Discussion:
An interesting idea
(too old to reply)
Gregg C Levine
2002-08-24 01:49:13 UTC
Permalink
Hello again from Gregg C Levine
Has anyone, besides me, given this one some thought? Setup Hercules to
run Linux for the S/390, run the whole shebang on Linux itself, I386
series Linux, then with working TUN/TAP connectivity have the box itself
connected to a local area network, which is reachable from the Internet.
All of that would work, only if the network is reachable on the instance
of Linux for the S/390. Run on the image of Linux for the S/390 the
usual functions that normal Linux does. The proof of this would be
having someone, me, even, send the Hercules, and Linux for the S/390,
e-mail. If it works, then the installation of say, an appropriate list
server software package would then follow. We'd need to find out what
list server programs the folks at Marist are using to host their
collection of lists first, before proceeding. Assuming that I'm not
breathing vapor ware, we'd have an almost viable alternative to this
clumsy method of managing our list, that Yahoo has. I'm not saying that
we should move this list onto that setup, more like a simple discussion
list, anything except sports, politics, and sex would be allowed. Oh,
and this would involve two big questions, not including hardware.
Question One, which Intel based distribution gets chosen? And Question
Two, which S/390 based distribution gets chosen?
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )
Per Jessen
2002-08-25 09:08:44 UTC
Permalink
On Fri, 23 Aug 2002 21:49:13 -0400, Gregg C Levine wrote:

>Hello again from Gregg C Levine
[proposal for running a mailing-list]

Hi Gregg,

if you'll allow me to sum up - you are proposing we install
a server and a permanent internet connection, then run majordomo,
mailman or something similar, then let it run the hercules-390 list.
I think your idea is excellent, even if somewhat over-engineered ;-)

I proposed to use http://sunsite.dk a while ago - they offered to host
the list including archiving - all free. Just because we're discussing
Hercules, it doesn't necessarily mean that the mailserver would be
best off by running on a Hercules based OS.

Regarding sunsite.dk hosting, I didn't want to step on anyones toes,
and didn't press the issue any further - I can't remember a particularly
overwhelmingly positive response :-(

/per




regards,
Per Jessen, Zurich
http://www.enidan.com - home of the J1 serial console.

Genius may have its limitations, but stupidity is not thus handicapped.
Paul Raulerson
2002-08-25 16:26:39 UTC
Permalink
Well, there is another point as well - L/390 running under Hercules under Linux is, well, slow. It is becoming pretty obvious that
Linux is not as efficient, by several orders of magnitude at least, as MVS, VM, or OS/390. A totally un-scientific but perhaps
indicative measure is the MIPS measurement displayed by
Hercules on the Hercules console. Here are some of the average readings, all take from the same machine
with the same hardware:

MVS --- Average during load: 19+ mips
Average during runtime 12+mips

VM/370 -- Average during load 20+mips
Average during run 18+ mips

Linux (SuSE 7.0) Average during load 9+ mips
Average during run 6+Mips

Linux (Redhat 7.2) Average during load 10+mips
Average during run 6+mips

Linux (ThinkBlue 64bit) Average during load 7+mips
Average during run 4+mips

Linux (Debian 3.0) Average during load 12+mips
Average during run 9+ mips

----
These are spot averages taken during a compiler run (COBOL on the IBM OS's, C under Linux)
the run job (calculate the interest and dollar amount of a investment over 100 years, 100 investments)
and a report job (list the input dataset) 1000 times. I modified the MIPS reporting code to store the
values (updated once every 500 milliseconds) to a data file for analysis.

Now I am wondering what could be at the core of this - surely part of it is the intense optimization that
MVS and VM have gone through, but just as surely, part of it has to be that Linux isn't doing so well.

A BIG part of it has to be Linux DASD; Linux seems to access and use an emulated 3390 volume about 100 -120 time slower than an IBM
O/S does.

Given all that, Linux/390 under Hercules under Linux/Intel is great for validating a product for Linux/390; but is way to slow to
develop under. Remember, a typical Intel box these days delivers a few hundred MIPS of CISC performance, and a PowerPC running at
330mhz delivers almost 900mips (peak) of RISC performance. While we can and do expect to see only 1/100th of that in emulation,
still - if MVS can 20+ mips out of some hardware, Linux should be able to as well. 20 mips would be just about on the verge of
usability, if I guess right.

Oh, the hardware:

1.2ghx PIII, 512 RAM, 80gig IDE drive with a hdparm of ~40/mb second.
Linux/Intel is RH7.2.

-Paul



----- Original Message -----
From: "Per Jessen" <per-***@public.gmane.org>
To: "Gregg C Levine" <hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org>; "Hercules-390" <hercules-***@yahoogroups.com>
Sent: Sunday, August 25, 2002 4:08 AM
Subject: Re: [hercules-390] An interesting idea


> On Fri, 23 Aug 2002 21:49:13 -0400, Gregg C Levine wrote:
>
> >Hello again from Gregg C Levine
> [proposal for running a mailing-list]
>
> Hi Gregg,
>
> if you'll allow me to sum up - you are proposing we install
> a server and a permanent internet connection, then run majordomo,
> mailman or something similar, then let it run the hercules-390 list.
> I think your idea is excellent, even if somewhat over-engineered ;-)
>
> I proposed to use http://sunsite.dk a while ago - they offered to host
> the list including archiving - all free. Just because we're discussing
> Hercules, it doesn't necessarily mean that the mailserver would be
> best off by running on a Hercules based OS.
>
> Regarding sunsite.dk hosting, I didn't want to step on anyones toes,
> and didn't press the issue any further - I can't remember a particularly
> overwhelmingly positive response :-(
>
> /per
>
>
>
>
> regards,
> Per Jessen, Zurich
> http://www.enidan.com - home of the J1 serial console.
>
> Genius may have its limitations, but stupidity is not thus handicapped.
>
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Gregg C Levine
2002-08-25 16:35:23 UTC
Permalink
Hello from Gregg C Levine
Thank you Paul. The position I have placed myself in, here, now, is that
of a theoretical scientist. Creating, and then proposing a theory. What
I am simply proposing here, is simply a test for us, anyway, of the
whole notion of running L/390 under Hercules. Also, the classic funny
question, what prompted you to choose that version number of Red Hat?
Also, I wasn't surprised by your numbers, only caught off guard.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> Sent: Sunday, August 25, 2002 12:27 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] An interesting idea
>
> Well, there is another point as well - L/390 running under Hercules
under Linux is,
> well, slow. It is becoming pretty obvious that
> Linux is not as efficient, by several orders of magnitude at least, as
MVS, VM, or
> OS/390. A totally un-scientific but perhaps
> indicative measure is the MIPS measurement displayed by
> Hercules on the Hercules console. Here are some of the average
readings, all take
> from the same machine
> with the same hardware:
>
> MVS --- Average during load: 19+ mips
> Average during runtime 12+mips
>
> VM/370 -- Average during load 20+mips
> Average during run 18+ mips
>
> Linux (SuSE 7.0) Average during load 9+ mips
> Average during run 6+Mips
>
> Linux (Redhat 7.2) Average during load 10+mips
> Average during run 6+mips
>
> Linux (ThinkBlue 64bit) Average during load 7+mips
> Average during run 4+mips
>
> Linux (Debian 3.0) Average during load 12+mips
> Average during run 9+ mips
>
> ----
> These are spot averages taken during a compiler run (COBOL on the IBM
OS's, C
> under Linux)
> the run job (calculate the interest and dollar amount of a
investment over 100 years,
> 100 investments)
> and a report job (list the input dataset) 1000 times. I modified the
MIPS reporting
> code to store the
> values (updated once every 500 milliseconds) to a data file for
analysis.
>
> Now I am wondering what could be at the core of this - surely part of
it is the
> intense optimization that
> MVS and VM have gone through, but just as surely, part of it has to be
that Linux
> isn't doing so well.
>
> A BIG part of it has to be Linux DASD; Linux seems to access and use
an emulated
> 3390 volume about 100 -120 time slower than an IBM
> O/S does.
>
> Given all that, Linux/390 under Hercules under Linux/Intel is great
for validating a
> product for Linux/390; but is way to slow to
> develop under. Remember, a typical Intel box these days delivers a few
hundred
> MIPS of CISC performance, and a PowerPC running at
> 330mhz delivers almost 900mips (peak) of RISC performance. While we
can and do
> expect to see only 1/100th of that in emulation,
> still - if MVS can 20+ mips out of some hardware, Linux should be able
to as well.
> 20 mips would be just about on the verge of
> usability, if I guess right.
>
> Oh, the hardware:
>
> 1.2ghx PIII, 512 RAM, 80gig IDE drive with a hdparm of ~40/mb second.
> Linux/Intel is RH7.2.
>
> -Paul
>
>
>
> ----- Original Message -----
> From: "Per Jessen" <per-***@public.gmane.org>
> To: "Gregg C Levine" <hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org>; "Hercules-390"
> <hercules-390-***@public.gmane.org>
> Sent: Sunday, August 25, 2002 4:08 AM
> Subject: Re: [hercules-390] An interesting idea
>
>
> > On Fri, 23 Aug 2002 21:49:13 -0400, Gregg C Levine wrote:
> >
> > >Hello again from Gregg C Levine
> > [proposal for running a mailing-list]
> >
> > Hi Gregg,
> >
> > if you'll allow me to sum up - you are proposing we install
> > a server and a permanent internet connection, then run majordomo,
> > mailman or something similar, then let it run the hercules-390 list.
> > I think your idea is excellent, even if somewhat over-engineered ;-)
> >
> > I proposed to use http://sunsite.dk a while ago - they offered to
host
> > the list including archiving - all free. Just because we're
discussing
> > Hercules, it doesn't necessarily mean that the mailserver would be
> > best off by running on a Hercules based OS.
> >
> > Regarding sunsite.dk hosting, I didn't want to step on anyones toes,
> > and didn't press the issue any further - I can't remember a
particularly
> > overwhelmingly positive response :-(
> >
> > /per
> >
> >
> >
> >
> > regards,
> > Per Jessen, Zurich
> > http://www.enidan.com - home of the J1 serial console.
> >
> > Genius may have its limitations, but stupidity is not thus
handicapped.
> >
> >
> >
> >
( Everything left intact at the request of the Time Lord. )
Rich Smrcina
2002-08-25 17:02:18 UTC
Permalink
Paul,

When you measured Linux, were you running with the Hercules console or with
the status display (with the MIPS counter and I/O rate, et. al.)? I've
noiticed a significant slow down of Linux with the console running and
displaying all of the interrupt messages. I can not quantify it, but it just
'seemed' to be faster.

On Sunday 25 August 2002 11:26 am, you wrote:
> Well, there is another point as well - L/390 running under Hercules under
> Linux is, well, slow. It is becoming pretty obvious that Linux is not as
> efficient, by several orders of magnitude at least, as MVS, VM, or OS/390.
> A totally un-scientific but perhaps indicative measure is the MIPS
> measurement displayed by
> Hercules on the Hercules console. Here are some of the average readings,
> all take from the same machine with the same hardware:
>
> MVS --- Average during load: 19+ mips
> Average during runtime 12+mips
>
> VM/370 -- Average during load 20+mips
> Average during run 18+ mips
>
> Linux (SuSE 7.0) Average during load 9+ mips
> Average during run 6+Mips
>
> Linux (Redhat 7.2) Average during load 10+mips
> Average during run 6+mips
>
> Linux (ThinkBlue 64bit) Average during load 7+mips
> Average during run 4+mips
>
> Linux (Debian 3.0) Average during load 12+mips
> Average during run 9+ mips
>
> ----
> These are spot averages taken during a compiler run (COBOL on the IBM OS's,
> C under Linux) the run job (calculate the interest and dollar amount of a
> investment over 100 years, 100 investments) and a report job (list the
> input dataset) 1000 times. I modified the MIPS reporting code to store the
> values (updated once every 500 milliseconds) to a data file for analysis.
>
> Now I am wondering what could be at the core of this - surely part of it is
> the intense optimization that MVS and VM have gone through, but just as
> surely, part of it has to be that Linux isn't doing so well.
>
> A BIG part of it has to be Linux DASD; Linux seems to access and use an
> emulated 3390 volume about 100 -120 time slower than an IBM O/S does.
>
> Given all that, Linux/390 under Hercules under Linux/Intel is great for
> validating a product for Linux/390; but is way to slow to develop under.
> Remember, a typical Intel box these days delivers a few hundred MIPS of
> CISC performance, and a PowerPC running at 330mhz delivers almost 900mips
> (peak) of RISC performance. While we can and do expect to see only 1/100th
> of that in emulation, still - if MVS can 20+ mips out of some hardware,
> Linux should be able to as well. 20 mips would be just about on the verge
> of usability, if I guess right.
>
> Oh, the hardware:
>
> 1.2ghx PIII, 512 RAM, 80gig IDE drive with a hdparm of ~40/mb second.
> Linux/Intel is RH7.2.
>
> -Paul
>
>
>
> ----- Original Message -----
> From: "Per Jessen" <per-***@public.gmane.org>
> To: "Gregg C Levine" <hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org>; "Hercules-390"
> <hercules-390-***@public.gmane.org> Sent: Sunday, August 25, 2002 4:08 AM
> Subject: Re: [hercules-390] An interesting idea
>
> > On Fri, 23 Aug 2002 21:49:13 -0400, Gregg C Levine wrote:
> > >Hello again from Gregg C Levine
> >
> > [proposal for running a mailing-list]
> >
> > Hi Gregg,
> >
> > if you'll allow me to sum up - you are proposing we install
> > a server and a permanent internet connection, then run majordomo,
> > mailman or something similar, then let it run the hercules-390 list.
> > I think your idea is excellent, even if somewhat over-engineered ;-)
> >
> > I proposed to use http://sunsite.dk a while ago - they offered to host
> > the list including archiving - all free. Just because we're discussing
> > Hercules, it doesn't necessarily mean that the mailserver would be
> > best off by running on a Hercules based OS.
> >
> > Regarding sunsite.dk hosting, I didn't want to step on anyones toes,
> > and didn't press the issue any further - I can't remember a particularly
> > overwhelmingly positive response :-(
> >
> > /per
> >
> >
> >
> >
> > regards,
> > Per Jessen, Zurich
> > http://www.enidan.com - home of the J1 serial console.
> >
> > Genius may have its limitations, but stupidity is not thus handicapped.
> >
> >
> >
> >
> > Community email addresses:
> > Post message: hercules-390-***@public.gmane.org
> > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > List owner: hercules-390-owner-***@public.gmane.org
> >
> > Files and archives at:
> > http://groups.yahoo.com/group/hercules-390
> >
> > Get the latest version of Hercules from:
> > http://www.conmicro.cx/hercules
> >
> > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

--
Rich Smrcina
Sytek Services, Inc.
Milwaukee, WI
rsmrcina-***@public.gmane.org
rsmrcina-***@public.gmane.org

Catch the WAVV! Stay for Requirements and the Free for All!
Update your S/390 skills in 4 days for a very reasonable price.
WAVV 2003 in Winston-Salem, NC.
April 25-29, 2003
For details see http://www.wavv.org
Greg Smith
2002-08-25 17:07:53 UTC
Permalink
Paul Raulerson wrote:

> A BIG part of it has to be Linux DASD; Linux seems to
> access and use an emulated 3390 volume about 100 -120
> time slower than an IBM O/S does.

This is a somewhat temporary situation. A bug in the
IBM linux CKD device driver was discovered last Feb;
the bug rarely shows up on a real machine (if at all)
but does on Hercules. IBM has fixed this in the April
code drop, and the fix should eventually make it into
the various distributions.

To workaround the bug, we added an i/o delay which is
most likely the cause of your observations. After you
have booted s/390 linux, you may be able to successfully
turn the delay off by issuing the `iodelay 0' hercules
console command.

One other note. Running a 64-bit operating system under
hercules on an ia32 (or any 32-bit) host will increase the
emulation instruction path and also use the host's registers
less efficiently.

Greg
w***@public.gmane.org
2002-08-26 14:36:54 UTC
Permalink
I never knew an S/390 ran a PowerPC 330 MHz, still why so expensive

Anyway, has anyone thought of a take over machine? That is, you boot it
up in like dos; it runs in protected mode and takes over the machine?

Though, I know it would be tricky, giving that it's hard to find a DOS
TCP stack. But then you would have total control of the system.

Or would it really give you that much of a speed improvement over
running inside of windows or Linux?
Gregg C Levine
2002-08-26 15:16:23 UTC
Permalink
Hello from Gregg C Levine
Not quite. The newer AS/400 based systems use a PowerPC processor.
S/390s use their own devices. As for a DOS TCP/IP stack. It exists.
E-mail me off line for details, if you want it.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: warlockd-+***@public.gmane.org [mailto:***@lords.com]
> Sent: Monday, August 26, 2002 10:37 AM
> To: hercules-390-***@public.gmane.org
> Subject: RE: [hercules-390] An interesting idea another one?
>
> I never knew an S/390 ran a PowerPC 330 MHz, still why so expensive
>
> Anyway, has anyone thought of a take over machine? That is, you boot
it
> up in like dos; it runs in protected mode and takes over the machine?
>
> Though, I know it would be tricky, giving that it's hard to find a DOS
> TCP stack. But then you would have total control of the system.
>
> Or would it really give you that much of a speed improvement over
> running inside of windows or Linux?
>
>
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
Paul Raulerson
2002-08-27 00:02:33 UTC
Permalink
Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes, AS/400, and RS/6000 machines)
all use the same PowerPC processor. Only the RS/6000 uses it natively though; the 400 and mainframes have software / microcode
layers that emulate a different processor. The S/390 Architecture is implemented primarily via Microcode these days.

