[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[dave]

Weird day...

Computer architect Gene Amdahl was born in on this day in 1922; he had a hand 
in the IBM 704 and the System/360, founded Amdahl Corporation (maker of /360 
clones), and devised Amdahl's Law in relation to parallel processing.

But we lost Jay W. Forrester in 2016; another computer pioneer, he invented 
core memory (remember that, with its destructive read cycle?).

Oh, and LSD was first synthesised in 1938 by Dr. Hofmann of Sandoz Labs, 
Switzerland; it had nothing to do with Berkeley and BSD, man...

-- Dave

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[krewat]

On 11/15/2019 4:54 PM, Dave Horsfall wrote:
> Computer architect Gene Amdahl was born in on this day in 1922; he had 
> a hand in the IBM 704 and the System/360, founded Amdahl Corporation 
> (maker of /360 clones), and devised Amdahl's Law in relation to 
> parallel processing.

How did Amdahl get away with making 360 clones? I would have thought 
that IBM would have crushed his bones into dust.

art k.

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[dave]

On Fri, 15 Nov 2019, Arthur Krewat wrote:

> How did Amdahl get away with making 360 clones? I would have thought 
> that IBM would have crushed his bones into dust.

Clones in the way that they were able to run OS/360; that's about all that 
I can remember.  Hitachi also came out with a clone, as did no doubt many 
other manufacturers; after all, the instruction set was public 
knowledge...

I dimly recall that some opcodes had undocumented side-effects, so in 
theory (and likely in practice) OS/360 could detect whether it was running 
on a clone, and "fail to proceed" (in Rolls Royce terms).

My favourite /360 instruction on our customised system was "SVC 254" which 
got you supervisor mode (long story, depending upon the statute of 
limitations).  The operators and I had a love-hate relationship i.e. they 
loved to hate me :-)  Soon after I got a job with them i.e. better to keep 
me inside the tent than out...

-- Dave

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[peter]

On 2019-Nov-16 09:42:47 +1100, Dave Horsfall <dave at horsfall.org> wrote:
>On Fri, 15 Nov 2019, Arthur Krewat wrote:
>> How did Amdahl get away with making 360 clones? I would have thought 
>> that IBM would have crushed his bones into dust.
>
>Clones in the way that they were able to run OS/360; that's about all that 
>I can remember.  Hitachi also came out with a clone, as did no doubt many 
>other manufacturers; after all, the instruction set was public 
>knowledge...

More than just the instruction set - IBM published a formal description of
the S/360 (in APL in the IBM Systems Journal issue that announced the
S/360).  The S/360 was (I believe) the first case where a company announced
a computer architecture (rather than an implementation) and implementations
were expected to precisely comply with the architecture (no more finding
undocumented instructions and side-effects and writing code that depended
on them).  This meant that clone makers could build a clone that accurately
emulated a S/360.

>I dimly recall that some opcodes had undocumented side-effects, so in 
>theory (and likely in practice) OS/360 could detect whether it was running 
>on a clone, and "fail to proceed" (in Rolls Royce terms).

AFAIR, the only "implementation defined" instruction was DIAGNOSE, OS/360
could presumably tell what it was running on by checking particular
DIAGNOSE function.  (VM/370 was paravirtualised and used DIAGNOSE to
communicate with the hypervisor - CP).

In the early PC era, it was not uncommon for applications to verify they
were running on a genuine IBM PC by looking for the copyright notice in the
BIOS - which clone makers countered by placing a "not" before an equivalent
copyright notice.

-- 
Peter Jeremy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 963 bytes
Desc: not available
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191116/18c1df3c/attachment.sig>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[clemc]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2920 bytes --]

Btw the way diagnose instruction worked was it used jumps into the
microcode.  Very cool.  I still have a TILT deck that is a 4 card program
written using mostly 360/67 diagnose instructions.

FYI. The “DAT” (data address translation-aka the vm unit) was a separate
box attached to the side of the CPU which was filled with incandescent
lamps.  Also remember that the console bell on the 360 was a large fire
alarm style bell

This program spelled TILT on the lights of the DAT box and sent a bell char
to the console every .5 sec like a large pinball machine.  Sadly it was a
standalone program that we could only run at night but very cool none the
less.

