IBM METALC curiosity.

#4146

While trying to understand a bit about IBM PL/S
and "Rand Compiler for PL/S", after watching a
YouTube presentation on this topic, my attention
had been caught by this old thread (2007):

  

about IBM Z/OS METALC.

Not sure I understand what was going on in the
discussion, a bit of help from the hercules MVS
group maybe may let me to understand the topic.

If I got the picture correctly, IBM METALC may
be used in a similar way of the old PL/S compiler
under MVS under a modern Z/OS (AMODE31/64) for
system programming.

It looks to my eyes that its library doesn't depend
from the XLC runtime environment, the METALC compiler
just generate standalone assembler (HLASM) source code as
the PL/S compiler and assembler code may be inlined
in the C code with the "__asm" directive as was for
PL/S with the CODEGEN directive.

The machine code should be indipendent from the
XLC runtime environment, any I/O operation
must be performed through assembler code using
directly the Z/OS control blocks (like DCB)
through the HLASM assembler macro libraries.

This code fragment, took from the public
IBM manuals:

     struct WTO_PARM {
       unsigned short len;
       unsigned short code;
       char text[80];
     } wto_buff = { 4+11, 0, "hello world" };

     __asm( " WTO  MF=(E,(%0)) " : : "r"(&wto_buff));

may give an idea of what I'm speaking about for
printing a message on the Z/OS console.

I hope this is a legit curiosity on this group
which may lead to an interesting discussion.

Peppe.

---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4147

Good evening,?

A little while back I wrote some C code, for the GCC compiler using embedded assembler, that made a few calls to MVS macros: WTO, WTOR, STIMER and WAIT. This was on MVS3.8j TK4-. I suspect the concept would work with new versions of the compiler on more modern versions of MVS.?

It is not wildly different to your IBM METALC example. The code and CLG JCL is attached. It was just a proof of concept, so don't take anything for granted with this. However, it appears to run OK under MVS 3.8j (TK4-).?

You can consider this code released as public domain.

The snippet for WTO ...

int mvsWto(char *msgText, int16_t routeCodes, int16_t desc) { ? ? ??

? ? int32_t ? ? wtoParmList[32]; ? ?// 32 full words ? ? ? ? ? ? ? ?

? ? int ? ? ? ? rc; ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??

? ? rc = intWtoParmList(wtoParmList, msgText, routeCodes, desc); ? ?

? ? if ( rc != 0 ) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ? ? ? return rc; ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ? __asm__("WTO ? MF=(E,(%[wtolist]))" ? ? ? ? ? ? ? ? ? ? ? ? ? ??

? ? ? ? ? ? : /* no outputs */ ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ? ? ? ? ? : /* inputs */ ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ? ? ? ? ? [wtolist] "r"(&wtoParmList) ? ? ? ? ? ? ? ? ? ? ? ? ? ??

? ? ? ? ? ? : /* Registers clobbered */ ? ? ? ? ? ? ? ? ? ? ? ? ? ??

? ? ? ? ? ? "0", "1", "15" ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ? ? ? ? ? ); ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

? ?return 0; ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

} ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??

Regards, Tony.

exwto2.jcl

exwto2.c

#4148

Thank you Tony, appreciated.

Yes the code is quite similar, it looks actually
METALC is modeled after GNUCC, beside other considerations,
GCC should be older.

But, by now, I'm not directly interested in
writing C code under MVS or Z/OS, rather to
understand if IBM had been on the track
(the thread is rather old, more than 15 years)
to switch from PL/S to C for system programming
for its mainframe OS.

It looks, from what I'm reading and tested
with GCC, that using C to access MVS control blocks,
is quite an involved game.

Just as an example this should be the DCB
data structure



translated from HLASM macro assembler in C which
allow to write I/O routines for IBM METALC.

I had quite the same problem with GCC: managing
the DCB fields require such a translation.

The author of this code fragment state that under Z/OS
the EDCDSECT conversion utility may be used to perform
this as an automatic task



Still curious about how this game is played, being METALC, GCC or
even PL/S, if even such compiler would be available, would
make any difference?

Peppe.