-Paul

----- Original Message -----
From: "Gregg C Levine" <hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Monday, August 26, 2002 10:16 AM
Subject: RE: [hercules-390] An interesting idea another one?


Hello from Gregg C Levine
Not quite. The newer AS/400 based systems use a PowerPC processor.
S/390s use their own devices. As for a DOS TCP/IP stack. It exists.
E-mail me off line for details, if you want it.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke." Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: warlockd-+***@public.gmane.org [mailto:***@lords.com]
> Sent: Monday, August 26, 2002 10:37 AM
> To: hercules-390-***@public.gmane.org
> Subject: RE: [hercules-390] An interesting idea another one?
>
> I never knew an S/390 ran a PowerPC 330 MHz, still why so expensive
>
> Anyway, has anyone thought of a take over machine? That is, you boot
it
> up in like dos; it runs in protected mode and takes over the machine?
>
> Though, I know it would be tricky, giving that it's hard to find a DOS
> TCP stack. But then you would have total control of the system.
>
> Or would it really give you that much of a speed improvement over
> running inside of windows or Linux?
>
>
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>




Community email addresses:
Post message: hercules-390-***@public.gmane.org
Subscribe: hercules-390-subscribe-***@public.gmane.org
Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
List owner: hercules-390-owner-***@public.gmane.org

Files and archives at:
http://groups.yahoo.com/group/hercules-390

Get the latest version of Hercules from:
http://www.conmicro.cx/hercules

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Gregg C Levine
2002-08-27 01:13:30 UTC
Permalink
Hello again from Gregg C Levine
Exactly my point Paul. But he asked, or mentioned those dratted AS/400
machines. I know that the RS/6000 descendants all use the PowerPC
processor natively. I wasn't aware though, that the AS/400 crowd, is
wearing microcode for that part of the solution.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> Sent: Monday, August 26, 2002 8:03 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] An interesting idea another one?
>
> Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes,
AS/400, and
> RS/6000 machines)
> all use the same PowerPC processor. Only the RS/6000 uses it natively
though; the
> 400 and mainframes have software / microcode
> layers that emulate a different processor. The S/390 Architecture is
implemented
> primarily via Microcode these days.
>
> -Paul
>
> ----- Original Message -----
> From: "Gregg C Levine" <hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org>
> To: <hercules-390-***@public.gmane.org>
> Sent: Monday, August 26, 2002 10:16 AM
> Subject: RE: [hercules-390] An interesting idea another one?
>
>
> Hello from Gregg C Levine
> Not quite. The newer AS/400 based systems use a PowerPC processor.
> S/390s use their own devices. As for a DOS TCP/IP stack. It exists.
> E-mail me off line for details, if you want it.
> -------------------
> Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
> ------------------------------------------------------------
> "The Force will be with you...Always." Obi-Wan Kenobi
> "Use the Force, Luke." Obi-Wan Kenobi
> (This company dedicates this E-Mail to General Obi-Wan Kenobi )
> (This company dedicates this E-Mail to Master Yoda )
>
>
>
> > -----Original Message-----
> > From: warlockd-+***@public.gmane.org [mailto:***@lords.com]
> > Sent: Monday, August 26, 2002 10:37 AM
> > To: hercules-390-***@public.gmane.org
> > Subject: RE: [hercules-390] An interesting idea another one?
> >
> > I never knew an S/390 ran a PowerPC 330 MHz, still why so expensive
> >
> > Anyway, has anyone thought of a take over machine? That is, you
boot
> it
> > up in like dos; it runs in protected mode and takes over the
machine?
> >
> > Though, I know it would be tricky, giving that it's hard to find a
DOS
> > TCP stack. But then you would have total control of the system.
> >
> > Or would it really give you that much of a speed improvement over
> > running inside of windows or Linux?
> >
> >
> >
> >
> > ------------------------ Yahoo! Groups Sponsor
> >
> > Community email addresses:
> > Post message: hercules-390-***@public.gmane.org
> > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > List owner: hercules-390-owner-***@public.gmane.org
> >
> > Files and archives at:
> > http://groups.yahoo.com/group/hercules-390
> >
> > Get the latest version of Hercules from:
> > http://www.conmicro.cx/hercules
> >
> > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> >
>
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
>
>
>
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
Jeffrey C Barnard
2002-08-27 02:13:42 UTC
Permalink
S/390's and the new z800/z900 processors are not PowerPC processors. I got to
talk to one of the designers once and he indicated that S/390's are part
hardwired, part microcoded. He said they designed their own chips. I actually
have one (IBM gave some out as key chains - not sure why I got one as they did
not have many to give out).

Regards,
Jeff

Paul Raulerson wrote:
>
> Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes, AS/400, and RS/6000 machines)
> all use the same PowerPC processor. Only the RS/6000 uses it natively though; the 400 and mainframes have software / microcode
> layers that emulate a different processor. The S/390 Architecture is implemented primarily via Microcode these days.
>
> -Paul


--
Jeffrey C Barnard
Barnard Software, Inc. http://www.bsiopti.com
Phone 407-323-4773 Fax 407-323-4775
bass
2002-08-27 02:20:17 UTC
Permalink
If I recall correctly, IBM has already announced(or has been speculated)
that the next series of the z900's (whatever they call it) will be based on
their next generation of a PowerPC processor.


Sam Bass

-----Original Message-----
From: Jeffrey C Barnard [mailto:bsiopti-***@public.gmane.org]
Sent: Monday, August 26, 2002 9:14 PM
To: hercules-390-***@public.gmane.org
Subject: Re: [hercules-390] An interesting idea another one?


S/390's and the new z800/z900 processors are not PowerPC processors. I got
to
talk to one of the designers once and he indicated that S/390's are part
hardwired, part microcoded. He said they designed their own chips. I
actually
have one (IBM gave some out as key chains - not sure why I got one as they
did
not have many to give out).

Regards,
Jeff

Paul Raulerson wrote:
>
> Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes,
AS/400, and RS/6000 machines)
> all use the same PowerPC processor. Only the RS/6000 uses it natively
though; the 400 and mainframes have software / microcode
> layers that emulate a different processor. The S/390 Architecture is
implemented primarily via Microcode these days.
>
> -Paul


--
Jeffrey C Barnard
Barnard Software, Inc. http://www.bsiopti.com
Phone 407-323-4773 Fax 407-323-4775



Community email addresses:
Post message: hercules-390-***@public.gmane.org
Subscribe: hercules-390-subscribe-***@public.gmane.org
Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
List owner: hercules-390-owner-***@public.gmane.org

Files and archives at:
http://groups.yahoo.com/group/hercules-390

Get the latest version of Hercules from:
http://www.conmicro.cx/hercules

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Paul Raulerson
2002-08-27 02:38:40 UTC
Permalink
S/390's and the new z800/z900 processors are most definitely Power based Chips. Note, I was inaccurate to say PowerPC, they are not
'PowerPC' chips, but an IBM modification of the base Motorola/IBM/Apple design. IBM's 'Power Architecture' is somewhat more
efficient than the chips in the Macintosh, though they are relatives of each other.

I guarantee you that at the core, the processing doing the work is a Power arch. It is indeed, the same chip powering the top of the
line RS/6000's and AS/400's, though you are correct they are surrounded by microcode and more.

-Paul

----- Original Message -----
From: "Jeffrey C Barnard" <bsiopti-***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Monday, August 26, 2002 9:13 PM
Subject: Re: [hercules-390] An interesting idea another one?


> S/390's and the new z800/z900 processors are not PowerPC processors. I got to
> talk to one of the designers once and he indicated that S/390's are part
> hardwired, part microcoded. He said they designed their own chips. I actually
> have one (IBM gave some out as key chains - not sure why I got one as they did
> not have many to give out).
>
> Regards,
> Jeff
>
> Paul Raulerson wrote:
> >
> > Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes, AS/400, and RS/6000 machines)
> > all use the same PowerPC processor. Only the RS/6000 uses it natively though; the 400 and mainframes have software / microcode
> > layers that emulate a different processor. The S/390 Architecture is implemented primarily via Microcode these days.
> >
> > -Paul
>
>
> --
> Jeffrey C Barnard
> Barnard Software, Inc. http://www.bsiopti.com
> Phone 407-323-4773 Fax 407-323-4775
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Peter D. Ward
2002-08-27 02:43:53 UTC
Permalink
Paul, I believe Jeff is correct on this one. See for instance:

http://www.research.ibm.com/journal/rd/435/check.html

PDW

Paul Raulerson wrote:

> S/390's and the new z800/z900 processors are most definitely Power based Chips. Note, I was inaccurate to say PowerPC, they are not
> 'PowerPC' chips, but an IBM modification of the base Motorola/IBM/Apple design. IBM's 'Power Architecture' is somewhat more
> efficient than the chips in the Macintosh, though they are relatives of each other.
>
> I guarantee you that at the core, the processing doing the work is a Power arch. It is indeed, the same chip powering the top of the
> line RS/6000's and AS/400's, though you are correct they are surrounded by microcode and more.
>
> -Paul
>
> ----- Original Message -----
> From: "Jeffrey C Barnard" <bsiopti-***@public.gmane.org>
> To: <hercules-390-***@public.gmane.org>
> Sent: Monday, August 26, 2002 9:13 PM
> Subject: Re: [hercules-390] An interesting idea another one?
>
> > S/390's and the new z800/z900 processors are not PowerPC processors. I got to
> > talk to one of the designers once and he indicated that S/390's are part
> > hardwired, part microcoded. He said they designed their own chips. I actually
> > have one (IBM gave some out as key chains - not sure why I got one as they did
> > not have many to give out).
> >
> > Regards,
> > Jeff
> >
> > Paul Raulerson wrote:
> > >
> > > Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes, AS/400, and RS/6000 machines)
> > > all use the same PowerPC processor. Only the RS/6000 uses it natively though; the 400 and mainframes have software / microcode
> > > layers that emulate a different processor. The S/390 Architecture is implemented primarily via Microcode these days.
> > >
> > > -Paul
> >
> >
> > --
> > Jeffrey C Barnard
> > Barnard Software, Inc. http://www.bsiopti.com
> > Phone 407-323-4773 Fax 407-323-4775
> >
> >
> >
> > Community email addresses:
> > Post message: hercules-390-***@public.gmane.org
> > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > List owner: hercules-390-owner-***@public.gmane.org
> >
> > Files and archives at:
> > http://groups.yahoo.com/group/hercules-390
> >
> > Get the latest version of Hercules from:
> > http://www.conmicro.cx/hercules
> >
> > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Paul Raulerson
2002-08-27 03:09:20 UTC
Permalink
Bet you several beers on this one Peter. The article you are referencing is from 1998 or so,
before IBM decided to admit that the core of the new CMOS systems was Power. Obviously,
S/390's execute S/390 instructions, but at the core, it is a processor instruction set emulation.

Hint: Port Flex to AIX, even on a B50, and be prepared to be amazed. It beats Intel hands down,
and beats *anything* else for emulation of a S/390.

-Paul


----- Original Message -----
From: "Peter D. Ward" <pdw-8vCV/***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Monday, August 26, 2002 9:43 PM
Subject: Re: [hercules-390] An interesting idea another one?


> Paul, I believe Jeff is correct on this one. See for instance:
>
> http://www.research.ibm.com/journal/rd/435/check.html
>
> PDW
>
> Paul Raulerson wrote:
>
> > S/390's and the new z800/z900 processors are most definitely Power based Chips. Note, I was inaccurate to say PowerPC, they are
not
> > 'PowerPC' chips, but an IBM modification of the base Motorola/IBM/Apple design. IBM's 'Power Architecture' is somewhat more
> > efficient than the chips in the Macintosh, though they are relatives of each other.
> >
> > I guarantee you that at the core, the processing doing the work is a Power arch. It is indeed, the same chip powering the top of
the
> > line RS/6000's and AS/400's, though you are correct they are surrounded by microcode and more.
> >
> > -Paul
> >
> > ----- Original Message -----
> > From: "Jeffrey C Barnard" <bsiopti-***@public.gmane.org>
> > To: <hercules-390-***@public.gmane.org>
> > Sent: Monday, August 26, 2002 9:13 PM
> > Subject: Re: [hercules-390] An interesting idea another one?
> >
> > > S/390's and the new z800/z900 processors are not PowerPC processors. I got to
> > > talk to one of the designers once and he indicated that S/390's are part
> > > hardwired, part microcoded. He said they designed their own chips. I actually
> > > have one (IBM gave some out as key chains - not sure why I got one as they did
> > > not have many to give out).
> > >
> > > Regards,
> > > Jeff
> > >
> > > Paul Raulerson wrote:
> > > >
> > > > Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes, AS/400, and RS/6000 machines)
> > > > all use the same PowerPC processor. Only the RS/6000 uses it natively though; the 400 and mainframes have software /
microcode
> > > > layers that emulate a different processor. The S/390 Architecture is implemented primarily via Microcode these days.
> > > >
> > > > -Paul
> > >
> > >
> > > --
> > > Jeffrey C Barnard
> > > Barnard Software, Inc. http://www.bsiopti.com
> > > Phone 407-323-4773 Fax 407-323-4775
> > >
> > >
> > >
> > > Community email addresses:
> > > Post message: hercules-390-***@public.gmane.org
> > > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > > List owner: hercules-390-owner-***@public.gmane.org
> > >
> > > Files and archives at:
> > > http://groups.yahoo.com/group/hercules-390
> > >
> > > Get the latest version of Hercules from:
> > > http://www.conmicro.cx/hercules
> > >
> > > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> > Community email addresses:
> > Post message: hercules-390-***@public.gmane.org
> > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > List owner: hercules-390-owner-***@public.gmane.org
> >
> > Files and archives at:
> > http://groups.yahoo.com/group/hercules-390
> >
> > Get the latest version of Hercules from:
> > http://www.conmicro.cx/hercules
> >
> > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Jay Maynard
2002-08-27 03:14:48 UTC
Permalink
On Mon, Aug 26, 2002 at 10:09:20PM -0500, Paul Raulerson wrote:
> Hint: Port Flex to AIX, even on a B50, and be prepared to be amazed. It
> beats Intel hands down, and beats *anything* else for emulation of a
> S/390.

That would be difficult at best, since Flex is heavily Intel x86-assembler
dependent...

One of these days, I'm gonna get gcc for AIX and build hercules on my C10,
just to see what it does (and serve as a springboard for the AS/400).
Paul Raulerson
2002-08-27 03:35:01 UTC
Permalink
The parts that port work well, but the parts like the tape driver and support for shared libraries, well,
they would take you about a day to figure out, and me just a day or two short of forever right now. :)

Contact me off list about whatever AIX software you need if you are really interested.

-Paul

----- Original Message -----
From: "Jay Maynard" <jmaynard-***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Monday, August 26, 2002 10:14 PM
Subject: Re: [hercules-390] An interesting idea another one?


> On Mon, Aug 26, 2002 at 10:09:20PM -0500, Paul Raulerson wrote:
> > Hint: Port Flex to AIX, even on a B50, and be prepared to be amazed. It
> > beats Intel hands down, and beats *anything* else for emulation of a
> > S/390.
>
> That would be difficult at best, since Flex is heavily Intel x86-assembler
> dependent...
>
> One of these days, I'm gonna get gcc for AIX and build hercules on my C10,
> just to see what it does (and serve as a springboard for the AS/400).
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Gregg C Levine
2002-08-27 03:45:24 UTC
Permalink
Hello again from Gregg C Levine
Did I start all of this? Nice to know that on this list I'm taken
seriously. It normally does not happen on the L390 list that lives at
Marist. It usually takes two or three posts, there.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> Sent: Monday, August 26, 2002 11:35 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] An interesting idea another one?
>
> The parts that port work well, but the parts like the tape driver and
support for
> shared libraries, well,
> they would take you about a day to figure out, and me just a day or
two short of
> forever right now. :)
>
> Contact me off list about whatever AIX software you need if you are
really
> interested.
>
> -Paul
>
> ----- Original Message -----
> From: "Jay Maynard" <jmaynard-***@public.gmane.org>
> To: <hercules-390-***@public.gmane.org>
> Sent: Monday, August 26, 2002 10:14 PM
> Subject: Re: [hercules-390] An interesting idea another one?
>
>
> > On Mon, Aug 26, 2002 at 10:09:20PM -0500, Paul Raulerson wrote:
> > > Hint: Port Flex to AIX, even on a B50, and be prepared to be
amazed. It
> > > beats Intel hands down, and beats *anything* else for emulation of
a
> > > S/390.
> >
> > That would be difficult at best, since Flex is heavily Intel
x86-assembler
> > dependent...
> >
> > One of these days, I'm gonna get gcc for AIX and build hercules on
my C10,
> > just to see what it does (and serve as a springboard for the
AS/400).
> >
> >
> > Community email addresses:
> > Post message: hercules-390-***@public.gmane.org
> > Subscribe: hercules-390-subscribe-***@public.gmane.org
> > Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> > List owner: hercules-390-owner-***@public.gmane.org
> >
> > Files and archives at:
> > http://groups.yahoo.com/group/hercules-390
> >
> > Get the latest version of Hercules from:
> > http://www.conmicro.cx/hercules
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
scott_blackledge
2002-09-04 02:10:52 UTC
Permalink
Paul,

You'd owe the beers. I work (for another week) about three
rows down from the HW designers. The z900 chips are custom. Jeff
is correct.