Clem

On Sat, Nov 16, 2019 at 2:24 AM Peter Jeremy <peter at rulingia.com> wrote:

> On 2019-Nov-16 09:42:47 +1100, Dave Horsfall <dave at horsfall.org> wrote:
> >On Fri, 15 Nov 2019, Arthur Krewat wrote:
> >> How did Amdahl get away with making 360 clones? I would have thought
> >> that IBM would have crushed his bones into dust.
> >
> >Clones in the way that they were able to run OS/360; that's about all
> that
> >I can remember.  Hitachi also came out with a clone, as did no doubt many
> >other manufacturers; after all, the instruction set was public
> >knowledge...
>
> More than just the instruction set - IBM published a formal description of
> the S/360 (in APL in the IBM Systems Journal issue that announced the
> S/360).  The S/360 was (I believe) the first case where a company announced
> a computer architecture (rather than an implementation) and implementations
> were expected to precisely comply with the architecture (no more finding
> undocumented instructions and side-effects and writing code that depended
> on them).  This meant that clone makers could build a clone that accurately
> emulated a S/360.
>
> >I dimly recall that some opcodes had undocumented side-effects, so in
> >theory (and likely in practice) OS/360 could detect whether it was
> running
> >on a clone, and "fail to proceed" (in Rolls Royce terms).
>
> AFAIR, the only "implementation defined" instruction was DIAGNOSE, OS/360
> could presumably tell what it was running on by checking particular
> DIAGNOSE function.  (VM/370 was paravirtualised and used DIAGNOSE to
> communicate with the hypervisor - CP).
>
> In the early PC era, it was not uncommon for applications to verify they
> were running on a genuine IBM PC by looking for the copyright notice in the
> BIOS - which clone makers countered by placing a "not" before an equivalent
> copyright notice.
>
> --
> Peter Jeremy
> _______________________________________________
> COFF mailing list
> COFF at minnie.tuhs.org
> https://minnie.tuhs.org/cgi-bin/mailman/listinfo/coff
>
-- 
Sent from a handheld expect more typos than usual
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191116/c7ea253d/attachment.html>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[dave]

On Sat, 16 Nov 2019, Peter Jeremy wrote:

> More than just the instruction set - IBM published a formal description 
> of the S/360 (in APL in the IBM Systems Journal issue that announced the 
> S/360).  The S/360 was (I believe) the first case where a company 
> announced a computer architecture (rather than an implementation) and 
> implementations were expected to precisely comply with the architecture 
> (no more finding undocumented instructions and side-effects and writing 
> code that depended on them).  This meant that clone makers could build a 
> clone that accurately emulated a S/360.

Ah, I'd forgotten about the APL documentation; thanks!  Talk about giving 
away the keys to the kingdom: Amdahl, Fujitsu, Hitachi...

> AFAIR, the only "implementation defined" instruction was DIAGNOSE, 
> OS/360 could presumably tell what it was running on by checking 
> particular DIAGNOSE function.  (VM/370 was paravirtualised and used 
> DIAGNOSE to communicate with the hypervisor - CP).

Another point I had forgotten :-(  Yep, the DIAG instruction was utterly 
implementation-dependent, and thus OS/360 could tell whether it was 
running on a clone or not.

Cut me some slack; I turned 67 last month :-(

> In the early PC era, it was not uncommon for applications to verify they 
> were running on a genuine IBM PC by looking for the copyright notice in 
> the BIOS - which clone makers countered by placing a "not" before an 
> equivalent copyright notice.

I remember the days of the "grey imports" (or "gray" for the Americans); 
if it ran Flight Simulator and not labelled "IBM" then it was technically 
illegal; shortly afterwards if it did *not* run FS then nobody would buy 
it :-)

How things change...