On Sun, 14 Jan 2024, Tony W wrote:

Good evening,
A little while back I wrote some C code, for the GCC compiler using embedded
assembler, that made a few calls to MVS macros: WTO, WTOR, STIMER and WAIT. This was
on MVS3.8j TK4-. I suspect the concept would work with new versions of the compiler on
more modern versions of MVS.
It is not wildly different to your IBM METALC example. The code and CLG JCL is
attached. It was just a proof of concept, so don't take anything for granted with
this. However, it appears to run OK under MVS 3.8j (TK4-).
You can consider this code released as public domain.
The snippet for WTO ...
int mvsWto(char *msgText, int16_t routeCodes, int16_t desc) {
int32_t wtoParmList[32]; // 32 full words
int rc;
rc = intWtoParmList(wtoParmList, msgText, routeCodes, desc);
if ( rc != 0 )
return rc;
__asm__("WTO MF=(E,(%[wtolist]))"
: /* no outputs */
: /* inputs */
[wtolist] "r"(&wtoParmList)
: /* Registers clobbered */
"0", "1", "15"
);
return 0;
}
Regards, Tony.
---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4149

Warning, it's definitely not a good example of style but it's GCC, it works in MVS 3.8 TK4- , it reads control blocks.

Some time ago I had some trouble (and had to ask to the group) how to do the same thing in PL/I(F) where there is no pointer arithmetic.

?

Ciao

Marco

ctrlblks.c

#4150

Peppe,

"Still curious about how this game is played, being METALC, GCC or

even PL/S, if even such compiler would be available, would
make any difference?"

PL/S hasnt?been used in decades. The modern version is PL/X-390. :)

Joe

Show quoted text

On Mon, Jan 15, 2024 at 3:13?AM Giuseppe Vitillaro <giuseppe@...> wrote:

Thank you Tony, appreciated.

Yes the code is quite similar, it looks actually
METALC is modeled after GNUCC, beside other considerations,
GCC should be older.

But, by now, I'm not directly interested in
writing C code under MVS or Z/OS, rather to
understand if IBM had been on the track
(the thread is rather old, more than 15 years)
to switch from PL/S to C for system programming
for its mainframe OS.

It looks, from what I'm reading and tested
with GCC, that using C to access MVS control blocks,
is quite an involved game.

Just as an example this should be the DCB
data structure

translated from HLASM macro assembler in C which
allow to write I/O routines for IBM METALC.

I had quite the same problem with GCC: managing
the DCB fields require such a translation.

The author of this code fragment state that under Z/OS
the EDCDSECT conversion utility may be used to perform
this as an automatic task

Still curious about how this game is played, being METALC, GCC or
even PL/S, if even such compiler would be available, would
make any difference?

Peppe.

On Sun, 14 Jan 2024, Tony W wrote:

> Good evening,
>
> A little while back I wrote some C code, for the GCC compiler using embedded
> assembler, that made a few calls to MVS macros: WTO, WTOR, STIMER and WAIT. This was
> on MVS3.8j TK4-. I suspect the concept would work with new versions of the compiler on
> more modern versions of MVS.
>
> It is not wildly different to your IBM METALC example. The code and CLG JCL is
> attached. It was just a proof of concept, so don't take anything for granted with
> this. However, it appears to run OK under MVS 3.8j (TK4-).
>
> You can consider this code released as public domain.
>
> The snippet for WTO ...
>
> int mvsWto(char *msgText, int16_t routeCodes, int16_t desc) {
> int32_t wtoParmList[32]; // 32 full words
> int rc;
> rc = intWtoParmList(wtoParmList, msgText, routeCodes, desc);
> if ( rc != 0 )
> return rc;
> __asm__("WTO MF=(E,(%[wtolist]))"
> : /* no outputs */
> : /* inputs */
> [wtolist] "r"(&wtoParmList)
> : /* Registers clobbered */
> "0", "1", "15"
> );
> return 0;
> }
>
>
> Regards, Tony.
>
>
>
>

---

Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail : giuseppe@...
ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia? Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4151

On Mon, 15 Jan 2024, Joe Monk wrote:

Peppe,
"Still curious about how this game is played, being METALC, GCC or
even PL/S, if even such compiler would be available, would
make any difference?"
PL/S hasnt?been used in decades. The modern version is PL/X-390. :)Apologies, I'm not "updated" on PL languages ;-)
I'm more a C/UNIX guy as you know.