Scott

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> Bet you several beers on this one Peter. The article you are
referencing is from 1998 or so,
> before IBM decided to admit that the core of the new CMOS systems
was Power. Obviously,
> S/390's execute S/390 instructions, but at the core, it is a
processor instruction set emulation.
>
> Hint: Port Flex to AIX, even on a B50, and be prepared to be amazed.
It beats Intel hands down,
> and beats *anything* else for emulation of a S/390.
>
> -Paul
>
>
> ----- Original Message -----
> From: "Peter D. Ward" <pdw-***@public.gmane.org>
> To: <hercules-390-F5Bj5G+***@public.gmane.org>
> Sent: Monday, August 26, 2002 9:43 PM
> Subject: Re: [hercules-390] An interesting idea another one?
>
>
> > Paul, I believe Jeff is correct on this one. See for instance:
> >
> > http://www.research.ibm.com/journal/rd/435/check.html
> >
> > PDW
> >
> > Paul Raulerson wrote:
> >
> > > S/390's and the new z800/z900 processors are most definitely
Power based Chips. Note, I was inaccurate to say PowerPC, they are
> not
> > > 'PowerPC' chips, but an IBM modification of the base
Motorola/IBM/Apple design. IBM's 'Power Architecture' is somewhat more
> > > efficient than the chips in the Macintosh, though they are
relatives of each other.
> > >
> > > I guarantee you that at the core, the processing doing the work
is a Power arch. It is indeed, the same chip powering the top of
> the
> > > line RS/6000's and AS/400's, though you are correct they are
surrounded by microcode and more.
> > >
> > > -Paul
> > >
> > > ----- Original Message -----
> > > From: "Jeffrey C Barnard" <bsiopti-***@public.gmane.org>
> > > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > > Sent: Monday, August 26, 2002 9:13 PM
> > > Subject: Re: [hercules-390] An interesting idea another one?
> > >
> > > > S/390's and the new z800/z900 processors are not PowerPC
processors. I got to
> > > > talk to one of the designers once and he indicated that
S/390's are part
> > > > hardwired, part microcoded. He said they designed their own
chips. I actually
> > > > have one (IBM gave some out as key chains - not sure why I got
one as they did
> > > > not have many to give out).
> > > >
> > > > Regards,
> > > > Jeff
> > > >
> > > > Paul Raulerson wrote:
> > > > >
> > > > > Not exactly - zSeries, iSeries, and pSeries machines (i.e.
Mainframes, AS/400, and RS/6000 machines)
> > > > > all use the same PowerPC processor. Only the RS/6000 uses
it natively though; the 400 and mainframes have software /
> microcode
> > > > > layers that emulate a different processor. The S/390
Architecture is implemented primarily via Microcode these days.
> > > > >
> > > > > -Paul
> > > > --
> > > > Jeffrey C Barnard
scott_blackledge
2002-09-04 02:10:30 UTC
Permalink
Paul,

You'd owe the beers. I work (for another week) about three
rows down from the HW designers. The z900 chips are custom. Jeff
is correct.

Scott

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> Bet you several beers on this one Peter. The article you are
referencing is from 1998 or so,
> before IBM decided to admit that the core of the new CMOS systems
was Power. Obviously,
> S/390's execute S/390 instructions, but at the core, it is a
processor instruction set emulation.
>
> Hint: Port Flex to AIX, even on a B50, and be prepared to be amazed.
It beats Intel hands down,
> and beats *anything* else for emulation of a S/390.
>
> -Paul
>
>
> ----- Original Message -----
> From: "Peter D. Ward" <pdw-***@public.gmane.org>
> To: <hercules-390-F5Bj5G+***@public.gmane.org>
> Sent: Monday, August 26, 2002 9:43 PM
> Subject: Re: [hercules-390] An interesting idea another one?
>
>
> > Paul, I believe Jeff is correct on this one. See for instance:
> >
> > http://www.research.ibm.com/journal/rd/435/check.html
> >
> > PDW
> >
> > Paul Raulerson wrote:
> >
> > > S/390's and the new z800/z900 processors are most definitely
Power based Chips. Note, I was inaccurate to say PowerPC, they are
> not
> > > 'PowerPC' chips, but an IBM modification of the base
Motorola/IBM/Apple design. IBM's 'Power Architecture' is somewhat more
> > > efficient than the chips in the Macintosh, though they are
relatives of each other.
> > >
> > > I guarantee you that at the core, the processing doing the work
is a Power arch. It is indeed, the same chip powering the top of
> the
> > > line RS/6000's and AS/400's, though you are correct they are
surrounded by microcode and more.
> > >
> > > -Paul
> > >
> > > ----- Original Message -----
> > > From: "Jeffrey C Barnard" <bsiopti-***@public.gmane.org>
> > > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > > Sent: Monday, August 26, 2002 9:13 PM
> > > Subject: Re: [hercules-390] An interesting idea another one?
> > >
> > > > S/390's and the new z800/z900 processors are not PowerPC
processors. I got to
> > > > talk to one of the designers once and he indicated that
S/390's are part
> > > > hardwired, part microcoded. He said they designed their own
chips. I actually
> > > > have one (IBM gave some out as key chains - not sure why I got
one as they did
> > > > not have many to give out).
> > > >
> > > > Regards,
> > > > Jeff
> > > >
> > > > Paul Raulerson wrote:
> > > > >
> > > > > Not exactly - zSeries, iSeries, and pSeries machines (i.e.
Mainframes, AS/400, and RS/6000 machines)
> > > > > all use the same PowerPC processor. Only the RS/6000 uses
it natively though; the 400 and mainframes have software /
> microcode
> > > > > layers that emulate a different processor. The S/390
Architecture is implemented primarily via Microcode these days.
> > > > >
> > > > > -Paul
> > > > --
> > > > Jeffrey C Barnard
Joe Zitzelberger
2002-09-04 12:57:25 UTC
Permalink
On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge wrote:
> Paul,
>
> You'd owe the beers. I work (for another week) about three
> rows down from the HW designers. The z900 chips are custom. Jeff
> is correct.
>
> Scott

That article makes the following very vague, but important statement:

"To simplify the logic design, the processor uses millicode [2] (a form
of Licensed Internal Code), which is the vertical microcode that
executes on this family of processors to implement many complex
elements of the architecture. Millicode consists of all of the
hardwired ESA/390 instructions plus 102 hardwired assist instructions
that can only be executed by millicode."


The question becomes 'What is executing the millicode'? IIRC Power
was, except in its low end consumer and embedded versions, was a set of
6 or 9 chips wired up to play well with each other. Did they borrow
some subset of those?


psychedelic-harry-9q/xBM6aKHVWk0Htik3J/***@public.gmane.org

Web Servers Do It With Cookies
Paul Raulerson
2002-09-05 00:31:25 UTC
Permalink
<grin> Trust me - that is a Power Architecture in there. :)

----- Original Message -----
From: "Joe Zitzelberger" <psychedelic-harry-9q/xBM6aKHVWk0Htik3J/***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Wednesday, September 04, 2002 7:57 AM
Subject: Re: [hercules-390] Re: An interesting idea another one?


>
> On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge wrote:
> > Paul,
> >
> > You'd owe the beers. I work (for another week) about three
> > rows down from the HW designers. The z900 chips are custom. Jeff
> > is correct.
> >
> > Scott
>
> That article makes the following very vague, but important statement:
>
> "To simplify the logic design, the processor uses millicode [2] (a form
> of Licensed Internal Code), which is the vertical microcode that
> executes on this family of processors to implement many complex
> elements of the architecture. Millicode consists of all of the
> hardwired ESA/390 instructions plus 102 hardwired assist instructions
> that can only be executed by millicode."
>
>
> The question becomes 'What is executing the millicode'? IIRC Power
> was, except in its low end consumer and embedded versions, was a set of
> 6 or 9 chips wired up to play well with each other. Did they borrow
> some subset of those?
>
>
> psychedelic-harry-9q/xBM6aKHVWk0Htik3J/***@public.gmane.org
>
> Web Servers Do It With Cookies
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
scott_blackledge
2002-09-06 02:37:49 UTC
Permalink
There *IS NO POWER ARCHITECTURE IN Z900 HARDWARE!*, from my personal
experience as a z900 Millicoder, dealing with the HW designers
directly. Custom chips run the millicoded instruction set. There
is nothing in the quoted article that even remotely implies the use
of Power architecture.

Scott

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> <grin> Trust me - that is a Power Architecture in there. :)
>
> ----- Original Message -----
> From: "Joe Zitzelberger" <psychedelic-harry-***@public.gmane.org>
> To: <hercules-390-F5Bj5G+***@public.gmane.org>
> Sent: Wednesday, September 04, 2002 7:57 AM
> Subject: Re: [hercules-390] Re: An interesting idea another one?
>
>
> >
> > On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge
wrote:
> > > Paul,
> > >
> > > You'd owe the beers. I work (for another week) about three
> > > rows down from the HW designers. The z900 chips are custom.
Jeff
> > > is correct.
> > >
> > > Scott
> >
> > That article makes the following very vague, but important
statement:
> >
> > "To simplify the logic design, the processor uses millicode [2] (a
form
> > of Licensed Internal Code), which is the vertical microcode that
> > executes on this family of processors to implement many complex
> > elements of the architecture. Millicode consists of all of the
> > hardwired ESA/390 instructions plus 102 hardwired assist
instructions
> > that can only be executed by millicode."
> >
> >
> > The question becomes 'What is executing the millicode'? IIRC
Power
> > was, except in its low end consumer and embedded versions, was a
set of
> > 6 or 9 chips wired up to play well with each other. Did they
borrow
> > some subset of those?
> >
> >
> > psychedelic-harry-***@public.gmane.org
Paul Raulerson
2002-09-06 12:03:43 UTC
Permalink
Okay Scott - what do you think those "custom chips" are? From my discussions with the hardware engineers, and from published IBM
promotional materials, I disagree.


----- Original Message -----
From: "scott_blackledge" <scottb2-jG/***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Thursday, September 05, 2002 9:37 PM
Subject: [hercules-390] Re: An interesting idea another one?


> There *IS NO POWER ARCHITECTURE IN Z900 HARDWARE!*, from my personal
> experience as a z900 Millicoder, dealing with the HW designers
> directly. Custom chips run the millicoded instruction set. There
> is nothing in the quoted article that even remotely implies the use
> of Power architecture.
>
> Scott
>
> --- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> > <grin> Trust me - that is a Power Architecture in there. :)
> >
> > ----- Original Message -----
> > From: "Joe Zitzelberger" <psychedelic-harry-***@public.gmane.org>
> > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > Sent: Wednesday, September 04, 2002 7:57 AM
> > Subject: Re: [hercules-390] Re: An interesting idea another one?
> >
> >
> > >
> > > On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge
> wrote:
> > > > Paul,
> > > >
> > > > You'd owe the beers. I work (for another week) about three
> > > > rows down from the HW designers. The z900 chips are custom.
> Jeff
> > > > is correct.
> > > >
> > > > Scott
> > >
> > > That article makes the following very vague, but important
> statement:
> > >
> > > "To simplify the logic design, the processor uses millicode [2] (a
> form
> > > of Licensed Internal Code), which is the vertical microcode that
> > > executes on this family of processors to implement many complex
> > > elements of the architecture. Millicode consists of all of the
> > > hardwired ESA/390 instructions plus 102 hardwired assist
> instructions
> > > that can only be executed by millicode."
> > >
> > >
> > > The question becomes 'What is executing the millicode'? IIRC
> Power
> > > was, except in its low end consumer and embedded versions, was a
> set of
> > > 6 or 9 chips wired up to play well with each other. Did they
> borrow
> > > some subset of those?
> > >
> > >
> > > psychedelic-harry-***@public.gmane.org
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
scott_blackledge
2002-09-08 02:49:21 UTC
Permalink
Paul,

I think the "custom chips" are just that, custom chips. As I said,
I work directly with these guys. They are not Power chips. In fact,
lets take a look at the quoted paragraph.

"To simplify the logic design, the processor uses millicode [2] (a
form
of Licensed Internal Code), which is the vertical microcode that
executes on this family of processors to implement many complex
elements of the architecture. Millicode consists of all of the
hardwired ESA/390 instructions plus 102 hardwired assist
instructions
that can only be executed by millicode."

Millicode is the name of the firmware/microcode that has been used
for a few generations of mainframes. "Horizontal" simply refers to a
specific type (executing in parallel). "Vertical" refers to the
current type of microcode written in assembler. A set of
instructions, both public and millicode, are executed directly by
hardware. The rest intercepted and then emulated by millicode. All
hardwired instructions can be executed by millicode.

There is absolutely nothing in the quoted paragraph that implies
the Power architecture. If you still disagree, please be more
specific (quotes please) about the information you have to the
contrary.

Scott


--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> Okay Scott - what do you think those "custom chips" are? From my
discussions with the hardware engineers, and from published IBM
> promotional materials, I disagree.
>
>
> ----- Original Message -----
> From: "scott_blackledge" <scottb2-***@public.gmane.org>
> To: <hercules-390-F5Bj5G+***@public.gmane.org>
> Sent: Thursday, September 05, 2002 9:37 PM
> Subject: [hercules-390] Re: An interesting idea another one?
>
>
> > There *IS NO POWER ARCHITECTURE IN Z900 HARDWARE!*, from my
personal
> > experience as a z900 Millicoder, dealing with the HW designers
> > directly. Custom chips run the millicoded instruction set. There
> > is nothing in the quoted article that even remotely implies the
use
> > of Power architecture.
> >
> > Scott
> >
> > --- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...>
wrote:
> > > <grin> Trust me - that is a Power Architecture in there. :)
> > >
> > > ----- Original Message -----
> > > From: "Joe Zitzelberger" <psychedelic-harry-***@public.gmane.org>
> > > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > > Sent: Wednesday, September 04, 2002 7:57 AM
> > > Subject: Re: [hercules-390] Re: An interesting idea another one?
> > >
> > >
> > > >
> > > > On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge
> > wrote:
> > > > > Paul,
> > > > >
> > > > > You'd owe the beers. I work (for another week) about
three
> > > > > rows down from the HW designers. The z900 chips are custom.
> > Jeff
> > > > > is correct.
> > > > >
> > > > > Scott
> > > >
> > > > That article makes the following very vague, but important
> > statement:
> > > >
> > > > "To simplify the logic design, the processor uses millicode
[2] (a
> > form
> > > > of Licensed Internal Code), which is the vertical microcode
that
> > > > executes on this family of processors to implement many
complex
> > > > elements of the architecture. Millicode consists of all of the
> > > > hardwired ESA/390 instructions plus 102 hardwired assist
> > instructions
> > > > that can only be executed by millicode."
> > > >
> > > >
> > > > The question becomes 'What is executing the millicode'? IIRC
> > Power
> > > > was, except in its low end consumer and embedded versions, was
a
> > set of
> > > > 6 or 9 chips wired up to play well with each other. Did they
> > borrow
> > > > some subset of those?
> > > >
> > > >
> > > > psychedelic-harry-***@public.gmane.org
Paul Raulerson
2002-09-08 03:14:26 UTC
Permalink
Look Scott - I am very sure you believe you are right, so why argue?
But I will leave you with two thoughts - where do you think the model for Power architecture came from? Hint, take a close look at
the millicode you are writing and then take another look at the Power instruction set.

and two- IBM has already announced that Mainframe will move to Power Architeture, along with using Infinband networking and a whole
bunch of other stuff. You have internal access to IBM Link, instead of being stubborn, go and talk the hardware engineers, look it
up in the memos published on IBM Link, or just go and ask one of the mainframe evangelists around.

Or do a Google search for yourself on it.

No possible reference or pointer I give you is going to get you to honestly re-evaluate what you believe to be the truth. So be it.

-Paul
----- Original Message -----
From: "scott_blackledge" <scottb2-jG/***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Saturday, September 07, 2002 9:49 PM
Subject: [hercules-390] Re: An interesting idea another one?


