL have a full complement of numeric types:
big nums, rationals, reals and complex numbers. At least some
versions of CL have arbitrary precision FP numbers.
What I really want is a programming language with support for
symbolic manipulation of formulas!
On Sat, 16 Nov 2019, SPC wrote:
I'm pretty sure that was the Fortran I learned. Yeah, it was not
C. But
it was math. I spent a bunch of time learning accumulated errors and
more time on floating point numbers. My dad was a theoretical physics
guy so I did some coding for him. I respected Fortran for what it could
do but I developed a hate for floating point. In my mind, floating
point numbers meant you couldn't handle the world you were working in.
It just felt like you could shift the domain you were working in so
integers could work. If you couldn't do that, you were admitting that
you were not accurate. My first FORTRAN textbook was titled
"FORTRAN with WATFOR and WATFIV". It
had a long print run as well.
Now *that* brings back memories (not necessarily
pleasant). WATFOR was as
ugly as sin
11 Nov
11 Nov
10:37 p.m.
On 11 Nov 2019 23:10 +0200, from arnold(a)skeeve.com (Arnold Robbins):
Looking at the V3 and V4 manuals, there is a reference
to the m6 macro
processor. The man page thereof refers to
A. D. Hall, The M6 Macroprocessor, Bell Telephone Laboratories, 1969
2. What's the history of m6, was it written in assembler? C?
If it was from 1969, it can't realistically have been written in C,
which at that time wouldn't be around in any form for another few
years.
--
Michael Kjörling • https://michael.kjorling.se • michael(a)kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”
12 Nov
12 Nov
1:09 a.m.
Doug can certainly give you more information, but very early in my
career at Bell Labs I saw an internal memo, by (I think) Doug and Bob
Morris, that had a taxonomy of macro systems based on a set of
decisions (things like can you define a macro inside a macro, rescan
expanded macros looking for macro definitions and uses, etc.) I
seem to recall there were close to a thousand different macro systems
possible. It was a brilliant paper, but I don't think it was ever
published. The Assembler for the IBM 7094 had an amazing macro
facility -- I recall that someone wrote a Lisp compiler entirely in
macros, and it would regularly recurse several hundred levels deep
while generating code. Someone told me once that Bob Morris invented
macros -- I can't vouch for this. Doug?
Steve
----- Original Message -----
From: "Arnold Robbins" <arnold(a)skeeve.com>
To:<tuhs@tuhs.org>
Cc:
Sent:Mon, 11 Nov 2019 23:10:26 +0200
Subject:[TUHS] History of m6?
Hi.
Doug McIlroy is probably the best person to answer this.
Looking at the V3 and V4 manuals, there is a reference to the m6
macro
processor. The man page thereof refers to
A. D. Hall, The M6 Macroprocessor, Bell Telephone Laboratories, 1969
1. Is this memo available, even in hardcopy that could be scanned?
2. What's the history of m6, was it written in assembler? C?
3. When and why was it replaced with m4 (written by DMR IIRC)?
More generally, what's the history of m6 prior to Unix?
IIRC, the macro processor in Software Tools was inspired by m4,
and in particular its immediate evaluation of its arguments during
definition.
I guess I'll also ask, how widespread was the use of macro processors
in high level languages? They were big for assembler, and PL/1 had
a macro language, but I don't know of any other contemporary
languages
that had them. Were the general purpose macro processors used a lot?
E.g. with Fortran or Cobol or ...
I'm just curious. :-)
Thanks,
Arnold
3:07 p.m.
On Mon, Nov 11, 2019 at 8:27 PM Steve Johnson <scj(a)yaccman.com> wrote:
I recall that someone wrote a Lisp compiler entirely
in macros, and it
would regularly recurse several hundred levels deep while generating code.
T'was before my time, but the legend has it that the original BLISS-10
bootstrap compiler was a set of TECO macros that Chuck Geschke (Adobe's
founder) wrote.
ᐧ
13 Nov
13 Nov
9:16 a.m.
'T'was before my time, but the legend has it that the original BLISS-10
bootstrap compiler was a set of TECO macros that Chuck Geschke (Adobe's
founder) wrote.'
Really? TECO = Tape Editor and Corrector
12:20 p.m.
Thomas Paulsen wrote:
'T'was before my time, but the legend has
it that the original BLISS-10
bootstrap compiler was a set of TECO macros that Chuck Geschke (Adobe's
founder) wrote.'