-- Dave

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[clemc]

On Sun, Nov 17, 2019 at 12:14 AM Dave Horsfall <dave at horsfall.org> wrote:

> On Sat, 16 Nov 2019, Peter Jeremy wrote:
>
> > More than just the instruction set - IBM published a formal description
> > of the S/360 (in APL in the IBM Systems Journal issue that announced the
> > S/360).  The S/360 was (I believe) the first case where a company
> > announced a computer architecture (rather than an implementation) and
> > implementations were expected to precisely comply with the architecture
> > (no more finding undocumented instructions and side-effects and writing
> > code that depended on them).  This meant that clone makers could build a
> > clone that accurately emulated a S/360.
>
> Ah, I'd forgotten about the APL documentation; thanks!  Talk about giving
> away the keys to the kingdom: Amdahl, Fujitsu, Hitachi...
>

The cat was already out and poking around with the publishing of:  IBM 360
Principles of Operation, DOC A22-6821-0
<http://bitsavers.trailing-edge.com/pdf/ibm/360/princOps/A22-6821-0_360PrincOps.pdf>.
The
APL version of spec just gave it more area to roam.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191118/bb3bd36e/attachment.html>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[bakul]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1718 bytes --]

Are you guys talking about “A formal description of System/360” by Falkoff, Iverson and Sussenguth? It uses an APL like notation but not exactly a S/360 emulator in APL! Much more concise than the S/360 POP.

http://www.cs.trinity.edu/~jhowland/class.files.cs2321.html/falkoff.pdf

>> On Nov 18, 2019, at 8:43 AM, Clem Cole <clemc at ccc.com> wrote:
> 
> 
> 
>> On Sun, Nov 17, 2019 at 12:14 AM Dave Horsfall <dave at horsfall.org> wrote:
>> On Sat, 16 Nov 2019, Peter Jeremy wrote:
>> 
>> > More than just the instruction set - IBM published a formal description 
>> > of the S/360 (in APL in the IBM Systems Journal issue that announced the 
>> > S/360).  The S/360 was (I believe) the first case where a company 
>> > announced a computer architecture (rather than an implementation) and 
>> > implementations were expected to precisely comply with the architecture 
>> > (no more finding undocumented instructions and side-effects and writing 
>> > code that depended on them).  This meant that clone makers could build a 
>> > clone that accurately emulated a S/360.
>> 
>> Ah, I'd forgotten about the APL documentation; thanks!  Talk about giving 
>> away the keys to the kingdom: Amdahl, Fujitsu, Hitachi...
> 
> The cat was already out and poking around with the publishing of:  IBM 360 Principles of Operation, DOC A22-6821-0. The APL version of spec just gave it more area to roam.
> _______________________________________________
> COFF mailing list
> COFF at minnie.tuhs.org
> https://minnie.tuhs.org/cgi-bin/mailman/listinfo/coff
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191118/b9b188d2/attachment.html>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[clemc]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 3067 bytes --]

I never said or tried to imply that it anything more than an architectural
description.  But I was saying S/360 was publically described before that
in the principles of operation by IBM.  At that point, you were fine to
clone it.

In fact, this is what put Ken O at Cal Data out of business a few years
later when he cloned the PDP-11 ISA (from the processor handbook
descriptions) and cloned the Unibus which had only been described in
schematics.   Thus Ken was careful later to only make bus repeaters/Caches
and the like at his new firm Able Computer (although he built, but I do not
believe he ever actually sold, a M68010 on a 22-bit QBUS at one point).

As it turns out, under those rules, its possible Amdahl might have been in
violation of IBM's IP, but again the difference was that the amount of
money he was taking from IBM was small, and IBM was (correctly worried)
about the Justice Dept.  I fact a few years later IBM was enjoined for
monopolistic practices (1969 and vacated without merit in 1980:
https://web.archive.org/web/20100408174629/http://www.hagley.lib.de.us/library/collections/manuscripts/findingaids/ibmantitrustpart2.ACC1980.htm
).


ᐧ

On Mon, Nov 18, 2019 at 1:45 PM Bakul Shah <bakul at bitblocks.com> wrote:

> Are you guys talking about “A formal description of System/360” by
> Falkoff, Iverson and Sussenguth? It uses an APL like notation but not
> exactly a S/360 emulator in APL! Much more concise than the S/360 POP.
>
> http://www.cs.trinity.edu/~jhowland/class.files.cs2321.html/falkoff.pdf
>
> On Nov 18, 2019, at 8:43 AM, Clem Cole <clemc at ccc.com> wrote:
>
> 
>
>
> On Sun, Nov 17, 2019 at 12:14 AM Dave Horsfall <dave at horsfall.org> wrote:
>
>> On Sat, 16 Nov 2019, Peter Jeremy wrote:
>>
>> > More than just the instruction set - IBM published a formal description
>> > of the S/360 (in APL in the IBM Systems Journal issue that announced
>> the
>> > S/360).  The S/360 was (I believe) the first case where a company
>> > announced a computer architecture (rather than an implementation) and
>> > implementations were expected to precisely comply with the architecture
>> > (no more finding undocumented instructions and side-effects and writing
>> > code that depended on them).  This meant that clone makers could build
>> a
>> > clone that accurately emulated a S/360.
>>
>> Ah, I'd forgotten about the APL documentation; thanks!  Talk about giving
>> away the keys to the kingdom: Amdahl, Fujitsu, Hitachi...
>>
>
> The cat was already out and poking around with the publishing of:  IBM
> 360 Principles of Operation, DOC A22-6821-0
> <http://bitsavers.trailing-edge.com/pdf/ibm/360/princOps/A22-6821-0_360PrincOps.pdf>. The
> APL version of spec just gave it more area to roam.
> _______________________________________________
> COFF mailing list
> COFF at minnie.tuhs.org
> https://minnie.tuhs.org/cgi-bin/mailman/listinfo/coff
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191118/ce2b09e8/attachment.html>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[peter]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 944 bytes --]

On 2019-Nov-18 10:45:13 -0800, Bakul Shah <bakul at bitblocks.com> wrote:
>Are you guys talking about “A formal description of System/360” by
>Falkoff, Iverson and Sussenguth? It uses an APL like notation but not
>exactly a S/360 emulator in APL! Much more concise than the S/360 POP.
>
>http://www.cs.trinity.edu/~jhowland/class.files.cs2321.html/falkoff.pdf

I was referring to exactly that but didn't have the reference handy.
Note that it's not "an APL like notation", it's the APL defined by Ken
Iverson in "A Programming Language"[1].  And, no-one said it was an
"emulator".

[1] http://www.softwarepreservation.org/projects/apl/Books/APROGRAMMING%20LANGUAGE

-- 
Peter Jeremy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 963 bytes
Desc: not available
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191120/63272abc/attachment.sig>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[cym224]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 969 bytes --]

On 19/11/2019, Peter Jeremy <peter at rulingia.com> wrote:
> On 2019-Nov-18 10:45:13 -0800, Bakul Shah <bakul at bitblocks.com> wrote:
>>Are you guys talking about “A formal description of System/360” by
>>Falkoff, Iverson and Sussenguth? It uses an APL like notation but not
>>exactly a S/360 emulator in APL! Much more concise than the S/360 POP.
>>
>>http://www.cs.trinity.edu/~jhowland/class.files.cs2321.html/falkoff.pdf
>
> I was referring to exactly that but didn't have the reference handy.
> Note that it's not "an APL like notation", it's the APL defined by Ken
> Iverson in "A Programming Language"[1].  And, no-one said it was an
> "emulator".
>
> [1] http://www.softwarepreservation.org/projects/apl/Books/APROGRAMMING%20LANGUAGE
>
> --
> Peter Jeremy

Blaauw and Brooks documented an enormous number of architectures that
way: https://www.pearson.com/us/higher-education/program/Blaauw-Computer-Architecture-Concepts-and-Evolution/PGM629.html

N.

>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[dave]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 388 bytes --]

On Mon, 18 Nov 2019, Clem Cole wrote:

> The cat was already out and poking around with the publishing of:  IBM 
> 360 Principles of Operation, DOC A22-6821-0. The APL version of spec 
> just gave it more area to roam.

Yeah, I know about PrincOps; it went into excruciating detail, leading to 
the observation that IBM manuals were designed to impress rather than 
inform...

-- Dave

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[clemc]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 3654 bytes --]

On Fri, Nov 15, 2019 at 5:27 PM Arthur Krewat <krewat at kilonet.net> wrote

>
> How did Amdahl get away with making 360 clones? I would have thought
> that IBM would have crushed his bones into dust.
>

There were a number of things going on.  Remember the world was different.
 First and foremost, was that IBM was so big and other large firms were
getting out the business such as GE and Xerox.  As I understand it from my
IBM friends that were there at the time was that IBM was concerned that the
justice dept would go after them.  The amount of business Amdahl bleed off
was small compared to what IBM was making, so I suspect that it played into
any business decisions.