Any way, I would be glad anyway to have access to the Rand's PL/S
compiler under MVS3.8j, if there is a way to use it on
an ancient platform.

For what I've seen, it would be a nice way to gain a better
understanding of this game, although I'm a little bit aged
to become as proficient as I'm in C using PL/I like languages.

But for what I'm reading here

  

this compiler may be defined ... a "chimera"?

Weird IBM in the XXI century is still so jelaous
about this compiler when IBM itself seems moving in the C direction with METALC.

Or I misunderstand and PL/X-390 is still the main road
under Z/OS?

Peppe.

---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4152

Peppe,

I don't have any inside info but talking to a few IBM developers who work on Db2 I got the impression that they are mainly using PL/X. It may or may not be case for other products. You might get some information if you asked on IBM-MAIN.

Khalid

Show quoted text

On 1/15/2024 5:24 AM, Giuseppe Vitillaro wrote:
<snip>
Weird IBM in the XXI century is still so jelaous
about this compiler when IBM itself seems moving in the C direction with METALC.

Or I misunderstand and PL/X-390 is still the main road
under Z/OS?

Peppe.

#4153

I believe if you ask on IBM-MAIN you will find that PL/X is still the main tool for development of z/OS.

Joe

Show quoted text

On Mon, Jan 15, 2024 at 5:24?AM Giuseppe Vitillaro <giuseppe@...> wrote:

On Mon, 15 Jan 2024, Joe Monk wrote:

> Peppe,
> "Still curious about how this game is played, being METALC, GCC or
> even PL/S, if even such compiler would be available, would
> make any difference?"
> PL/S hasnt?been used in decades. The modern version is PL/X-390. :)

Apologies, I'm not "updated" on PL languages ;-)
I'm more a C/UNIX guy as you know.

Any way, I would be glad anyway to have access to the Rand's PL/S
compiler under MVS3.8j, if there is a way to use it on
an ancient platform.

For what I've seen, it would be a nice way to gain a better
understanding of this game, although I'm a little bit aged
to become as proficient as I'm in C using PL/I like languages.

But for what I'm reading here

? ?

this compiler may be defined ... a "chimera"?

Weird IBM in the XXI century is still so jelaous
about this compiler when IBM itself seems
moving in the C direction with METALC.

Or I misunderstand and PL/X-390 is still the main road
under Z/OS?

Peppe.

---

Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail : giuseppe@...
ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia? Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4154

I found this post quite interesting.  I wrote lots of PL/AS code in the 1980's and 90's when I worked for IBM.  I still have a lot of the code but have no way to compile it with TK4-.  I had never heard of Rand's RL/S until today.  Did Rand ever release their RL/S compiler to the public?  It would be fun to see my PL/AS code could be adapted to compile with RL/S.

Charles Bailey

Show quoted text

On 2024-01-15 05:24, Giuseppe Vitillaro wrote:
On Mon, 15 Jan 2024, Joe Monk wrote:

Peppe,
"Still curious about how this game is played, being METALC, GCC or
even PL/S, if even such compiler would be available, would
make any difference?"
PL/S hasnt?been used in decades. The modern version is PL/X-390. :)Apologies, I'm not "updated" on PL languages ;-)
I'm more a C/UNIX guy as you know.

Any way, I would be glad anyway to have access to the Rand's PL/S
compiler under MVS3.8j, if there is a way to use it on
an ancient platform.

For what I've seen, it would be a nice way to gain a better
understanding of this game, although I'm a little bit aged
to become as proficient as I'm in C using PL/I like languages.

But for what I'm reading here

  

this compiler may be defined ... a "chimera"?

Weird IBM in the XXI century is still so jelaous
about this compiler when IBM itself seems moving in the C direction with METALC.

Or I misunderstand and PL/X-390 is still the main road
under Z/OS?

Peppe.