> Paul,
>
> I think the "custom chips" are just that, custom chips. As I said,
> I work directly with these guys. They are not Power chips. In fact,
> lets take a look at the quoted paragraph.
>
> "To simplify the logic design, the processor uses millicode [2] (a
> form
> of Licensed Internal Code), which is the vertical microcode that
> executes on this family of processors to implement many complex
> elements of the architecture. Millicode consists of all of the
> hardwired ESA/390 instructions plus 102 hardwired assist
> instructions
> that can only be executed by millicode."
>
> Millicode is the name of the firmware/microcode that has been used
> for a few generations of mainframes. "Horizontal" simply refers to a
> specific type (executing in parallel). "Vertical" refers to the
> current type of microcode written in assembler. A set of
> instructions, both public and millicode, are executed directly by
> hardware. The rest intercepted and then emulated by millicode. All
> hardwired instructions can be executed by millicode.
>
> There is absolutely nothing in the quoted paragraph that implies
> the Power architecture. If you still disagree, please be more
> specific (quotes please) about the information you have to the
> contrary.
>
> Scott
>
>
> --- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> > Okay Scott - what do you think those "custom chips" are? From my
> discussions with the hardware engineers, and from published IBM
> > promotional materials, I disagree.
> >
> >
> > ----- Original Message -----
> > From: "scott_blackledge" <scottb2-***@public.gmane.org>
> > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > Sent: Thursday, September 05, 2002 9:37 PM
> > Subject: [hercules-390] Re: An interesting idea another one?
> >
> >
> > > There *IS NO POWER ARCHITECTURE IN Z900 HARDWARE!*, from my
> personal
> > > experience as a z900 Millicoder, dealing with the HW designers
> > > directly. Custom chips run the millicoded instruction set. There
> > > is nothing in the quoted article that even remotely implies the
> use
> > > of Power architecture.
> > >
> > > Scott
> > >
> > > --- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...>
> wrote:
> > > > <grin> Trust me - that is a Power Architecture in there. :)
> > > >
> > > > ----- Original Message -----
> > > > From: "Joe Zitzelberger" <psychedelic-harry-***@public.gmane.org>
> > > > To: <hercules-390-F5Bj5G+***@public.gmane.org>
> > > > Sent: Wednesday, September 04, 2002 7:57 AM
> > > > Subject: Re: [hercules-390] Re: An interesting idea another one?
> > > >
> > > >
> > > > >
> > > > > On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge
> > > wrote:
> > > > > > Paul,
> > > > > >
> > > > > > You'd owe the beers. I work (for another week) about
> three
> > > > > > rows down from the HW designers. The z900 chips are custom.
> > > Jeff
> > > > > > is correct.
> > > > > >
> > > > > > Scott
> > > > >
> > > > > That article makes the following very vague, but important
> > > statement:
> > > > >
> > > > > "To simplify the logic design, the processor uses millicode
> [2] (a
> > > form
> > > > > of Licensed Internal Code), which is the vertical microcode
> that
> > > > > executes on this family of processors to implement many
> complex
> > > > > elements of the architecture. Millicode consists of all of the
> > > > > hardwired ESA/390 instructions plus 102 hardwired assist
> > > instructions
> > > > > that can only be executed by millicode."
> > > > >
> > > > >
> > > > > The question becomes 'What is executing the millicode'? IIRC
> > > Power
> > > > > was, except in its low end consumer and embedded versions, was
> a
> > > set of
> > > > > 6 or 9 chips wired up to play well with each other. Did they
> > > borrow
> > > > > some subset of those?
> > > > >
> > > > >
> > > > > psychedelic-harry-***@public.gmane.org
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Fish
2002-09-08 03:40:07 UTC
Permalink
http://zdnet.com.com/2100-1103-919690.html
http://news.com.com/2100-1001-919579.html

- From the sounds of it, the zSeries isn't YET based on the Power
architecture, but it WILL be sometime around the year 2006??

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org


> -----Original Message-----
> From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> Sent: Saturday, September 07, 2002 8:14 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] Re: An interesting idea another one?
>
>
> Look Scott - I am very sure you believe you are right, so why
> argue?

<snip>
Mark Perry
2002-09-08 09:18:58 UTC
Permalink
Hi All,
I have taken the long and arduous task of zooming in on the Key ring picture
I had uploaded previously.

see in the files section: G7_IBM_KEY RING_CHIP_ZOOM.jpg

In the zoom in I am pretty sure those experts amongst you will be able to
discern the origin of the z/900 CPU.

Ciao
Mark
-----Original Message-----
From: Fish [mailto:fish-6N/dkqvhA+***@public.gmane.org]
Sent: 08 September 2002 05:40
To: hercules-390-***@public.gmane.org
Subject: RE: [hercules-390] Re: An interesting idea another one?



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

http://zdnet.com.com/2100-1103-919690.html
http://news.com.com/2100-1001-919579.html

- From the sounds of it, the zSeries isn't YET based on the Power
architecture, but it WILL be sometime around the year 2006??

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org


> -----Original Message-----
> From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> Sent: Saturday, September 07, 2002 8:14 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] Re: An interesting idea another one?
>
>
> Look Scott - I am very sure you believe you are right, so why
> argue?

<snip>

-----BEGIN PGP SIGNATURE-----
Version: PGP 7.0.4

iQA/AwUBPXrGlkj11/TE7j4qEQJLtgCeMvs+a/5Hzf7wu8gk1JOwXIfZkqUAoKkC
mDigEb4WeKW7au2IP/ErmxCP
=sSY9
-----END PGP SIGNATURE-----


Yahoo! Groups Sponsor
ADVERTISEMENT



Community email addresses:
Post message: hercules-390-***@public.gmane.org
Subscribe: hercules-390-subscribe-***@public.gmane.org
Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
List owner: hercules-390-owner-***@public.gmane.org

Files and archives at:
http://groups.yahoo.com/group/hercules-390

Get the latest version of Hercules from:
http://www.conmicro.cx/hercules

Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.


[Non-text portions of this message have been removed]
zapzap50
2002-09-08 09:53:30 UTC
Permalink
Nice one
I knew the operating system could make the difference <grin>
Bruno

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Mark Perry" <***@h...> wrote:
> Hi All,
> I have taken the long and arduous task of zooming in on the Key ring
picture
> I had uploaded previously.
>
> see in the files section: G7_IBM_KEY RING_CHIP_ZOOM.jpg
>
> In the zoom in I am pretty sure those experts amongst you will be
able to
> discern the origin of the z/900 CPU.
>
> Ciao
> Mark
> -----Original Message-----
> From: Fish [mailto:fish-***@public.gmane.org]
> Sent: 08 September 2002 05:40
> To: hercules-390-F5Bj5G+***@public.gmane.org
> Subject: RE: [hercules-390] Re: An interesting idea another one?
>
>
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> http://zdnet.com.com/2100-1103-919690.html
> http://news.com.com/2100-1001-919579.html
>
> - From the sounds of it, the zSeries isn't YET based on the Power
> architecture, but it WILL be sometime around the year 2006??
>
> - --
> "Fish" (David B. Trout)
> fish-***@public.gmane.org
>
>
> > -----Original Message-----
> > From: Paul Raulerson [mailto:praulerson-***@public.gmane.org]
> > Sent: Saturday, September 07, 2002 8:14 PM
> > To: hercules-390-F5Bj5G+***@public.gmane.org
> > Subject: Re: [hercules-390] Re: An interesting idea another one?
> >
> >
> > Look Scott - I am very sure you believe you are right, so why
> > argue?
>
> <snip>
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXrGlkj11/TE7j4qEQJLtgCeMvs+a/5Hzf7wu8gk1JOwXIfZkqUAoKkC
> mDigEb4WeKW7au2IP/ErmxCP
> =sSY9
> -----END PGP SIGNATURE-----
>
>
> Yahoo! Groups Sponsor
> ADVERTISEMENT
>
>
>
> Community email addresses:
> Post message: hercules-390-F5Bj5G+***@public.gmane.org
> Subscribe: hercules-390-subscribe-F5Bj5G+***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-F5Bj5G+***@public.gmane.org
> List owner: hercules-390-owner-F5Bj5G+***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
>
>
> [Non-text portions of this message have been removed]
Volker Bandke
2002-09-08 10:15:20 UTC
Permalink
He cheated! He overpainted the zilog Z80A logo! Shame on him....






With kind Regards |\ _,,,---,,_
ZZZzz /,`.-'`' -. ;-;;,
Volker Bandke |,4- ) )-,_. ,\ ( `'-'
(BSP GmbH) '---''(_/--' `-'\_)

If it wasn't so warm out today, it would be cooler.

(Another Wisdom from my fortune cookie jar)
Paul Raulerson
2002-09-08 14:12:08 UTC
Permalink
ROTFLMAO!!!!!!!
----- Original Message -----
From: "Mark Perry" <mark.perry-***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Sunday, September 08, 2002 4:18 AM
Subject: RE: [hercules-390] Re: An interesting idea another one?


> Hi All,
> I have taken the long and arduous task of zooming in on the Key ring picture
> I had uploaded previously.
>
> see in the files section: G7_IBM_KEY RING_CHIP_ZOOM.jpg
>
> In the zoom in I am pretty sure those experts amongst you will be able to
> discern the origin of the z/900 CPU.
>
> Ciao
> Mark
> -----Original Message-----
> From: Fish [mailto:fish-6N/dkqvhA+***@public.gmane.org]
> Sent: 08 September 2002 05:40
> To: hercules-390-***@public.gmane.org
> Subject: RE: [hercules-390] Re: An interesting idea another one?
>
>
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> http://zdnet.com.com/2100-1103-919690.html
> http://news.com.com/2100-1001-919579.html
>
> - From the sounds of it, the zSeries isn't YET based on the Power
> architecture, but it WILL be sometime around the year 2006??
>
> - --
> "Fish" (David B. Trout)
> fish-6N/dkqvhA+***@public.gmane.org
>
>
> > -----Original Message-----
> > From: Paul Raulerson [mailto:praulerson-qMZMQ02z5afQT0dZR+***@public.gmane.org]
> > Sent: Saturday, September 07, 2002 8:14 PM
> > To: hercules-390-***@public.gmane.org
> > Subject: Re: [hercules-390] Re: An interesting idea another one?
> >
> >
> > Look Scott - I am very sure you believe you are right, so why
> > argue?
>
> <snip>
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXrGlkj11/TE7j4qEQJLtgCeMvs+a/5Hzf7wu8gk1JOwXIfZkqUAoKkC
> mDigEb4WeKW7au2IP/ErmxCP
> =sSY9
> -----END PGP SIGNATURE-----
>
>
> Yahoo! Groups Sponsor
> ADVERTISEMENT
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
>
>
> [Non-text portions of this message have been removed]
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
scott_blackledge
2002-09-09 02:52:38 UTC
Permalink
Paul,

That is a rather condescending reply. I've given you my reasons
(personal experience with today's HW developers) along with an
explanation of the quote. I also asked for quotes that make you
believe that there is Power architecture in *today's* z900. All I
get is this rather vague reply.

I've read the articles that Fish provided and nothing there states
that there is Power architecture in *today's* z900. The articles
certainly state that the future of the eSeries is convergence but
that is the future. I'm talking about now. Also, if I take your
hint, it sounds like you think the Power architecture was developed
from the mainframe architecture. This may be true, but it really
has nothing to do with what is in the z900 chipset *today*.

Lastly, since you don't know me, it is rather presumptuous to
assume that I wouldn't give any of your references any serious
thought. Its rather a moot point, though, since you haven't
provided any.

Scott

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul Raulerson" <***@h...> wrote:
> Look Scott - I am very sure you believe you are right, so why argue?
> But I will leave you with two thoughts - where do you think the
model for Power architecture came from? Hint, take a close look at
> the millicode you are writing and then take another look at the
Power instruction set.
>
> and two- IBM has already announced that Mainframe will move to Power
Architeture, along with using Infinband networking and a whole
> bunch of other stuff. You have internal access to IBM Link, instead
of being stubborn, go and talk the hardware engineers, look it
> up in the memos published on IBM Link, or just go and ask one of the
mainframe evangelists around.
>
> Or do a Google search for yourself on it.
>
> No possible reference or pointer I give you is going to get you to
honestly re-evaluate what you believe to be the truth. So be it.
>
> -Paul
> ----- Original Message -----
> From: "scott_blackledge" <scottb2-***@public.gmane.org>
> To: <hercules-390-F5Bj5G+***@public.gmane.org>
> Sent: Saturday, September 07, 2002 9:49 PM
> Subject: [hercules-390] Re: An interesting idea another one?
>
>
> > Paul,
> >
> > I think the "custom chips" are just that, custom chips. As I
said,
> > I work directly with these guys. They are not Power chips. In
fact,
> > lets take a look at the quoted paragraph.
> >
> > "To simplify the logic design, the processor uses millicode
[2] (a
> > form
> > of Licensed Internal Code), which is the vertical microcode
that
> > executes on this family of processors to implement many
complex
> > elements of the architecture. Millicode consists of all of the
> > hardwired ESA/390 instructions plus 102 hardwired assist
> > instructions
> > that can only be executed by millicode."
> >
> > Millicode is the name of the firmware/microcode that has been used
> > for a few generations of mainframes. "Horizontal" simply refers
to a
> > specific type (executing in parallel). "Vertical" refers to the
> > current type of microcode written in assembler. A set of
> > instructions, both public and millicode, are executed directly by
> > hardware. The rest intercepted and then emulated by millicode.
All
> > hardwired instructions can be executed by millicode.
> >
> > There is absolutely nothing in the quoted paragraph that implies
> > the Power architecture. If you still disagree, please be more
> > specific (quotes please) about the information you have to the
> > contrary.
> >
> > Scott
Fish
2002-09-05 08:07:33 UTC
Permalink
<snip>

> "To simplify the logic design, the processor uses millicode [2]
> (a form of Licensed Internal Code), which is the vertical microcode
> that executes on this family of processors to implement many
> complex elements of the architecture. Millicode consists of all of
> the
> hardwired ESA/390 instructions plus 102 hardwired assist
> instructions that can only be executed by millicode."

<snip>

Hmmm... "vertical microcode"...

it's a little fuzzy after all these years, but I seem to remember the
documentation for the AS/400 mentioning "vertical" and "horizontal"
microcode too, didn't it?

As I recall, the various compilers generated "machine instructions"
(MI) which were then processed by the vertical microcode at execution
time, the o/p of which was then of course processed by the
*horizontal* microcode which did the actual work.

I personally interpreted this to mean that the vertical microcode
essentially just "compiled" the CISC "machine instructions" into the
RISC instructions that the actual underlying hardware actually
executed.

That is to say, the actual hardware (the CPUs that "actually executed
the instructions" in the real sense that we're all familiar with
(i.e. the horizontal microcode layer)) was simply a RISC setup that
was simply "fed" instructions from the slightly higher-level vertical
microcode layer that more or less simply "compiled" (or rather
"interpreted" at "execution" (run) time) what the human being user
was supposed to understand as being "the machine instructions".

I.e.:

CISC [pseudo-]"instruction" -->
--> "vertical" microcode -->
--> bona fide RISC instructions -->
--> horizontal microcode.

At least that was MY interpretation / understanding of how it worked.
<shrug>

Never did find out whether I actually had it right though. Can anyone
in an authoritative position wrt this issue confirm/deny whether or
not I do?

Thanks. :)

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org
Ronald Tatum
2002-09-05 16:31:15 UTC
Permalink
Fish,
Your post jogged my memory a bit; sure enough, there was a paper in the
December 1969 Computing Surveys about microprogramming. I found the pub in
my basement stash.
Here's the abstract:
"Microprogramming is introduced as a vehicle for implementing the control
function of a computer. Some contemporary systems using this technique are
described and several justifications are given for its application. Finally
a set of specific research problem areas are suggested shich center around
the viability and value of the most general use of writable control stores."
The paper is sixteen pages, including the one-page bibliography. No, I
*won't* transcribe the paper for you, even though you have accused me of
being overly prolix :-). I would, however, be willing to scan the bloody
thing and use Rob Storey's neat program to convert the images to
(non-editable) .PDF format if you'd like a copy for yourself or to post to
the files area if anyone's interested in such "old stuff".
Regards,
Ron Tatum
----- Original Message -----
From: "Fish" <fish-6N/dkqvhA+***@public.gmane.org>
To: <hercules-390-***@public.gmane.org>
Sent: Thursday, September 05, 2002 3:07 AM
Subject: RE: [hercules-390] Re: An interesting idea another one?


>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> <snip>
>
> > "To simplify the logic design, the processor uses millicode [2]
> > (a form of Licensed Internal Code), which is the vertical microcode
> > that executes on this family of processors to implement many
> > complex elements of the architecture. Millicode consists of all of
> > the
> > hardwired ESA/390 instructions plus 102 hardwired assist
> > instructions that can only be executed by millicode."
>
> <snip>
>
> Hmmm... "vertical microcode"...
>
> it's a little fuzzy after all these years, but I seem to remember the
> documentation for the AS/400 mentioning "vertical" and "horizontal"
> microcode too, didn't it?
>
> As I recall, the various compilers generated "machine instructions"
> (MI) which were then processed by the vertical microcode at execution
> time, the o/p of which was then of course processed by the
> *horizontal* microcode which did the actual work.
>
> I personally interpreted this to mean that the vertical microcode
> essentially just "compiled" the CISC "machine instructions" into the
> RISC instructions that the actual underlying hardware actually
> executed.
>
> That is to say, the actual hardware (the CPUs that "actually executed
> the instructions" in the real sense that we're all familiar with
> (i.e. the horizontal microcode layer)) was simply a RISC setup that
> was simply "fed" instructions from the slightly higher-level vertical
> microcode layer that more or less simply "compiled" (or rather
> "interpreted" at "execution" (run) time) what the human being user
> was supposed to understand as being "the machine instructions".
>
> I.e.:
>
> CISC [pseudo-]"instruction" -->
> --> "vertical" microcode -->
> --> bona fide RISC instructions -->
> --> horizontal microcode.
>
> At least that was MY interpretation / understanding of how it worked.
> <shrug>
>
> Never did find out whether I actually had it right though. Can anyone
> in an authoritative position wrt this issue confirm/deny whether or
> not I do?
>
> Thanks. :)
>
> - --
> "Fish" (David B. Trout)
> fish-6N/dkqvhA+***@public.gmane.org
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXcQxEj11/TE7j4qEQJifgCg3N7rnLZrQ9ZkciC7jWM04h4rmGsAoKIs
> nFiR8JsHbyUna+JR4TU3gR/j
> =tu2v
> -----END PGP SIGNATURE-----
>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
CTB
2002-09-05 17:00:26 UTC
Permalink
Ron,

You took me back to when I started on the 360s in the fall of 1966. Of
the original 360s, the 30, 40 and 50 were all microprogrammed. The 65
an higher were hard wired. I think starting with the 80 the rest were
all microprogrammed (but I wouldn't stand very hard on that). I
remember hearing that the channels on the 65 and above were the same box
as the 360/30 but with different microcode.

Times do change.