Amdahl legitimized a lot of IBM's practices and in many ways, and as the
Harvard Business School teaches, "It is always better to have a small weak
competitor than none at all."  At the time the price difference (certainly
for universities) was huge.   I remember when CMU was making the decision
to replace the IBM 360/67, IBM bid a 370/168 an Amdahl a 470/v7 and DEC a
PDP-20 (DEC won).   IIRC: The difference in quotes between IBM and Amdahl
was a factor of two.  DEC was cheaper still, although, in the end, CMU had
to buy 2 of them to do what they had been doing with the 360 previously.

Also, remember the IBM OS's were published and thus all the sites had the
full sources. It was built on-site for that specific installation and each
site tended to have made small local mods.    PTF's (Program
Temporary Fixes - *aka* patches) came out from IBM as the source and you
applied the PTFs yourself (my memory was for most PFTs we had about a 2-3
month lag from recent from IBM before we had them in to the system).  But
folks took the original code whole cloth and changed it too.   For
instance, CMU took the TSS source which did not work (crashed every few
hours) after a year or so, had fixed it to work reliably enough to be the
system running on the 360/67 24/7.  On the other hand, U Mich took the TSS
sources and rewrote it completely to create MTS.  The IBM/MIT team created
the precursor to VM.  Also, the user-level code like compilers was
reasonably movable between different OS (for instance my first paying
programming job was in the CMU's IBM shop moving York/APL which was written
for OS/360 to CMU's TSS).

Also, a lot of Amdahl customers ran MTS, not the basic IBM OS.  Also, at
one point I was told by one of my former CMU co-workers who has moved to
IBM to work on TSS, that there were more TSS customers on Amdahl equipment
than IBM.  But IBM kept the TSS group alive for a long time.  [I'm not sure
what the relationship was for support to be honest as I never lived it].

Note from my later LCC days, I was also under the impression that a lot of
TSS sites were the ones that AIX/370 targeted to get them back into the IBM
fold.   But it also took a lower-cost model of the 370 before that happened
because they were targeting University types.

Finally, folks with Amdahl machines just looked at the PTFs and
reimplemented them if they used an IBM specific trick (how they got them
I'm not sure).   I do remember that it was not unusual for SHARE (the IBM
user's group) to have Amdahl specific PTFs and often fixes/updates to the
IBM code that came from folks rewriting them.  I don't know how that all
worked.  I just got a stack of PTFs and had to deal with them, but I do
remember some came from IBM some from SHARE and in the later seeing
comments about running on Amdahl.
ᐧ
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20191115/ee0cdb97/attachment.html>

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[krewat]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 718 bytes --]

On 11/15/2019 6:19 PM, Clem Cole wrote:
> As I understand it from my IBM friends that were there at the time was 
> that IBM was concerned that the justice dept would go after them.  The 
> amount of business Amdahl bleed off was small compared to what IBM was 
> making, so I suspect that it played into any business decisions.

Now I understand ;) I was so isolated from IBM in my early life, that I 
didn't really get involved with them until the early 2000's. That was on 
RS-6000's and AIX on the hardware side, and then IBM outsourcing on the 
system administration side at one of my Fortune 100 customers.

They were still the elephant in the room, but they didn't seem as 
predatory as I had heard.

art k.

[COFF] In Memoriam: Jay W. Forrester, happy birthday Gene Amdahl, and LSD

[dave]

Weird day...

Computer architect Gene Amdahl was born in on this day in 1922; he had a 
hand in the IBM 704 and the System/360, founded Amdahl Corporation (maker 
of /360 clones), and devised Amdahl's Law in relation to parallel 
processing.

But we lost Jay W. Forrester in 2016; another computer pioneer, he 
invented core memory (remember that, with its destructive read cycle?).

Oh, and LSD was first synthesised in 1938 by Dr. Hofmann of Sandoz Labs, 
Switzerland; it had nothing to do with Berkeley and BSD, man...

-- Dave