---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia Phone:+39.075.585-5518
----------------------------------------------------------------------------- 

#4155

As I wrote, I'm not a PL/I guy, I began to use
PL/I when I first met MVS3.8j under hercules,
around ten years ago.

So, if even I will need it, I'm able to use
GCC, with its inline assembly feature, to,
I guess, perform almost the same task.

But I've to admit that the Rand PL/S compiler
is an interesting piece of software.

Its history is even more interesting, quoting from
the PL/S Wiki page:

"The Rand RL/S language and compiler were publicly announced in August,
1976, at a Share meeting where numerous copies of the RL/S Language
Reference Manual were distributed to attendees.? IBM quickly responded
claiming the language was proprietary.? While Rand admitted no wrong doing
in developing RL/S, Rand management agreed not to distribute the compiler
publicly.? Rand, however, continued to use the RL/S language and compiler
internally for many projects."

It would be interesting to understand if IBM is still
against distributing this compiler, it should be now
almost half a century old.

Maybe in our days the "Rand management", it looks the
company still exists:

??? https://www.rand.org/about/locations/santa-monica.html

may be available to share, or sell, copies of this compiler
without any issue with IBM?

Peppe.

#4156

Yes, Rand released it. It wasnt?even an IBM product, Rand?reverse engineered the compiler from available docs, and sold it for $100.

IBM made them go retrieve every?single tape and destroy it, and promise to never again distribute it, in exchange for not being sued.

Joe

Show quoted text

On Mon, Jan 15, 2024 at 9:55?PM Charles Bailey <txlogicguy@...> wrote:

I found this post quite interesting.? I wrote lots of PL/AS code in the
1980's and 90's when I worked for IBM.? I still have a lot of the code
but have no way to compile it with TK4-.? I had never heard of Rand's
RL/S until today.? Did Rand ever release their RL/S compiler to the
public?? It would be fun to see my PL/AS code could be adapted to
compile with RL/S.

Charles Bailey

On 2024-01-15 05:24, Giuseppe Vitillaro wrote:
> On Mon, 15 Jan 2024, Joe Monk wrote:
>
>> Peppe,
>> "Still curious about how this game is played, being METALC, GCC or
>> even PL/S, if even such compiler would be available, would
>> make any difference?"
>> PL/S hasnt?been used in decades. The modern version is PL/X-390. :)
>
> Apologies, I'm not "updated" on PL languages ;-)
> I'm more a C/UNIX guy as you know.
>
> Any way, I would be glad anyway to have access to the Rand's PL/S
> compiler under MVS3.8j, if there is a way to use it on
> an ancient platform.
>
> For what I've seen, it would be a nice way to gain a better
> understanding of this game, although I'm a little bit aged
> to become as proficient as I'm in C using PL/I like languages.
>
> But for what I'm reading here
>
>? ?
>
> this compiler may be defined ... a "chimera"?
>
> Weird IBM in the XXI century is still so jelaous
> about this compiler when IBM itself seems moving in the C direction
> with METALC.
>
> Or I misunderstand and PL/X-390 is still the main road
> under Z/OS?
>
> Peppe.
>
> ---
>
> Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail : giuseppe@...
> ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia
> Phone:+39.075.585-5518
> -----------------------------------------------------------------------------
>
>
>
>
>
>
>

#4157

On Tue, 16 Jan 2024, Joe Monk wrote:

Yes, Rand released it. It wasnt?even an IBM product, Rand?reverse engineered
the compiler from available docs, and sold it for $100.
IBM made them go retrieve every?single tape and destroy it, and promise to
never again distribute it, in exchange for not being sued.
JoeYep, understood.

Buf half a century ticked over the clock, isn't it?

Impossible to think may be different in our days,
at the beginning of XXI century?

MVT and MVS are now just toys for hobbists, not
OS for real iron.

This is something I can't really understand, it
looks more like  ... what the word in english?  ... stubborness?
obstinacy? ... please suggest to an italian speaker the
correct translation for "puntiglio" in english ... than
a real legal issue.

Maybe IBM may have changed its mind?

Peppe.