Carter

Ronald Tatum wrote:

>Fish,
> Your post jogged my memory a bit; sure enough, there was a paper in the
>December 1969 Computing Surveys about microprogramming. I found the pub in
>my basement stash.
> Here's the abstract:
>"Microprogramming is introduced as a vehicle for implementing the control
>function of a computer. Some contemporary systems using this technique are
>described and several justifications are given for its application. Finally
>a set of specific research problem areas are suggested shich center around
>the viability and value of the most general use of writable control stores."
>
>
gelasco
2002-09-05 19:31:03 UTC
Permalink
> Ron,
>
> You took me back to when I started on the 360s in the fall of 1966. Of
> the original 360s, the 30, 40 and 50 were all microprogrammed. The 65
> an higher were hard wired. I think starting with the 80 the rest were
> all microprogrammed (but I wouldn't stand very hard on that). I
> remember hearing that the channels on the 65 and above were the same box
> as the 360/30 but with different microcode.
>

The 65 had a kind of "microcode" called ROS (Read Only Storage),
implemented thru paper tape punches.

Gilson Lasco
gelasco-***@public.gmane.org
broidoj
2002-09-05 21:37:36 UTC
Permalink
Carter,

I'm only slightly more certain than you are, but I think I can
refine that a bit (if you're really curious, and want almost 100%
reliable history, you could always tap into the awesome memory of
Shmuel Metz over on the IBM-MAIN list).

I do know for a fact, however, that the model 65 and 67 were
microprogrammed. The microcode was stored in CROS ("Capacitive Read
Only Storage") (see

http://portal.acm.org/citation.cfmid=800148.804859&coll=portal&dl=ACM

), consisting of several large, hinged frames in the front box, just
behind the lights on the right side. What you could see was
beautifully machined alloy plates, criss-crossed with evenly torqued
braces. Inside, it couldn't have been simpler. There was a large
circuit board, essentially a ground plane, about 3' by 4'. Next,
was a thin dialectric layer. Finally, there was another circuit
board, etched so that at a given point there was metal where a b'1'
was desired and substrate where a b'0' was desired. Each point in
the matrix was addressible as leads were also printed. A voltage
was applied and the capacitance was measured. If it was nearly
zero, the bit was zero.

That may seem bizarre, today, especially since the entire microcode
set of that machine can now be stored in a fraction of the capacity
of a chip the size of a pin head, but there were even more bizarre
ROS implementations; IBM tried everything! The model 30, for
example, used plastic sheets the size of a standard, 80-column
punched card, with holes equally appropriately punched. They were
inserted between metallic plates on one side and wire brushes on the
other. Where there was continuity through one of the holes, b'1'
was read.

The neatest ROS I ever encountered, however, was in the 2848, a
channel-sized box which handled 2260 terminals. The 2260, itself,
was dumb, being mainly a monitor and keyboard with little or no
logic. Indeed, even the dot matrix which made-up the terminal's
character set was stored in the 2848. The storage was a matrix of
hand-wired ferrite cores, each core easily visible and therefore
much larger than the cores in the CPU's potted, heated, 64Kb main
memory modules. This matrix was mounted behind glass in a frame
about 2' on a side which hung on the left short wall of the 2848.
If one positioned oneself at just the right angle, one could
actually make out the alphabet, for there was only a core at a
matrix intersection where a b'1' was desired.

As I recall, models 75, 91, 95, 195 and 168 were all hard-wired, at
least to a large extent. I can't recall the implementation of the
165, but I believe it was microcoded as were the lower models, of
course. As for the channels, I believe that those attachable to the
65 and 67 were entirely standard, as were the 2365 core boxes
(footprint of a large office desk and about 6' tall), each of which
cost roughly $50,000 in 1972 and contained 256Kb (that's kilobytes,
not megabytes) of 0.7 microsecond memory in four small (7" on a
side, perhaps) cubes of 64Kb each. As I mentioned above, these core
modules were potted (filled with resin) and heated to 90 degrees F.
under the control of very sensitive thermostats. This isolated the
delicate cores from the abrupt and almost unavoidable temperature
changes which could cause them to crack. In the same room, we had a
DEC PDP-10 (model KI) which had visible, unheated, unpotted and even
unprotected core arrays. When the machine was delivered, my
boss, "One Shot" Norman Wattenberger (the finest SysProg I've ever
come across, with an IQ well over 200) and I couldn't wait, so we
followed the installation instructions, cabled it up and started
playing with it over a long weekend. In the process, I got curious
and stuck my finger under the plexiglass sheet mounted on nylon
standoffs, which was a core array's only protection. Of course, I
broke at least one core and/or spiderweb-thin wire and the array had
to be replaced. To my eternal embarrassment, I never owned-up to
this to anyone other than Norman, and the DEC engineers spent hours
spinning their wheels trying to figure out what happened. Anyhow,
the point is that because the array was exposed to the elements, the
cores were constantly breaking. DEC complained that our machine
room temperature wasn't constant, so we installed a mechanical
recording thermometer/ hygrometer. We found that the temperature
varied through only four degrees F. and the humidity was very
constant and appropriate. At these tense, accusatory hardware
meetings, you should have seen the smiles on the faces of the IBM
contingent. TOPS-10 kept a statistic of memory parity errors per
hour, and it was almost always non-zero. In the entire time we had
the model 67, which was about five years, we had one (1) memory
parity error. Of course, that set off the alarm, which was a large
fire alarm bell mounted just above the CROS frames, and shut the
machine down. Our dedicated IBM CE (we had full, prime shift
coverage, if you can believe it) replaced a card and we were off and
running again in minutes. Those really were the days and I'd give
up all of the computing power at my current command to go back to
our single MIP and those heady, pioneering times to work with that
great but entirely insane bunch of people.

Regards,
Bomber

--- In hercules-390-F5Bj5G+***@public.gmane.org, CTB <***@c...> wrote:

> <snip> Of the original 360s, the 30, 40 and 50 were all
> microprogrammed. The 65 and higher were hard wired. I think
> starting with the 80 the rest were all microprogrammed (but I
> wouldn't stand very hard on that). I remember hearing that the
> channels on the 65 and above were the same box as the 360/30 but
> with different microcode.
CTB
2002-09-05 21:51:27 UTC
Permalink
Bomber,

That goes to show what age will do for your memory. I remember the
2260's well. Before the days of TSO, under MFT at least, they could be
supported under GAM (Graphics Access Method) which gave a
semi-priviledge execution level between the application and normal task
level. I wrote a very simplistic time sharing monitor to support
multiple users (the maximum was 8, but we only had at most 3 terminals).
Since it took about 1/2 sec to swap a user in and out, performance was
not too impresive.

Carter

broidoj wrote:

>Carter,
>
>I'm only slightly more certain than you are, but I think I can
>refine that a bit (if you're really curious, and want almost 100%
>reliable history, you could always tap into the awesome memory of
>Shmuel Metz over on the IBM-MAIN list).
>
>I do know for a fact, however, that the model 65 and 67 were
>microprogrammed. The microcode was stored in CROS ("Capacitive Read
>Only Storage") (see
>
>http://portal.acm.org/citation.cfmid=800148.804859&coll=portal&dl=ACM
>
>), consisting of several large, hinged frames in the front box, just
>behind the lights on the right side. What you could see was
>beautifully machined alloy plates, criss-crossed with evenly torqued
>braces. Inside, it couldn't have been simpler. There was a large
>circuit board, essentially a ground plane, about 3' by 4'. Next,
>was a thin dialectric layer. Finally, there was another circuit
>board, etched so that at a given point there was metal where a b'1'
>was desired and substrate where a b'0' was desired. Each point in
>the matrix was addressible as leads were also printed. A voltage
>was applied and the capacitance was measured. If it was nearly
>zero, the bit was zero.
>
>That may seem bizarre, today, especially since the entire microcode
>set of that machine can now be stored in a fraction of the capacity
>of a chip the size of a pin head, but there were even more bizarre
>ROS implementations; IBM tried everything! The model 30, for
>example, used plastic sheets the size of a standard, 80-column
>punched card, with holes equally appropriately punched. They were
>inserted between metallic plates on one side and wire brushes on the
>other. Where there was continuity through one of the holes, b'1'
>was read.
>
>The neatest ROS I ever encountered, however, was in the 2848, a
>channel-sized box which handled 2260 terminals. The 2260, itself,
>was dumb, being mainly a monitor and keyboard with little or no
>logic. Indeed, even the dot matrix which made-up the terminal's
>character set was stored in the 2848. The storage was a matrix of
>hand-wired ferrite cores, each core easily visible and therefore
>much larger than the cores in the CPU's potted, heated, 64Kb main
>memory modules. This matrix was mounted behind glass in a frame
>about 2' on a side which hung on the left short wall of the 2848.
>If one positioned oneself at just the right angle, one could
>actually make out the alphabet, for there was only a core at a
>matrix intersection where a b'1' was desired.
>
>As I recall, models 75, 91, 95, 195 and 168 were all hard-wired, at
>least to a large extent. I can't recall the implementation of the
>165, but I believe it was microcoded as were the lower models, of
>course. As for the channels, I believe that those attachable to the
>65 and 67 were entirely standard, as were the 2365 core boxes
>(footprint of a large office desk and about 6' tall), each of which
>cost roughly $50,000 in 1972 and contained 256Kb (that's kilobytes,
>not megabytes) of 0.7 microsecond memory in four small (7" on a
>side, perhaps) cubes of 64Kb each. As I mentioned above, these core
>modules were potted (filled with resin) and heated to 90 degrees F.
>under the control of very sensitive thermostats. This isolated the
>delicate cores from the abrupt and almost unavoidable temperature
>changes which could cause them to crack. In the same room, we had a
>DEC PDP-10 (model KI) which had visible, unheated, unpotted and even
>unprotected core arrays. When the machine was delivered, my
>boss, "One Shot" Norman Wattenberger (the finest SysProg I've ever
>come across, with an IQ well over 200) and I couldn't wait, so we
>followed the installation instructions, cabled it up and started
>playing with it over a long weekend. In the process, I got curious
>and stuck my finger under the plexiglass sheet mounted on nylon
>standoffs, which was a core array's only protection. Of course, I
>broke at least one core and/or spiderweb-thin wire and the array had
>to be replaced. To my eternal embarrassment, I never owned-up to
>this to anyone other than Norman, and the DEC engineers spent hours
>spinning their wheels trying to figure out what happened. Anyhow,
>the point is that because the array was exposed to the elements, the
>cores were constantly breaking. DEC complained that our machine
>room temperature wasn't constant, so we installed a mechanical
>recording thermometer/ hygrometer. We found that the temperature
>varied through only four degrees F. and the humidity was very
>constant and appropriate. At these tense, accusatory hardware
>meetings, you should have seen the smiles on the faces of the IBM
>contingent. TOPS-10 kept a statistic of memory parity errors per
>hour, and it was almost always non-zero. In the entire time we had
>the model 67, which was about five years, we had one (1) memory
>parity error. Of course, that set off the alarm, which was a large
>fire alarm bell mounted just above the CROS frames, and shut the
>machine down. Our dedicated IBM CE (we had full, prime shift
>coverage, if you can believe it) replaced a card and we were off and
>running again in minutes. Those really were the days and I'd give
>up all of the computing power at my current command to go back to
>our single MIP and those heady, pioneering times to work with that
>great but entirely insane bunch of people.
>
>Regards,
>Bomber
>
>
>
Fish
2002-09-05 23:17:42 UTC
Permalink
<snip>

> [...] Those really were the days and I'd give up all
> of the computing power at my current command to go back
> to our single MIP and those heady, pioneering times to
> work with that great but entirely insane bunch of people.
>
> Regards,
> Bomber

(sigh!)

</me: eyes begin to glaze over as I too fondly reminisce about "The
Good Old Days"...>

=)

Yeah, man, dems really WERE "Da Good Old Days!" weren't they? :)

(I don't know about others but I personally enjoy the shit outta
hearing such stories! Perhaps we should start a new group??)
- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org
Bona fide mainframe "Old Fart"
broidoj
2002-09-05 20:44:18 UTC
Permalink
Ron,

Oh, I'm interested, big time! This interest was sparked when,
working at my first mainframe job in 1972, I stumbled across the
microcode listing for my employer's 360/67. The code was written in
100-bit so-called micro-orders, made-up mostly of bit switches and
various address fields. Everything was emulated, from the PSW to
the GPRs and, of course, all of the instructions. I discovered that
the main function of the Diagnose instruction on that machine was to
force a branch to a microcode address. I found the micro-order loop
which was associated with the STOP button. When the machine was in
manual, this disabled loop ran and checked for one thing: a
depression of the START button. I wrote a program on April 1, 1973
which used X'83' to branch to this address and, of course, stop the
machine. My buddies and I ran into the machine room and started
hollering at the lead operator, accusing her of bumping the button
with her elbow (nearly impossible as the 1052 was in the way). She
sheepishly apologized, I hit the green START button with an
exasperated exression on my face, and ran back into the Systems Hole
to run it again. After the third time, we could no longer stifle
our hysteria and spilled the beans. Such was life
of "professionals" at a university.

As for the microcode listing, I photocopied it and still have it
floating around in the attic somewhere.

Regards,
Bomber
"Old Stuff R Us"

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Ronald Tatum" <***@d...> wrote:

> I would, however, be willing to scan the bloody thing and use
> Rob Storey's neat program to convert the images to (non-
> editable) .PDF format if you'd like a copy for yourself or to post
> to the files area if anyone's interested in such "old stuff".
Fish
2002-09-05 22:25:25 UTC
Permalink
> Ron,

<snip amusing 'stop button microcode' story>

(LOL!!) I *love* it! :`D

I rather enjoy hearing about these types of old-time pranks.

Anyone else have one to share? :)

</me: stands back for sudden flood of stories>
- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org
broidoj
2002-09-06 01:17:38 UTC
Permalink
--- In hercules-390-F5Bj5G+***@public.gmane.org, "Fish" <***@i...> wrote:

> I rather enjoy hearing about these types of old-time pranks.

I have many more Norman stories, if you like. Some are programming
oriented, some elevator oriented, car (Lamborghini Miura, Maserati
Bora, DeThomaso Pantera), motorcycle, you name it.
Gregg C Levine
2002-09-06 01:28:11 UTC
Permalink
Hello from Gregg C Levine
I'd love to see the computer related ones. I mean, I happen to know that
geniuses have a different sense of humor then normal people. It shows in
how Linux behaves for example. In fact, can everyone who has a story of
that nature provide a few? I'd especially like to see the microcode for
that practical joke.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: broidoj [mailto:broidoj-***@public.gmane.org]
> Sent: Thursday, September 05, 2002 9:18 PM
> To: hercules-390-***@public.gmane.org
> Subject: [hercules-390] Re: Old-timer's pranks (was: An interesting
idea another
> one?)
>
> --- In hercules-390-F5Bj5G+***@public.gmane.org, "Fish" <***@i...> wrote:
>
> > I rather enjoy hearing about these types of old-time pranks.
>
> I have many more Norman stories, if you like. Some are programming
> oriented, some elevator oriented, car (Lamborghini Miura, Maserati
> Bora, DeThomaso Pantera), motorcycle, you name it.
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
broidoj
2002-09-06 01:40:29 UTC
Permalink
Gregg,

I'm too tired right now, but in the near I will tell one Norman story
at a time until the off topic police come after me. As for the joke,
there was no microcode. We tried, but couldn't figure out how to get
it to run our own microcode stored in Main memory. All we did was
use X'83', Diagnose, to branch to the micro-order address (in ROS,
not Main) which was the start of the manual state disabled loop.
Now, to bed!

Jeff

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Gregg C Levine" <***@w...> wrote:
> Hello from Gregg C Levine
> I'd love to see the computer related ones. I mean, I happen to know
that
> geniuses have a different sense of humor then normal people. It
shows in
> how Linux behaves for example. In fact, can everyone who has a
story of
> that nature provide a few? I'd especially like to see the microcode
for
> that practical joke.
jeffsavit
2002-09-06 19:48:36 UTC
Permalink
>From the 'net:

See "Computer Releated Horror Stories, Folklore and Anecdotes"
http://www.antioffline.com/HUM/computer.folklore.from.net.rumors.html
and "The Size Stages of Field Services"
http://nemesis.lonestar.org/stories/stages.html

I broke up at this one:
Why do you put a message in the system that reports the equivalent
of: "Hi there!! I just did something stupid and committed suicide!
I will crash in a few moments and lose everybody's work. In the
meantime, here is some music to listen to!" BEEP BEEP BEEP ..."

--- In hercules-390-F5Bj5G+***@public.gmane.org, "broidoj" <***@g...> wrote:
> Gregg,
>
> I'm too tired right now, but in the near I will tell one Norman story
> at a time until the off topic police come after me. As for the joke,
> there was no microcode. We tried, but couldn't figure out how to get
> it to run our own microcode stored in Main memory. All we did was
> use X'83', Diagnose, to branch to the micro-order address (in ROS,
> not Main) which was the start of the manual state disabled loop.
> Now, to bed!
>
> Jeff
>
> --- In hercules-390-F5Bj5G+***@public.gmane.org, "Gregg C Levine" <***@w...> wrote:
> > Hello from Gregg C Levine
> > I'd love to see the computer related ones. I mean, I happen to know
> that
> > geniuses have a different sense of humor then normal people. It
> shows in
> > how Linux behaves for example. In fact, can everyone who has a
> story of
> > that nature provide a few? I'd especially like to see the microcode
> for
> > that practical joke.
Greg Smith
2002-09-07 04:29:43 UTC
Permalink
I went to a IBM programming class when ESA was new. This
was v3 when access registers, data spaces, hiper spaces,
and linear datasets were introduced.