Really? TECO = Tape Editor and Corrector
Why not? TECO is both an editor and a programming language. The first
Emacs was written in TECO.
12:50 p.m.
On Wed, Nov 13, 2019 at 7:21 AM Lars Brinkhoff <lars(a)nocrew.org> wrote:
Thomas Paulsen wrote:
Pi calculated with teco: http://www.iwriteiam.nl/HaPi_TECO_macro.html
This impresses me, partly because it looks like someone fell asleep on a
keyboard.
Running it just now:
qsb$ tecoc mung pi.tec,30
314159265358979323846264338327
Mike Markowski
'T'was before my time, but the legend
has it that the original BLISS-10
bootstrap compiler was a set of TECO macros that Chuck Geschke (Adobe's
founder) wrote.'
Really? TECO = Tape Editor and Corrector
Why not? TECO is both an editor and a programming language. The first
Emacs was written in TECO.
1:02 p.m.
Mike Markowski wrote:
Pi calculated with teco:
http://www.iwriteiam.nl/HaPi_TECO_macro.html
This impresses me, partly because it looks like someone fell asleep on a
keyboard.
As does this:
https://news.ycombinator.com/item?id=13519471
Perhaps COFF is more appropriate for TECO discussions.
7:19 p.m.
Paul Winalski wrote:
So the story goes. But I'm not sure sure. In November 1976, the name
went from ? and ?MACS to E and EMACS. Maybe "editor" was implied, but
it doesn't come up in the email conversation.
The first
Emacs was written in TECO.
Which is how it got its name, of course. Emacs =
"Editor MACroS". 
7:21 p.m.
Lars Brinkhoff writes:
Paul Winalski wrote:
So the story goes. But I'm not sure sure. In November 1976, the name
went from ? and ?MACS to E and EMACS. Maybe "editor" was implied, but
it doesn't come up in the email conversation.
For what it's worth, I have some vague memory of attending a lunchtime
presentation at BTL in maybe the summer of 72 or 73 on EMACS as TECO
editor macros.
Jon
The first
Emacs was written in TECO.
Which is how it got its name, of course. Emacs =
"Editor MACroS".
14 Nov
14 Nov
9:26 a.m.
'Why not? TECO is both an editor and a programming language. The first
Emacs was written in TECO.'
Hic Rhodos - Hic Salta!
I have a account on a remote twenex PDP10. There is a editor named emacs. This is a very
archaic piece of software. It doesn't know any teco commands no matter how I tried.
I'm pretty sure that this is teco-emacs. I draw my own conclusions from these
observations which are far away from all these myths.
10:53 a.m.
I suggest continuing this on COFF.
Thomas Paulsen wrote:
I have a account on a remote twenex PDP10. There is a
editor named
emacs. This is a very archaic piece of software. It doesn't know any
teco commands no matter how I tried. I'm pretty sure that this is
teco-emacs.
Yes, it should be TECO Emacs. Normal use of Emacs rarely needs TECO
commands.
To get a TECO minibuffer type Meta-Altmode (Esc Esc). You should get a
small window at the top of the terminal in which you can enter TECO
commands. Execute them with double altmode as you would in any TECO.
I draw my own conclusions from these observations
which are far away
from all these myths.
I try to stay with the facts. I actually use TECO Emacs almost daily,
and I have built it from sources. Some other information is based on
email conversations among those who wrote Emacs.
10:53 p.m.
On Nov 13, 2019, at 10:16, Thomas Paulsen
<thomas.paulsen(a)firemail.de> wrote:
'T'was before my time, but the legend has it that the original BLISS-10
bootstrap compiler was a set of TECO macros that Chuck Geschke (Adobe's
founder) wrote.'
Really? TECO = Tape Editor and Corrector
Yes, and the TAPE in the mnemonic is paper tape.
In an former live we once had a text formatter running on OS/8
(pdp-8) written in TECO.
jaap
1632
days inactive
1638
days old
29 comments
20 participants
participants (20)
-
Arnold Robbins -
Bakul Shah -
Barry Stanly -
Clem cole -
Clem Cole
-
Dave Horsfall -
Dennis Boone -
Jaap Akkerhuis -
Jon Steinhart -
Larry McVoy -
Lars Brinkhoff -
Leah Neukirchen -
Michael Kjörling -
Mike Markowski -
Nemo Nusquam -
Paul Winalski -
SPC -
Steve Johnson -
Theodore Y. Ts'o -
Thomas Paulsen