On Mon, Jan 15, 2024 at 9:55?PM Charles Bailey <txlogicguy@...> wrote:
      I found this post quite interesting.? I wrote lots of PL/AS code
      in the
      1980's and 90's when I worked for IBM.? I still have a lot of
      the code
      but have no way to compile it with TK4-.? I had never heard of
      Rand's
      RL/S until today.? Did Rand ever release their RL/S compiler to
      the
      public?? It would be fun to see my PL/AS code could be adapted
      to
      compile with RL/S.

      Charles Bailey

      On 2024-01-15 05:24, Giuseppe Vitillaro wrote:
      > On Mon, 15 Jan 2024, Joe Monk wrote:
      >
      >> Peppe,
      >> "Still curious about how this game is played, being METALC,
      GCC or
      >> even PL/S, if even such compiler would be available, would
      >> make any difference?"
      >> PL/S hasnt?been used in decades. The modern version is
      PL/X-390. :)
      >
      > Apologies, I'm not "updated" on PL languages ;-)
      > I'm more a C/UNIX guy as you know.
      >
      > Any way, I would be glad anyway to have access to the Rand's
      PL/S
      > compiler under MVS3.8j, if there is a way to use it on
      > an ancient platform.
      >
      > For what I've seen, it would be a nice way to gain a better
      > understanding of this game, although I'm a little bit aged
      > to become as proficient as I'm in C using PL/I like languages.
      >
      > But for what I'm reading here
      >
      >? ?
      >
      > this compiler may be defined ... a "chimera"?
      >
      > Weird IBM in the XXI century is still so jelaous
      > about this compiler when IBM itself seems moving in the C
      direction
      > with METALC.
      >
      > Or I misunderstand and PL/X-390 is still the main road
      > under Z/OS?
      >
      > Peppe.
      >
      > ---
      >
      > Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail :
      giuseppe@...
      > ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia
      > Phone:+39.075.585-5518
      >---------------------------------------------------------------------------
      --
      >
      >
      >
      >
      >
      >
      >
---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4158

Giuseppe Vitillaro wrote:

[...]
This is something I can't really understand,
it looks more like  ... what the word in english?
... stubborness? obstinacy? ... please suggest
to an italian speaker the correct translation
for "puntiglio" in english ...pique?




-- 
"Fish" (David B. Trout)
Software Development Laboratories

mail: fish@...

#4159

Show quoted text

-----Original Message-----
From: [email protected] <[email protected]> On Behalf Of Giuseppe
Vitillaro
Sent: Tuesday, January 16, 2024 6:28 PM
To: [email protected]
Subject: Re: [H390-MVS] IBM METALC curiosity.

On Tue, 16 Jan 2024, Joe Monk wrote:

Yes, Rand released it. It wasnt even an IBM product, Rand reverse
engineered the compiler from available docs, and sold it for $100.
IBM made them go retrieve every single tape and destroy it, and
promise to never again distribute it, in exchange for not being sued.
JoeYep, understood.

Buf half a century ticked over the clock, isn't it?

Impossible to think may be different in our days, at the beginning of XXI century?

MVT and MVS are now just toys for hobbists, not OS for real iron.

This is something I can't really understand, it looks more like  ... what the word
in english?  ... stubborness?
obstinacy? ... please suggest to an italian speaker the correct translation for
"puntiglio" in english ... than a real legal issue.

Maybe IBM may have changed its mind?If anything, IBM is more protective about its intellectual property these days.
From previous comments I see its still used.



Peppe.
Dave



On Mon, Jan 15, 2024 at 9:55?PM Charles Bailey <txlogicguy@...> wrote:
      I found this post quite interesting.  I wrote lots of PL/AS code
      in the
      1980's and 90's when I worked for IBM.  I still have a lot of
      the code
      but have no way to compile it with TK4-.  I had never heard of
      Rand's
      RL/S until today.  Did Rand ever release their RL/S compiler to
      the
      public?  It would be fun to see my PL/AS code could be adapted
      to
      compile with RL/S.