This was a lab type of class, and naturally, I sat in the
very back of the room. We were reviewing how to do DIV
(data in virtual) on linear datasets in HLLs (hi-level-
languages) then had a lab exercise on the same.

The execercise was to code a div call to read a certain
page in the linear dataset using an hll, and then pass
the data we read to a subroutine that the instructor had
written. It calculated a checksum of the data passed to
it and if the return value matched a certain number, then
you successfully completed the exercise.

Most everyone started programming the exercise in cobol.
My lab partner & I programmed it in pl/i and successfully
completed the exercise in about a minute.

So, I went exploring. Somehow, I managed to bring up
an os/ditto ispf screen. I wondered `did they update
ditto to work on linear datasets ?' So I brought up
the linear dataset that we were using for the class exercise.
I quickly discovered that I could zap the portion of the
dataset and make everyone's program fail.

We had the instructor running around like crazy. I was zapping
the dataset so that sometimes the students' programs would
fail then sometimes work. Unfotunately, my lab partner gave
me away by literally rotfl.

A few years later when openedition was new, I went to a
Share hands-on lab. In a typical fubar, it turned out that
the lab was scheduled after they had packed up the es-9000
they brought for the occassion. But IBM worked out a kludge
where you could telnet from an aix workstation to somewhere
in ibm-land then tn3270 to a mvs v4 mainframe.

Again, in the back of the room, I got signed on to the remote
mainframe. The tso logon proc dumped you right into omvs.
I ran a few demos then tried `tso isfisp' and found myself
in ispf sdsf. Tried a `/d t' and it worked !! (ie I could
issue operator commands).

A number of IBMers were walking around the room checking
out how the lab was going. One came by me and I showed
him what I could do. He quickly called a couple of others
over. All we had to do was pick the deed.

All the lab computers had a sign over top of them specifying
the userid you were to use for that station. Over across the
room was a very pretty young woman that was having a helluva
time just getting connected and was being assisted by at least
a 1/2 dozen helpful IBMers.

We watched. After a couple of minutes she got the tso logon
prompt. After the she got the logon in progress message all
the helpful IBMers cheered. That's when I let loose with
the `/c u=xxx' command ;-) `No No come back come back'

Someone in the crowd that had gathered around me snickered
too loud and one of the helpful IBMers noticed and gave us
a glare that woulda frozen water ;-)

Moral is: Don't trust the guy in the back of the classroom

Greg
halfmeg
2002-09-07 16:07:07 UTC
Permalink
> Greg Smith <gsmith-***@public.gmane.org> wrote:

><snip>

> Moral is: Don't trust the guy in the back of the classroom


Hmm, this reminded me of a class in SCO UNIX for the Air Force which
I attended. The instructor was telling us how the passwords were
encrypted and that we couldn't decipher them from the gibberish which
we saw displayed when we looked at the data. Sure enough, it was
gibberish from what I had used for my password. A slight problem
developed for the instructor however after lunch was over, she could
not sign back into the system. Her password had been changed. Not
by any deciphering scheme or a good guess. It had simply been
replaced with the encrypted string that I copied from my password.

They changed the access rights of the file after this.

Phil - hey, I was a newbie to UNIX just seeing what it would do :)
broidoj
2002-09-07 17:58:40 UTC
Permalink
Good ones! Keep 'em coming!

--- In hercules-390-F5Bj5G+***@public.gmane.org, Greg Smith <***@n...> wrote:

> Moral is: Don't trust the guy in the back of the classroom
S. Vetter
2002-09-07 18:15:40 UTC
Permalink
Being a nubie to IBM mainframes many years back, I ended up with one of
those core reports. I found two bugs in the program that I wrote. But one I
just had to turn into my boss who was a Systems type person. I went to his
office and told him about the problem. (I should point out at this point in
time we printed everything and used recycled paper) He looked through the
core dump when I stoped him after a minute. I said let's look at the program
listing and we should find the bug there. I had him flip to some particular
page and there it was. A spider had been impressed neatly into the paper in
the middle of the program listing! Thats where the program bug was found...

Scott
Fish
2002-09-07 23:33:51 UTC
Permalink
(Heh!) Good one, Greg. :)

I got a short one for you.

A friend of mine where I worked was good friends the VM System
Administrator and got into this good natured disagreement with
another fellow coworker over who was better at
<something-or-other>[1] (each of course claiming THEY were better at
it than the other):

(My friend): "You KNOW I'm better at it than you."
(Coworker): "Are not! *I'm* better and everyone knows it!"
(My friend): "Nope. *I'm* better. And *some* day you're going to
admit it to the world. <wiggles eyebrows mischievously>"
(Coworker): "NEVER!"

So later that evening after person X leaves to go home, my friend
visits his VM System Admin friend and explains to him what he wants
to do. The VM Admin agrees to help and together they make a slight
modification to his CMS profile exec.

The next day person X arrives at work and logs onto VM/CMS and is
greeted with a nice "Good morning, Person X! To logon, please answer
correctly the following question: Who's the absolute best at
<something-or-other>? <So-and-so> (my friend's name) or you?"

Of course if he tries answering anything OTHER than my friend's name
he is immediately logged off. :)

He immediately went up to my friend and asked him to fix his
shenanigans but my friend of course refused and suggested he go talk
to the VM Admin. So he went to the VM Admin who of course *also*
refused (basically telling him to just answer the question and
correctly and then he could logon and fix it for himself, which
eventually, after grumbling, he finally does).

But it was only after he finally logged on and went to fix it that he
saw that part of this specially crafted profile exec was to also send
every logged on user a personal message from him stating that he had
finally admitted that my friend was indeed MUCH better at
<such-and-such> than him (thereby explicitly "proclaiming to the
world" that which he said he'd never do).

>;-)

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org

[1] I forget what the disagreement was actually over. Maybe softball
or something? Anyway it's not really that important as far as the
telling of the story goes.


> -----Original Message-----
> From: Greg Smith [mailto:gsmith-***@public.gmane.org]
> Sent: Friday, September 06, 2002 9:30 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] Re: Old-timer's pranks
>
>
> I went to a IBM programming class when ESA was new. This
> was v3 when access registers, data spaces, hiper spaces,
> and linear datasets were introduced.

<snip>
John Alvord
2002-09-07 23:44:02 UTC
Permalink
On Sat, 7 Sep 2002 16:33:51 -0700, "Fish" <fish-6N/dkqvhA+***@public.gmane.org> wrote:

>
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>(Heh!) Good one, Greg. :)
>
>I got a short one for you.
>
>A friend of mine where I worked was good friends the VM System
>Administrator and got into this good natured disagreement with
>another fellow coworker over who was better at
><something-or-other>[1] (each of course claiming THEY were better at
>it than the other):
>
>(My friend): "You KNOW I'm better at it than you."
>(Coworker): "Are not! *I'm* better and everyone knows it!"
>(My friend): "Nope. *I'm* better. And *some* day you're going to
>admit it to the world. <wiggles eyebrows mischievously>"
>(Coworker): "NEVER!"
>
>So later that evening after person X leaves to go home, my friend
>visits his VM System Admin friend and explains to him what he wants
>to do. The VM Admin agrees to help and together they make a slight
>modification to his CMS profile exec.
>
>The next day person X arrives at work and logs onto VM/CMS and is
>greeted with a nice "Good morning, Person X! To logon, please answer
>correctly the following question: Who's the absolute best at
><something-or-other>? <So-and-so> (my friend's name) or you?"
>
>Of course if he tries answering anything OTHER than my friend's name
>he is immediately logged off. :)
>
>He immediately went up to my friend and asked him to fix his
>shenanigans but my friend of course refused and suggested he go talk
>to the VM Admin. So he went to the VM Admin who of course *also*
>refused (basically telling him to just answer the question and
>correctly and then he could logon and fix it for himself, which
>eventually, after grumbling, he finally does).
>
>But it was only after he finally logged on and went to fix it that he
>saw that part of this specially crafted profile exec was to also send
>every logged on user a personal message from him stating that he had
>finally admitted that my friend was indeed MUCH better at
><such-and-such> than him (thereby explicitly "proclaiming to the
>world" that which he said he'd never do).

IPL CMS (NOPROF

is a good way around those games.

My personal favorite was a trick played on me by Ted van Duyn while I
was at Amdahl. I was known as a hot shot in VM (wrote several
products, solved impossible customer problems).

One day Ted invited me into his office, sat me down at his terminal
and said "What's wrong?". The screen was blank. I typed in some
characters... nothing and the screen remained blank. After 20 minutes
of struggle, no change.

He had found some sort of translate function that was translating all
output to blanks. And he was running a program that watched for a
certain character sequence like "ASDFG" which would then revert the
translate function. On all other input it would ignore the input and
continue.

I bought him lunch.

john alvord
Fish
2002-09-07 23:54:57 UTC
Permalink
> IPL CMS (NOPROF
>
> is a good way around those games.

It's been a while, but wouldn't a directory entry with AUTOCR prevent
one from being able to do that?

(And I can't recall anymore how they did it (I don't remember the
details), but I believe they managed to come up with some way to
prevent him from simply dropping into CP too. I.e. "#CP IPL CMS"
didn't work either. NO CP commands would work as I recall. Don't
remember for sure or how they did it, but I *think* they managed to
accomplisht that too 'cause anyone even remotely familiar with CP/CMS
would be smart enough to try #CP IPL CMS.)

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org
malvert5125
2002-09-08 09:40:11 UTC
Permalink
Hi,

Yes. When you log on to VM you will get a VMREAD at the time you will
IPL CMS. When you have the statement PARM AUTOCR in your directory you
won't get the VMREAD and your CMS is IPL-led immediatly. The SYSPROF
EXEC will setup your CMS environment and call the PROFILE EXEC on the
A-disk if it exists. (This is for VM/ESA, the older systems use a
DMK-member to do the stuff the SYSPROF EXEC will do. The CP has to be
generated again to get the changes effective.)

However, when you do an IPL CMS when you are already logged on, you
will get a VMREAD and then you can do an ACCESS 191 A (NOPROF to IPL
your CMS without running the profile exec. Even a NOSPROF is
possible to prevent the SYSPROF EXEC to be executed. The PARM AUTOCR
statement is not in effect now, since you are already logged on.

It should be possible to trap anything you can do on your CMS I guess,
including IPL and CP commands. There would be some ways to do it, but
I never had any reason to try it myself, so I couldn't tell for sure.

Regards, Berry.

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Fish" <***@i...> wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> > IPL CMS (NOPROF
> >
> > is a good way around those games.
>
> It's been a while, but wouldn't a directory entry with AUTOCR
prevent
> one from being able to do that?
>
> (And I can't recall anymore how they did it (I don't remember the
> details), but I believe they managed to come up with some way to
> prevent him from simply dropping into CP too. I.e. "#CP IPL CMS"
> didn't work either. NO CP commands would work as I recall. Don't
> remember for sure or how they did it, but I *think* they managed to
> accomplisht that too 'cause anyone even remotely familiar with
CP/CMS
> would be smart enough to try #CP IPL CMS.)
>
> - --
> "Fish" (David B. Trout)
> fish-***@public.gmane.org
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXqRyEj11/TE7j4qEQKI6QCfe9zl+zOsqAm9MPRqe+3sCRRHF6QAoI5H
> od883eGOMpFoIDfQ+EA8xuYB
> =rrBA
> -----END PGP SIGNATURE-----
mvandere
2002-09-09 04:44:58 UTC
Permalink
Yup,

Did this once..... execs to cover every possible commmand (and feign
the appropriate responses). Even wrote an exec to update all the
users execs to insert 'exec' in front of every possible command....
Not to mention an Xedit macro called 'save' et al, to stop them
removing the 'exec', or writing new ones, didn't quite stoop to
hiding the 'exec' when they started editing.

You kind of give the game away when you stop PA1 coming up 'CP READ',
but never mind. Still quite good fun and sure shows the true power of
VM.

Cheers,

Mark

--- In hercules-390-F5Bj5G+***@public.gmane.org, "malvert5125" <***@h...> wrote:
> Hi,
>
> Yes. When you log on to VM you will get a VMREAD at the time you
will
> IPL CMS. When you have the statement PARM AUTOCR in your directory
you
> won't get the VMREAD and your CMS is IPL-led immediatly. The
SYSPROF
> EXEC will setup your CMS environment and call the PROFILE EXEC on
the
> A-disk if it exists. (This is for VM/ESA, the older systems use a
> DMK-member to do the stuff the SYSPROF EXEC will do. The CP has to
be
> generated again to get the changes effective.)
>
> However, when you do an IPL CMS when you are already logged on, you
> will get a VMREAD and then you can do an ACCESS 191 A (NOPROF to
IPL
> your CMS without running the profile exec. Even a NOSPROF is
> possible to prevent the SYSPROF EXEC to be executed. The PARM
AUTOCR
> statement is not in effect now, since you are already logged on.
>
> It should be possible to trap anything you can do on your CMS I
guess,
> including IPL and CP commands. There would be some ways to do it,
but
> I never had any reason to try it myself, so I couldn't tell for
sure.
>
> Regards, Berry.
>
> --- In hercules-390-F5Bj5G+***@public.gmane.org, "Fish" <***@i...> wrote:
> >
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > > IPL CMS (NOPROF
> > >
> > > is a good way around those games.
> >
> > It's been a while, but wouldn't a directory entry with AUTOCR
> prevent
> > one from being able to do that?
> >
> > (And I can't recall anymore how they did it (I don't remember the
> > details), but I believe they managed to come up with some way to
> > prevent him from simply dropping into CP too. I.e. "#CP IPL CMS"
> > didn't work either. NO CP commands would work as I recall. Don't
> > remember for sure or how they did it, but I *think* they managed
to
> > accomplisht that too 'cause anyone even remotely familiar with
> CP/CMS
> > would be smart enough to try #CP IPL CMS.)
> >
> > - --
> > "Fish" (David B. Trout)
> > fish-***@public.gmane.org
> >
> > -----BEGIN PGP SIGNATURE-----
> > Version: PGP 7.0.4
> >
> > iQA/AwUBPXqRyEj11/TE7j4qEQKI6QCfe9zl+zOsqAm9MPRqe+3sCRRHF6QAoI5H
> > od883eGOMpFoIDfQ+EA8xuYB
> > =rrBA
> > -----END PGP SIGNATURE-----
barahul
2002-09-08 04:08:30 UTC
Permalink
It probably would not qualify as a prank as such, but...

During my university years I developed a server that needed to send
large chunks of data to another service machine. Given various side
requirements, using a virtual printer was the best way to handle this
(due to record length issues and wanting to avoid yucky things like
NETDATA format). Unfortunately, the largest printer that could be
defined on the system (VM/SP or VM/XA... this was 1993) only allowed
for 204 characters per line, and I needed at least 255, and preferably
a lot more.

Then my eye fell on CMS PIPELINES and the PRINTMC stage. Using that
stage it was possible to write much longer records to the virtual
printer. In order to test both functionality and performance, I wrote
a little EXEC that created a 65537 character string that got piped
through the printmc stage.

The EXEC never completed because the VM system decided to crash at
that moment. The sysadmin informed me that there was a hardware
failure on one of the CPUs, and after the lengthy re-IPL (a 3090-600E)
I proceeded with my testing. Again, the mainframe went down hard as
my EXEC was running.

Showing the little EXEC to the sysadmin resulted in: "No worries, that
could not have caused it." followed by "Oh well, we can test that
easily." As soon as the next re-IPL was done, I logged in on the
first available 3270, ran the EXEC and it completed without a glitch.

About 15 seconds later, alarms went off and the machine was on its
knees again.

When the university reported the issue to IBM (tagged urgent since the
site was the central BITNET node for Belgium at that time), IBM's
response was along the lines of: "Who on earth would try something
like that???"

They did have a fix for it in a matter of 1-2 days though.

A couple of months later a friendly sysadmin at a university in Israel
learnt about the little issue I found and decided to see if their VM
already included that fix or not. About 30 minutes later he reported
it apparently did not...

Kris
madhackbt
2002-09-08 05:43:50 UTC
Permalink
Delurking for this thread...

Two things to report on this one. The first being my witnessing the
first ever voice activated 3278 terminal. The manager, for whom this
little joke was set up for, absolutley could not believe his eyes.
There was the systems programmer demonstrating the device; <In a loud
voice> "Backspace, A..., B..., 1..., no, not 2, 1...., Backspace...
Backspace....". Of course, the device had a few bugs, like the other
systems programmer who could not hear what the "voice activation
unit" was saying as he tried to type in using the keyboard that had
been run to his cubicle from the 3278 where it was being demonstrated.

The second, was from my the person who trained me in RACF Data
Security (What I am by training). We had a mini-RACF command war,
where I had revoked his ID, and then cancelled his TSO session.
Thinking I had won the war, I was surprised that he was pounding the
keys on his terminal and chuckling to himself. I could not
understand or figure how he got on the system. Well, in a few
minutes, I got a call from a console operator asking me if I wanted
my ID to be revoked... you know the WTO to the console for SPECIAL
userids :) The guy did not even have to sign on to a terminal to
make my life difficult. Taught me a valuable lesson!

Bill
Jim Keohane
2002-09-08 16:41:22 UTC
Permalink
Merrill Lynch scheduled an entire weekend for the wholesale conversion
from 7074 to 360. In addition to the 7074 merging and combining tapes into a
single string of 3 - 4 dozen tapes, the 2 360/65's were also running 7074
emulation. As soon as a tape came off the 7074 it was processed on the 1st
360/65 whose output tapes were then fed to next 360/65.