      Charles Bailey

      On 2024-01-15 05:24, Giuseppe Vitillaro wrote:
      > On Mon, 15 Jan 2024, Joe Monk wrote:
      >
      >> Peppe,
      >> "Still curious about how this game is played, being METALC,
      GCC or
      >> even PL/S, if even such compiler would be available, would
      >> make any difference?"
      >> PL/S hasnt?been used in decades. The modern version is
      PL/X-390. :)
      >
      > Apologies, I'm not "updated" on PL languages ;-)
      > I'm more a C/UNIX guy as you know.
      >
      > Any way, I would be glad anyway to have access to the Rand's
      PL/S
      > compiler under MVS3.8j, if there is a way to use it on
      > an ancient platform.
      >
      > For what I've seen, it would be a nice way to gain a better
      > understanding of this game, although I'm a little bit aged
      > to become as proficient as I'm in C using PL/I like languages.
      >
      > But for what I'm reading here
      >
      >   
      >
      > this compiler may be defined ... a "chimera"?
      >
      > Weird IBM in the XXI century is still so jelaous
      > about this compiler when IBM itself seems moving in the C
      direction
      > with METALC.
      >
      > Or I misunderstand and PL/X-390 is still the main road
      > under Z/OS?
      >
      > Peppe.
      >
      > ---
      >
      > Giuseppe Vitillaro                  |  E-Mail :
      giuseppe@...
      > ex-IBM, ex-CNR                      |  06123 Perugia
      > Phone:+39.075.585-5518
      >---------------------------------------------------------------------------
      --
      >
      >
      >
      >
      >
      >
      >








---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4160

>>IBM made them go retrieve every?single tape and destroy it, and promise to
>>never again distribute it, in exchange for not being sued.

>This is something I can't really understand, it
>looks more like ... what the word in english? ... stubborness?
>obstinacy? ... than a real legal issue.

>Maybe IBM may have changed its mind?

I doubt if IBM could have sued RAND successfully. Back then, the arguments were about look and feel, and trying to patent a programming language would fail as it would be deemed a mere mathematical formula. Now, good-cop-bad-cop, IBM could have suggested that an operation busy modifying MVT might want to keep in good graces for the times it needed to call for support or such. IBM has good reasons not to have PL/S or a workalike out in the wild.

It looks like someone could start with PL360 and the available documentation and if it took three programmers for RAND to bring RL/S to production, how long would it take one programmer?

As for as Metal C as it applies to MVS, a GCC library could invoke the SVCs seamlessly, if someone wants to do the work and if anyone would use it.

#4161

On Tue, 16 Jan 2024, Fish Fish wrote:

Giuseppe Vitillaro wrote:

[...]
This is something I can't really understand,
it looks more like  ... what the word in english?
... stubborness? obstinacy? ... please suggest
to an italian speaker the correct translation
for "puntiglio" in english ...pique?


Thank you, but not really, from what I can read the nearest
italian word to "pique" seems "dispetto".

The word "puntglio" is when you insist on a
point just because you are not likely to admit
you were wrong, even against your own interest.

A "dispetto" is different, it may be done
just for the pleasure of doing it.

Peppe.

---

Giuseppe Vitillaro                  |  E-Mail : giuseppe@...
ex-IBM, ex-CNR                      |  06123 Perugia  Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4162

"MVT and MVS are now just toys for hobbists, not
OS for real iron."

You know how IBM insists on maintaining forward compatibility, even in z/OS?

Like, how you can take an object module from OS/360, re-link on z/OS, and execute?

Joe

Show quoted text

On Tue, Jan 16, 2024 at 12:27?PM Giuseppe Vitillaro <giuseppe@...> wrote:

On Tue, 16 Jan 2024, Joe Monk wrote:

> Yes, Rand released it. It wasnt?even an IBM product, Rand?reverse engineered
> the compiler from available docs, and sold it for $100.
> IBM made them go retrieve every?single tape and destroy it, and promise to
> never again distribute it, in exchange for not being sued.
> Joe

Yep, understood.

Buf half a century ticked over the clock, isn't it?

Impossible to think may be different in our days,
at the beginning of XXI century?

MVT and MVS are now just toys for hobbists, not
OS for real iron.

This is something I can't really understand, it
looks more like? ... what the word in english?? ... stubborness?
obstinacy? ... please suggest to an italian speaker the
correct translation for "puntiglio" in english ... than
a real legal issue.

Maybe IBM may have changed its mind?

Peppe.

>
> On Mon, Jan 15, 2024 at 9:55?PM Charles Bailey <txlogicguy@...> wrote:
>? ? ? ?I found this post quite interesting.? I wrote lots of PL/AS code
>? ? ? ?in the
>? ? ? ?1980's and 90's when I worked for IBM.? I still have a lot of
>? ? ? ?the code
>? ? ? ?but have no way to compile it with TK4-.? I had never heard of
>? ? ? ?Rand's
>? ? ? ?RL/S until today.? Did Rand ever release their RL/S compiler to
>? ? ? ?the
>? ? ? ?public?? It would be fun to see my PL/AS code could be adapted
>? ? ? ?to
>? ? ? ?compile with RL/S.
>
>? ? ? ?Charles Bailey
>
>? ? ? ?On 2024-01-15 05:24, Giuseppe Vitillaro wrote:
>? ? ? ?> On Mon, 15 Jan 2024, Joe Monk wrote:
>? ? ? ?>
>? ? ? ?>> Peppe,
>? ? ? ?>> "Still curious about how this game is played, being METALC,
>? ? ? ?GCC or
>? ? ? ?>> even PL/S, if even such compiler would be available, would
>? ? ? ?>> make any difference?"
>? ? ? ?>> PL/S hasnt?been used in decades. The modern version is
>? ? ? ?PL/X-390. :)
>? ? ? ?>
>? ? ? ?> Apologies, I'm not "updated" on PL languages ;-)
>? ? ? ?> I'm more a C/UNIX guy as you know.
>? ? ? ?>
>? ? ? ?> Any way, I would be glad anyway to have access to the Rand's
>? ? ? ?PL/S
>? ? ? ?> compiler under MVS3.8j, if there is a way to use it on
>? ? ? ?> an ancient platform.
>? ? ? ?>
>? ? ? ?> For what I've seen, it would be a nice way to gain a better
>? ? ? ?> understanding of this game, although I'm a little bit aged
>? ? ? ?> to become as proficient as I'm in C using PL/I like languages.
>? ? ? ?>
>? ? ? ?> But for what I'm reading here
>? ? ? ?>
>? ? ? ?>? ?
>? ? ? ?>
>? ? ? ?> this compiler may be defined ... a "chimera"?
>? ? ? ?>
>? ? ? ?> Weird IBM in the XXI century is still so jelaous
>? ? ? ?> about this compiler when IBM itself seems moving in the C
>? ? ? ?direction
>? ? ? ?> with METALC.
>? ? ? ?>
>? ? ? ?> Or I misunderstand and PL/X-390 is still the main road
>? ? ? ?> under Z/OS?
>? ? ? ?>
>? ? ? ?> Peppe.
>? ? ? ?>
>? ? ? ?> ---
>? ? ? ?>
>? ? ? ?> Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail :
>? ? ? ?giuseppe@...
>? ? ? ?> ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia
>? ? ? ?> Phone:+39.075.585-5518
>? ? ? ?>---------------------------------------------------------------------------
>? ? ? ?--
>? ? ? ?>
>? ? ? ?>
>? ? ? ?>
>? ? ? ?>
>? ? ? ?>
>? ? ? ?>
>? ? ? ?>
>
>
>
>
>
>
>
>
>

---

Giuseppe Vitillaro? ? ? ? ? ? ? ? ? |? E-Mail : giuseppe@...
ex-IBM, ex-CNR? ? ? ? ? ? ? ? ? ? ? |? 06123 Perugia? Phone:+39.075.585-5518
-----------------------------------------------------------------------------

#4163

On Tue, 16 Jan 2024, Joe Monk wrote:

"MVT and MVS are now just toys for hobbists, not
OS for real iron."
You know how IBM insists on maintaining forward compatibility, even in z/OS?
Like, how you can take an object module from OS/360, re-link on z/OS, and
execute?
Joe

Who asked for that?

If I even would find a PL/S compiler I would
use it under MVS3.8j and only under MVS3.8j.