Process had been running trouble free for over 30 hours. Perhaps 80 - 90
% done when IBM CE's said to put 1st 360/65 into STOP as a precaution during
some other maintenance. CE's estimated 30 minutes.

Vic Stannish, project overseer, was in machine room but unaware of CE
maintenance. The o/p tapes on 1st 360/65 were mounted on drives 364 and 366.
Tape 365 was offline for some maintenance issue.

I opened the vacuum columns on 365 and dumped in a few dozen feet of
magnetic tape from the trash. I then mounted an "official" looking tape on
360/65. I then put tape in READY but switched READY light with CB FUSE so it
looked like drive went kablooey! Cigar-smoking Vic came over and saw
everything was halted. I looked as despondent as I could while pointing out
the ruined tape. It all had to be redone from start!

He almost swallowed his cigar. When it was explained as a joke to Vic he
then called down his boss, Ross Clemens, and pulled the joke on him.

CE's gave the go-ahead and processing resumed to finish successfully.

- Jim Keohane

<Signature>
<Name>Jim Keohane</Name>
<Quote>You must be certifiable to even think of using PKI</Quote>
</Signature>
broidoj
2002-09-06 11:47:53 UTC
Permalink
OK, at least two of you asked for it.

When I met him in 1972, Norman was a contractor. He was hired by
the university for one project at a time and paid a per dieum which
was, in those days, astronomical; if I recall correctly, it was
something like $300/day. At the time, I was making a salary of
$10,994 per year, which worked out to about $42.28 a day. Four
years later, I'd become the Senior Sysprog, but was still only
making $75 a day. You get the idea: Norman was very, very well
respected. The trouble was, the university didn't like to pay its
bills on time. Most often, he'd invoice them and, six months later,
he'd still be hassling them for his check.

Eventually, we replaced our 360/67 with a 370/158, selling the
former to another university, RPI. A few months later, we added a
370/168, finally beginning to fill-up our immense machine room. The
MIS administration investigated ways to couple the two machines and
were on the verge of trying to install Boeing's shared spool code
for HASP II when we were warned that the added code was very
sensitive. We'd just recovered from an embarrassingly long six-
crash-a-day period, so they decided to hire Norman to write
something from scratch.

The only trouble was that he'd already told them that he'd never
perform another contract there due to the administration's
reluctance to pay his bills within 30 days. They pushed, he
resisted and, finally, after weeks of this fruitless negotiation, he
told them he'd do it for $20,000 just to get them off his back. To
everyone's surprise, especially Norman's, they accepted and cut him
a check for the full amount in advance. The first thing he did was
take the money and the old 1968 Citroën DS-19 I'd sold him when I
bought my used 1972 Citroën SM (I still have it!) and trade every
penny, nut and bolt for a brand new Maserati Bora.

He wrote the code in one weekend and installed and debugged it the
next weekend. That works out to $5,000 per day, and though I've no
doubt that it's not a world record, I've never heard of a SysProg
getting paid anywhere near that amount, even at the compensation
peak in this country just before Y2K. Shortly after that, I was
appointed acting Systems Programming Manager as my boss was AWOL.
When he finally returned from his three-week vacation after six
months, and they realized he didn't have a good excuse, they offered
me his job. I'd had enough management, and convinced them to hire
Norman instead, which they did. He was a lousy manager, but the
best I've ever worked for. I know that sounds contradictory, but
perhaps you'll get the idea...

I'm having trouble remembering all the details of his shared spool
implementation, but it was elegant, simple and so efficient that the
overhead was virtually undetectable. As far as I know, his code
never caused HASP (later retrofitted to early JES2) to crash, and
the code ran there for years. Both systems just ran as normal, but
one could submit a job from either system to run on the other and
output could be retrieved from either as well. He used a pseudo
device he called RDR20000 to send jobs from one system to the
other. He was questioned often as to the meaning of the name but
always smiled and quietly stated that it was meaningless. The
astute reader will surely figure it out.

More Norman stories, anyone? How about the two elevator shuffle?

Regards,
Bomber

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Gregg C Levine" wrote:

> I'd love to see the computer related ones. I mean, I happen to
> know that geniuses have a different sense of humor then normal
> people. It shows in how Linux behaves for example. In fact, can
> everyone who has a story of that nature provide a few?
Fish
2002-09-05 22:00:44 UTC
Permalink
> Fish,
> Your post jogged my memory a bit; sure enough, there was a paper
> in the December 1969 Computing Surveys about microprogramming. I
> found the pub in my basement stash.
> Here's the abstract:
> "Microprogramming is introduced as a vehicle for implementing the
> control function of a computer. Some contemporary systems using
> this technique are described and several justifications are given
> for its application. Finally a set of specific research problem
> areas are suggested shich center around the viability and value of
> the most general use of writable control stores."

Sounds interesting!

> The paper is sixteen pages, including the one-page bibliography.
> No, I *won't* transcribe the paper for you, even though you have
> accused me of being overly prolix :-).

(heh!) When I did, I hope I also mentioned (and if I haven't I shall
do so now) that I readily admit that I happen to be guilty of the
same thing from time to time -- sometimes more often than not![1] :)

> I would, however, be willing to scan the bloody
> thing and use Rob Storey's neat program to convert the images to
> (non-editable) .PDF format if you'd like a copy for yourself or to
> post to the files area if anyone's interested in such "old stuff".

Please! Thanks. :)

> Regards,
> Ron Tatum

Regarding your propensity to make wordy posts: I wouldn't mind it so
much, Ron, if you would simply stick a blank line here and there
(say, in between your paragraphs for instance?).

Doing so makes one's posts MUCH easier to read.[2] :)

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org

[1] It seems to come and go with me. Sometimes I ramble, other times
I don't [too much]. (I do know I'm more likely to do it the tireder
(more tired? "tireder *should* be a word IMO!) I get, and given that
I suffer from a sleeping disorder and am almost *always* exhausted, I
tend to ramble more often than not. (Like now for example! <sigh!>)

[2] And much more likely to BE read! Personally, if I find reading a
person's post to be "too much work" for me (due to its poor
formatting for example), I won't even bother reading it. No matter
how good (valuable) the subject matter may be, if it's poorly
formatted and too cumbersome to read, I won't even bother reading it.
Sorry, but on high volume lists, that's just the way I am. <shrug>
w***@public.gmane.org
2002-08-27 14:49:16 UTC
Permalink
Neat, have a picture of that keychain?

-----Original Message-----
From: Jeffrey C Barnard [mailto:bsiopti-***@public.gmane.org]
Sent: Monday, August 26, 2002 9:14 PM
To: hercules-390-***@public.gmane.org
Subject: Re: [hercules-390] An interesting idea another one?

S/390's and the new z800/z900 processors are not PowerPC processors. I
got to
talk to one of the designers once and he indicated that S/390's are part
hardwired, part microcoded. He said they designed their own chips. I
actually
have one (IBM gave some out as key chains - not sure why I got one as
they did
not have many to give out).

Regards,
Jeff

Paul Raulerson wrote:
>
> Not exactly - zSeries, iSeries, and pSeries machines (i.e. Mainframes,
AS/400, and RS/6000 machines)
> all use the same PowerPC processor. Only the RS/6000 uses it natively
though; the 400 and mainframes have software / microcode
> layers that emulate a different processor. The S/390 Architecture is
implemented primarily via Microcode these days.
>
> -Paul


--
Jeffrey C Barnard
Barnard Software, Inc. http://www.bsiopti.com
Phone 407-323-4773 Fax 407-323-4775



Community email addresses:
Post message: hercules-390-***@public.gmane.org
Subscribe: hercules-390-subscribe-***@public.gmane.org
Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
List owner: hercules-390-owner-***@public.gmane.org

Files and archives at:
http://groups.yahoo.com/group/hercules-390

Get the latest version of Hercules from:
http://www.conmicro.cx/hercules

Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
Jeffrey C Barnard
2002-08-27 18:04:06 UTC
Permalink
Sorry, do not have a way to post a picture.

The keychain says

IBM
ESA/390 CMOS Microprocessor
25 Million Transistors
Poughkeepsie & Yorktown Labs, USA

The chip is embedded in clear plastic above the writing.

Jeff

warlockd-+***@public.gmane.org wrote:
>
> Neat, have a picture of that keychain?

--
Jeffrey C Barnard
Barnard Software, Inc. http://www.bsiopti.com
Phone 407-323-4773 Fax 407-323-4775
Gregg C Levine
2002-08-27 18:11:10 UTC
Permalink
Hello again from Gregg C Levine
Do you have a digital camera Jeffrey? Take one, and then upload it to
the photos section of the files area. Supposedly it's a relatively quick
and easy affair. If you don't, then take a regular picture, and scan it,
and then repeat the first steps.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: Jeffrey C Barnard [mailto:bsiopti-***@public.gmane.org]
> Sent: Tuesday, August 27, 2002 2:04 PM
> To: hercules-390-***@public.gmane.org
> Subject: Re: [hercules-390] An interesting idea another one?
>
> Sorry, do not have a way to post a picture.
>
> The keychain says
>
> IBM
> ESA/390 CMOS Microprocessor
> 25 Million Transistors
> Poughkeepsie & Yorktown Labs, USA
>
> The chip is embedded in clear plastic above the writing.
>
> Jeff
>
> warlockd-+***@public.gmane.org wrote:
> >
> > Neat, have a picture of that keychain?
>
> --
> Jeffrey C Barnard
> Barnard Software, Inc. http://www.bsiopti.com
> Phone 407-323-4773 Fax 407-323-4775
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
Michael Short
2002-09-04 19:28:01 UTC
Permalink
I good description of the z900 can be found in the redbook IBM eServer
zSeries 900 Technical Guide, SG24-5975-00 which can be found at
www.redbooks.ibm.com . From the description I would say it is done with
custom chips, although I believe the direction is to use a RISC engine with
microcode similar to the AS-400.

>From the book for 10way to a 16way machine:

on a 5" x 5" MultiChip Module

20 PU chips of which 16 can be CPUs (47 million transtistors in each);
three of the others are service processors and one is a spare
2 Storage Control chips
8 4MB chips for level 2 cache
4 Memory Bus Adapters for 24-1GB bidirectional I/O attachments
1 Clock chip

2.5 billion transistors
101 glass ceramic layers
6 thin film layers with a km of wiring
4224 pins





To: hercules-390-***@public.gmane.org
Joe Zitzelberger cc: (bcc: Michael Short/Towers Perrin)
<psychedelic-***@mi Subject: Re: [hercules-390] Re: An interesting idea another one?
ndless.com>

09/04/2002 08:57 AM
Please respond to
hercules-390







On Tuesday, September 3, 2002, at 10:10 PM, scott_blackledge wrote:
> Paul,
>
> You'd owe the beers. I work (for another week) about three
> rows down from the HW designers. The z900 chips are custom. Jeff
> is correct.
>
> Scott

That article makes the following very vague, but important statement:

"To simplify the logic design, the processor uses millicode [2] (a form
of Licensed Internal Code), which is the vertical microcode that
executes on this family of processors to implement many complex
elements of the architecture. Millicode consists of all of the
hardwired ESA/390 instructions plus 102 hardwired assist instructions
that can only be executed by millicode."


The question becomes 'What is executing the millicode'? IIRC Power
was, except in its low end consumer and embedded versions, was a set of
6 or 9 chips wired up to play well with each other. Did they borrow
some subset of those?


psychedelic-harry-9q/xBM6aKHVWk0Htik3J/***@public.gmane.org

Web Servers Do It With Cookies



Community email addresses:
Post message: hercules-390-***@public.gmane.org
Subscribe: hercules-390-subscribe-***@public.gmane.org
Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
List owner: hercules-390-owner-***@public.gmane.org

Files and archives at:
http://groups.yahoo.com/group/hercules-390

Get the latest version of Hercules from:
http://www.conmicro.cx/hercules

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Paul A. Scott
2002-09-06 16:00:31 UTC
Permalink
I know exactly what you mean. Like those annoying, psychotic foot notes. I
just move on to the next post.

Paul

> Regarding your propensity to make wordy posts: I wouldn't mind it so
> much, Ron, if you would simply stick a blank line here and there
> (say, in between your paragraphs for instance?).
>
> Doing so makes one's posts MUCH easier to read.[2] :)
>
> - --
> "Fish" (David B. Trout)
> fish-6N/dkqvhA+***@public.gmane.org
>
> [1] It seems to come and go with me. Sometimes I ramble, other times
> I don't [too much]. (I do know I'm more likely to do it the tireder
> (more tired? "tireder *should* be a word IMO!) I get, and given that
> I suffer from a sleeping disorder and am almost *always* exhausted, I
> tend to ramble more often than not. (Like now for example! <sigh!>)
>
> [2] And much more likely to BE read! Personally, if I find reading a
> person's post to be "too much work" for me (due to its poor
> formatting for example), I won't even bother reading it. No matter
> how good (valuable) the subject matter may be, if it's poorly
> formatted and too cumbersome to read, I won't even bother reading it.
> Sorry, but on high volume lists, that's just the way I am. <shrug>
Gregg C Levine
2002-09-06 16:06:54 UTC
Permalink
Hello from Gregg C Levine
Indeed! I've been getting signals from my brother of all people who want
me to do the same thing. It won't work. I sometimes type the way I talk.
Fish, Paul, can we agree on one thing? Hercules. Its why we are all
here. And I like Fish's footnotes, they make his posts more interesting,
and imply that he's spending more time writing them, then usual three to
five minutes that we spend.
-------------------
Gregg C Levine hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org
------------------------------------------------------------
"The Force will be with you...Always." Obi-Wan Kenobi
"Use the Force, Luke."  Obi-Wan Kenobi
(This company dedicates this E-Mail to General Obi-Wan Kenobi )
(This company dedicates this E-Mail to Master Yoda )



> -----Original Message-----
> From: Paul A. Scott [mailto:pscott-***@public.gmane.org]
> Sent: Friday, September 06, 2002 12:01 PM
> To: hercules-390-***@public.gmane.org
> Subject: [hercules-390] Re: An interesting idea another one?
>
> I know exactly what you mean. Like those annoying, psychotic foot
notes. I
> just move on to the next post.
>
> Paul
>
> > Regarding your propensity to make wordy posts: I wouldn't mind it so
> > much, Ron, if you would simply stick a blank line here and there
> > (say, in between your paragraphs for instance?).
> >
> > Doing so makes one's posts MUCH easier to read.[2] :)
> >
> > - --
> > "Fish" (David B. Trout)
> > fish-6N/dkqvhA+***@public.gmane.org
> >
> > [1] It seems to come and go with me. Sometimes I ramble, other times
> > I don't [too much]. (I do know I'm more likely to do it the tireder
> > (more tired? "tireder *should* be a word IMO!) I get, and given that
> > I suffer from a sleeping disorder and am almost *always* exhausted,
I
> > tend to ramble more often than not. (Like now for example! <sigh!>)
> >
> > [2] And much more likely to BE read! Personally, if I find reading a
> > person's post to be "too much work" for me (due to its poor
> > formatting for example), I won't even bother reading it. No matter
> > how good (valuable) the subject matter may be, if it's poorly
> > formatted and too cumbersome to read, I won't even bother reading
it.
> > Sorry, but on high volume lists, that's just the way I am. <shrug>
>
>
> ------------------------ Yahoo! Groups Sponsor
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
sukato
2002-09-06 16:40:54 UTC
Permalink
Okay, let me make my point more plainly. Fish shouldn't rag on other people
for the way they write their posts, when his posts are just as oddly written. In
other words, people in glass houses shouldn't throw stones. And, since I
agree that this list is for the discussion of hercules, now that I've made my
point clear, I won't discuss it anymore regardless of the hellfire that may rain
down on me for saying it. Have a nice day :)

Paul

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Gregg C Levine" <***@w...> wrote:
> Hello from Gregg C Levine
> Indeed! I've been getting signals from my brother of all people who want
> me to do the same thing. It won't work. I sometimes type the way I talk.
> Fish, Paul, can we agree on one thing? Hercules. Its why we are all
> here. And I like Fish's footnotes, they make his posts more interesting,
> and imply that he's spending more time writing them, then usual three to
> five minutes that we spend.
> -------------------
> Gregg C Levine hansolofalcon-***@public.gmane.org
Adam Thornton
2002-09-06 16:56:59 UTC
Permalink
On a Linux 2.4.17 system, when I fire up Hercules (I'm finally trying to
move off vmnet to tun/tap), I get this (after the disks come up):

HHC894I Error setting MTU for tun: No such device
HHC897I Error setting driving system IP addr for tun: No such device
HHC897I Error setting Hercules IP addr for tun: No such device
HHC897I Error setting netmask for tun: No such device
HHC898I Error getting flags for tun: No such device
HHC848I 0400 configuration failed: hercifc rc=4
HHC038I Initialization failed for device 0400

But here's what I can't figure: lsmod shows that the tun module is
loaded, and ls -l /dev/net/tun shows:
crw-rw---- 1 root hercules 10, 200 Sep 6 11:34 /dev/net/tun

The configuration file has:
0400 3088 CTCI /dev/net/tun 1500 10.1.1.2 10.1.1.1 255.255.255.255
0401 3088 CTCI /dev/net/tun 1500 10.1.1.2 10.1.1.1 255.255.255.255

Which, if I understand correctly should say that the host is 10.1.1.1,
the other end is 10.1.1.2, and there's a point-to-point CTC link between
'em.