Even if created for MVT it should run
under MVS3.8j beside OS dependencies
I wouldn't expect from a compiler.

Why I should need to relink it on Z/OS?

Peppe.

#4164

On Tue, 16 Jan 2024, Laird Heal via??wrote:

>>IBM made them go retrieve every?single tape and destroy it, and promise to
>>never again distribute it, in exchange for not being sued.
>This is something I can't really understand, it
>looks more like ... what the word in english? ... stubborness?
>obstinacy? ... than a real legal issue.
>Maybe IBM may have changed its mind?
I doubt if IBM could have sued RAND successfully. Back then, the arguments
were about look and feel, and trying to patent a programming language would
fail as it would be deemed a mere mathematical formula. Now,
good-cop-bad-cop, IBM could have suggested that an operation busy modifying
MVT might want to keep in good graces for the times it needed to call for
support or such. IBM has good reasons not to have PL/S or a workalike out in
the wild.

This match with what I'm reading on Wiki, it
looks like the Rand Company just didn't find any real
interest in being sued by IBM by such an
economical irrilevant thing.

Reading between the lines it seems that Rand Company
itself didn't dismiss their compiler internally, they
just didn't sell or distribute it anymore after IBM
expressed its concerns.

It looks like someone could start with PL360 and the available documentation
and if it took three programmers for RAND to bring RL/S to production, how
long would it take one programmer?
As for as Metal C as it applies to MVS, a GCC library could invoke the SVCs
seamlessly, if someone wants to do the work and if anyone would use it.

Yes, the asm inline GCC feature can do that,
on any reasonable platform from what I know.

The GCC compiler just emit assembly code
and may add assembler language source code whatever
is needed to C code, C code that at the end
born as a portable "assmbler" language ;-)

Actually, as I wrote, IBM METALC seems modeled,
even for its syntax, on GCC.

But, just for the sake of the history, it looks like
PL/S had been probably in its age (1960-1970)
among the first between languages which support
this feature.

True of false?

Peppe.

#4165

In the '60s, there was BSL (basic systems language). It is the predecessor of PL/S.

Joe

Show quoted text

On Tue, Jan 16, 2024 at 3:31?PM Giuseppe Vitillaro <giuseppe@...> wrote:

On Tue, 16 Jan 2024, Laird Heal via??wrote:

>>IBM made them go retrieve every?single tape and destroy it, and promise to
>>never again distribute it, in exchange for not being sued.
>This is something I can't really understand, it
>looks more like ... what the word in english? ... stubborness?
>obstinacy? ... than a real legal issue.
>Maybe IBM may have changed its mind?
I doubt if IBM could have sued RAND successfully. Back then, the arguments
were about look and feel, and trying to patent a programming language would
fail as it would be deemed a mere mathematical formula. Now,
good-cop-bad-cop, IBM could have suggested that an operation busy modifying
MVT might want to keep in good graces for the times it needed to call for
support or such. IBM has good reasons not to have PL/S or a workalike out in
the wild.

This match with what I'm reading on Wiki, it
looks like the Rand Company just didn't find any real
interest in being sued by IBM by such an
economical irrilevant thing.

Reading between the lines it seems that Rand Company
itself didn't dismiss their compiler internally, they
just didn't sell or distribute it anymore after IBM
expressed its concerns.

It looks like someone could start with PL360 and the available documentation
and if it took three programmers for RAND to bring RL/S to production, how
long would it take one programmer?
As for as Metal C as it applies to MVS, a GCC library could invoke the SVCs
seamlessly, if someone wants to do the work and if anyone would use it.

Yes, the asm inline GCC feature can do that,
on any reasonable platform from what I know.

The GCC compiler just emit assembly code
and may add assembler language source code whatever
is needed to C code, C code that at the end
born as a portable "assmbler" language ;-)

Actually, as I wrote, IBM METALC seems modeled,
even for its syntax, on GCC.

But, just for the sake of the history, it looks like
PL/S had been probably in its age (1960-1970)
among the first between languages which support
this feature.

True of false?

Peppe.

开云体育

IBM METALC curiosity.