Both hercules and hercifc are owned by group hercules and are SGID. So
I don't see what it is that I'm missing. Anyone?

Adam
--
adam-uX/***@public.gmane.org
"My eyes say their prayers to her / Sailors ring her bell / Like a moth
mistakes a light bulb / For the moon and goes to hell." -- Tom Waits
barahul
2002-09-06 19:10:58 UTC
Permalink
--- In hercules-390-F5Bj5G+***@public.gmane.org, Adam Thornton <***@p...> wrote:
> HHC894I Error setting MTU for tun: No such device
> HHC897I Error setting driving system IP addr for tun: No such device
> HHC897I Error setting Hercules IP addr for tun: No such device
> HHC897I Error setting netmask for tun: No such device
> HHC898I Error getting flags for tun: No such device
> HHC848I 0400 configuration failed: hercifc rc=4
> HHC038I Initialization failed for device 0400

> Both hercules and hercifc are owned by group hercules and are SGID.
So
> I don't see what it is that I'm missing. Anyone?

I believe they should be owned root, and SUID (at least hercifc should
be) so that the proper settings can be passed through to the tun
device.

Kris
Fish
2002-09-07 21:29:09 UTC
Permalink
> Okay, let me make my point more plainly. Fish shouldn't rag on
> other people

I wasn't "ragging". I was simply offering a suggestion.

> for the way they write their posts, when his posts are just as
> oddly written.

Not always though.

> In other words, people in glass houses shouldn't throw stones.

<yawn!>

> And, since I agree that this list is for the discussion of
> hercules, now that I've made my point clear, I won't discuss it
> anymore
> regardless of the hellfire that may rain down on me for saying it.

<<GASP!>>

He said a *NAUGHTY WORD(tm)*!!!

You need to stop doing that Paul, because it upsets me!!

Please stop using naughty words in your post!![1]

>;->

> Have a nice day :)

Thanks! I think I shall. :)

> Paul

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org

[1] Or I shall be forced to censor you so NO ONE will be able to see
your posts![2]

[2] Since, after all, MY personally brand of morality is beyond
reproach and therefore should justifiably be enforced upon
*everyone*, right?
Fish
2002-09-07 21:16:25 UTC
Permalink
Well, it's like this Gregg: it's one thing to offer a suggestion for
a minor change to the physical formatting of a post (the insertion of
a simple blank line here and there), but an entirely different matter
altogether to ask someone to ask someone to change the actual content
of their post simply because certain letter sequences happen to
affect their delicate sensibilities.

One is an entirely appropriate suggestion regarding a post's
*physical* formatting whereas the other is an entirely
*inappropriate* attempt to censor the poster.

The two are *worlds* apart IMHO.

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org


> -----Original Message-----
> From: Gregg C Levine [mailto:hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org]
> Sent: Friday, September 06, 2002 9:07 AM
> To: hercules-390-***@public.gmane.org
> Subject: RE: [hercules-390] Re: An interesting idea another one?
> Importance: High
>
>
> Hello from Gregg C Levine
> Indeed! I've been getting signals from my brother of all people who
> want me to do the same thing. It won't work. I sometimes type the
> way I talk. Fish, Paul, can we agree on one thing? Hercules. Its
> why we are all here. And I like Fish's footnotes, they make his
> posts more interesting, and imply that he's spending more time
> writing them, then usual three to five minutes that we spend.
S. Vetter
2002-09-07 21:28:45 UTC
Permalink
Sounds like someone wanting to force me to be politically correct. To
them I say, too friggen' bad...

Speaking about political correctness, didn't this B.S. start when Clinton
was in office? You know the guy who said oral sex wasn't sex? Was he
politiclly correct?

Scott

-----------------

Fish wrote:

>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Well, it's like this Gregg: it's one thing to offer a suggestion for
> a minor change to the physical formatting of a post (the insertion of
> a simple blank line here and there), but an entirely different matter
> altogether to ask someone to ask someone to change the actual content
> of their post simply because certain letter sequences happen to
> affect their delicate sensibilities.
>
> One is an entirely appropriate suggestion regarding a post's
> *physical* formatting whereas the other is an entirely
> *inappropriate* attempt to censor the poster.
>
> The two are *worlds* apart IMHO.
>
> - --
> "Fish" (David B. Trout)
> fish-6N/dkqvhA+***@public.gmane.org
>
> > -----Original Message-----
> > From: Gregg C Levine [mailto:hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org]
> > Sent: Friday, September 06, 2002 9:07 AM
> > To: hercules-390-***@public.gmane.org
> > Subject: RE: [hercules-390] Re: An interesting idea another one?
> > Importance: High
> >
> >
> > Hello from Gregg C Levine
> > Indeed! I've been getting signals from my brother of all people who
> > want me to do the same thing. It won't work. I sometimes type the
> > way I talk. Fish, Paul, can we agree on one thing? Hercules. Its
> > why we are all here. And I like Fish's footnotes, they make his
> > posts more interesting, and imply that he's spending more time
> > writing them, then usual three to five minutes that we spend.
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXpsqUj11/TE7j4qEQJw6QCfRbWJAz5aoFm+BOnV9K4adv9kaX4AoJoR
> e0x/LpCQztBRzCdkosYc1VnB
> =TpVY
> -----END PGP SIGNATURE-----
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
S. Vetter
2002-09-07 21:29:28 UTC
Permalink
My appologies for getting way off topic!

Scott

----------------

Fish wrote:

>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Well, it's like this Gregg: it's one thing to offer a suggestion for
> a minor change to the physical formatting of a post (the insertion of
> a simple blank line here and there), but an entirely different matter
> altogether to ask someone to ask someone to change the actual content
> of their post simply because certain letter sequences happen to
> affect their delicate sensibilities.
>
> One is an entirely appropriate suggestion regarding a post's
> *physical* formatting whereas the other is an entirely
> *inappropriate* attempt to censor the poster.
>
> The two are *worlds* apart IMHO.
>
> - --
> "Fish" (David B. Trout)
> fish-6N/dkqvhA+***@public.gmane.org
>
> > -----Original Message-----
> > From: Gregg C Levine [mailto:hansolofalcon-XfrvlLN1Pqtfpb/***@public.gmane.org]
> > Sent: Friday, September 06, 2002 9:07 AM
> > To: hercules-390-***@public.gmane.org
> > Subject: RE: [hercules-390] Re: An interesting idea another one?
> > Importance: High
> >
> >
> > Hello from Gregg C Levine
> > Indeed! I've been getting signals from my brother of all people who
> > want me to do the same thing. It won't work. I sometimes type the
> > way I talk. Fish, Paul, can we agree on one thing? Hercules. Its
> > why we are all here. And I like Fish's footnotes, they make his
> > posts more interesting, and imply that he's spending more time
> > writing them, then usual three to five minutes that we spend.
>
> -----BEGIN PGP SIGNATURE-----
> Version: PGP 7.0.4
>
> iQA/AwUBPXpsqUj11/TE7j4qEQJw6QCfRbWJAz5aoFm+BOnV9K4adv9kaX4AoJoR
> e0x/LpCQztBRzCdkosYc1VnB
> =TpVY
> -----END PGP SIGNATURE-----
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
broidoj
2002-09-06 20:51:18 UTC
Permalink
As with almost everything in life, to each his or her own... I
happen to enjoy those psychotic footnotes.

--- In hercules-390-F5Bj5G+***@public.gmane.org, "Paul A. Scott" <***@s...> wrote:

> I know exactly what you mean. Like those annoying, psychotic foot
> notes. I just move on to the next post.
Fish
2002-09-07 21:08:47 UTC
Permalink
(Note to Paul: feel free to move on to the next post. >;->)

Annoying? Perhaps.

Whenever I happen to use them maybe (which isn't all the time of
course).

But "psychotic"? Really??

You think my footnotes[1] are *psychotic*?!

(Hmph!)

;-)

- --
"Fish" (David B. Trout)
fish-6N/dkqvhA+***@public.gmane.org

[1] Footnotes? What footnotes? I don't see any footnotes! >;->


> -----Original Message-----
> From: Paul A. Scott [mailto:pscott-***@public.gmane.org]
> Sent: Friday, September 06, 2002 9:01 AM
> To: hercules-390-***@public.gmane.org
> Subject: [hercules-390] Re: An interesting idea another one?
> Importance: High
>
>
> I know exactly what you mean. Like those annoying, psychotic foot
> notes. I just move on to the next post.
>
> Paul
S. Vetter
2002-09-07 18:00:26 UTC
Permalink
Here is one of the pranks I pulled:

In our shop we had a rather gullable VTAM systems programmer. So one
day I changed his TSO logon proc and created a few ISPF screens. So one
day he made some changes to VTAM and goes to lunch. After lunch he
signs onto TSO via VTAM, but only to get a VTAM logon screen of one of
our vendors. He tried to logon again, but now gets a logon screen to
the Burroughs company. By this time he's going some serious mubling and
head scratches. And tries again (and the rest of us systems programers
are looking over his cubicle walls trying desparitly holding back some
snickers). But this time he gets the Secritary of State's drivers
licence bureau. Now that he is getting more worried and tries one last
time and gets a screen for the NSA (National Security Agency), of course
with the line saying his PU and LU address has been recorded and is
being traced. With a startled look on his face, we yell at him to turn
off his terminal so he couldn't be no longer tracked down! Of course he
does...Mean while we all slither down the cubicle walls laughing like
hell! I did finially fess up to my prank after he calmed down and we
could tell him without bursting into laughter about 10 minutes later...

Scott
S. Vetter
2002-09-07 18:09:51 UTC
Permalink
Being a notorius practical joker that I am... I will not admit to
having performed this prank: At one company that I worked at we had a
large turn over of programmers and just went through a massive heiring
process. So much so, the phone book changed daily and lasting for
weeks. It got to the point no one knew who each other was any more.
One systems programmer came in very early and decided to change this
problem of not knowing who each other was. On each 3278 was a key to
lock the device. Of course the keys could be inserted into any 3278 but
just not turned to unlock it. And this is precicely what happened. The
keys were all taken out and randomly selected and put back in, but not
unlocked. So when the programmers started coming back in, they all
tried to unlock their terminal, but to no avail. So some time had to go
by and people started roaming the floor to find what key did work on
their terminal and hope that when they got back to their desk their
terminal would be unlocked by the proper key. Management wasn't too
happy, but the many of us knew where many other sats and said hello in
passing.

Scott

P.S. The systems programmer who pulled this stunt also did this to
themselves as well as not not look like the culpret of this prank.
There was another side benefit for this systems programmer, but they
wont tell...
Paul Yahoo
2002-09-07 18:16:08 UTC
Permalink
With the P/390 and the VM/ESA AD disk, a MOUNT command was included. The
syntax was:

MOUNT 181 ( QUERY

MOUNT 181 C:\HERCULES\TAPE.AWS

It looks like they used a SETDIAG type CCW.

Would it be possible to add that support into Hercules. I would be happy
to provide traces of the channel commands used.

Paul
S. Vetter
2002-09-07 18:25:34 UTC
Permalink
And last for the day: The company that I worked for hired a CICS
systems programmer who I grew to dispise and for good reason. He would
make a change to MVS, which I was solely resonsible for and on the next
IPL a problem would show up and I had to come in the middle of the night
to find out what change he made and to fix the problem that he created.
Of course he didn't have to come in since it wasn't his problem and he
wouldn't write down what changes he made. Of course management wouldn't
stop him from making changes or force him to write down any changes he
made. So, dispite the phrase "vengence is mine - says the lord", I
figured he needed a little help. I created a little program which would
knock down his CICS test region at any and at all time of day or night
and with various undetectable reasons. After two weeks I figured he had
enough and pulled my program. Moral of the story: Don't piss off your
friendly MVS systems programmer!
Jim Keohane
2002-09-07 18:47:38 UTC
Permalink
I remember the panic decades ago when a critical month end program
started to produce 1000's of operator messages:

"The Analyst, NAME WITHHELD, assured me the program could never get to this
point."

I immediately felt for the poor programmer who had been obviously asking
the analyst what to do in a certain case only to have the analyst say "Don't
worry. It can't happen!"

- Jim Keohane

<Signature>
<Name>Jim Keohane</Name>
<Quote>You must be certifiable to even think of using PKI</Quote>
</Signature>
S. Vetter
2002-09-07 20:14:41 UTC
Permalink
I wonder how many programmers were told that something like only to have
happened!

Scott

--------------

Jim Keohane wrote:

> I remember the panic decades ago when a critical month end program
> started to produce 1000's of operator messages:
>
> "The Analyst, NAME WITHHELD, assured me the program could never get to this
> point."
>
> I immediately felt for the poor programmer who had been obviously asking
> the analyst what to do in a certain case only to have the analyst say "Don't
> worry. It can't happen!"
>
> - Jim Keohane
>
> <Signature>
> <Name>Jim Keohane</Name>
> <Quote>You must be certifiable to even think of using PKI</Quote>
> </Signature>
>
>
> Community email addresses:
> Post message: hercules-390-***@public.gmane.org
> Subscribe: hercules-390-subscribe-***@public.gmane.org
> Unsubscribe: hercules-390-unsubscribe-***@public.gmane.org
> List owner: hercules-390-owner-***@public.gmane.org
>
> Files and archives at:
> http://groups.yahoo.com/group/hercules-390
>
> Get the latest version of Hercules from:
> http://www.conmicro.cx/hercules
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Jim Keohane
2002-09-07 18:55:02 UTC
Permalink
The CCAP programmers had shown me how to enter instructions into S/360
via toggle switches.

When 2nd shift supervisor, Izzy, came in System B was always bereft of
all peripherals. At business shutdown he would grab some DASD and TAPEs and
unit record from systems A and C via switching stations.

One time Izzy found a note on console, allegedly from IBM C.E., stating
"just hit start. had to hit stop before maintenance on ..."

He hit start only to see console message loop saying:

"screw you Izzy"

over and over again on a system with no peripherals! SIO and TIO is a
dangerous thing in the hands of a warped apprentice.

- Jim Keohane

<Signature>
<Name>Jim Keohane</Name>
<Quote>You must be certifiable to even think of using PKI</Quote>
</Signature>
Graham Goodwin
2002-09-09 01:05:07 UTC
Permalink
... and positively the last word on this.

<snipped account of nasty profile exec>

If yer program in CMS loops on a read, say via the Rexx pull statement, there
is still the "secret" get out of jail card by backspacing one character on
the CMS command line, so that the cursor wraps back to the end of the
previous line and hitting enter. You can then type HX and interrupt the
program in the conventional way. Obviously the CMS people got pissed off
getting stuck in this way too...

Graham.
Jim Pierson (Hercules)
2002-09-09 00:35:10 UTC
Permalink
I had this print operator that was a real pain is the hind
quarters. Trying to get her to print my compile listings was
close to pulling teeth. After many incidents of fighting
with her to print my stuff, I'd finally had enough. Our
print que screens at the time were CICS transactions. I got
this evil look on my face and proceeded to go into CECI and
setup a timer that issues a START TRANSACTION(CSSF) TERM(XXXX)
at a particular interval, I think it was 30 minutes. The poor
girl to this day has never figured out who (and how) her
terminal was getting logged off of CICS. She had accused every
operator in the room, but never fingered me for the dirty deed.


Jim
Cam Farnell
2002-09-09 04:21:19 UTC
Permalink
Ok, so were telling stories. In 1972 I was one of a couple of *very*
junior programmers, and occasional evening operator, at a local
community college. Our work was done on a 360/65 but the college, not
being large enough to justify a machine of that size, ran via an RJE
terminal connected to a 65 at a university 100 miles away. The day
operator, in retrospect not a bad guy, was a fellow by the name of Ron
Payne and at the time we found him to be officious, too slow and
generally rather annoying. Sitting proudly on his desk near the RJE
terminal was a triangular name plate with plug-in plastic letters
reading 'RON PAYNE'. One time while serving as evening operator it
occurred to me that by rearranging the letters you could turn "RON
PAYNE" into "RONNY APE" and thats just what I did, leaving it for him to
find the next morning. Next evening the sign was back to his real name
and I revised it again. The following evening it was back to RON PAYNE
but this time was wrapped in a solid layer of scotch tape. Well, it
wasn't that busy in the evening so I laboriously removed the scotch
tape, put the sign back to RONNY APE and then wrapped that in lots of
scotch tape.

We never saw the sign again, so something else was in order. The RJE
terminal consisted of a card reader, a line printer, a teletype and a
PDP-11 all connected via an unbelievably fast 9600 baud half-duplex
modem. To get the terminal up you clacked a "ipl" deck in the reader,
twiddled some switches on the front panel of the PDP-11, the cards read
in and it sprang to life by printing MULTILEAVING INITIATED on the
teletype and then, once communication with the 65 had been established,
REMOTE-9 SIGNED ON. Noise on the line would cause it to print COMM ERR a
couple of times an hour. We thought it might be fun to change the
messages so we read in the "ipl" deck as data, printed it and went
looking for messages without finding any. Well of course not, we were
assuming EBCDIC but the PDP-11 would be ASCII. So we converted to ASCII
and printed it out again, still without finding any message text.
D.a.m.n. That had us stumped briefly. Then we noticed that the first
five letters of INITIATED were of the form ABACA and that was a
reasonably unusual sequence. Whipped up a PL/I program to search for
such a sequence and found only one. Ah-ha. Translated based on that and
in a series of iterations decoded all the messages; they were not in any
code we knew about. Using a keypunch (there was no card punch at the
terminal) we copied the machine code portions of the card and then
slowly and carefully (there were lots of strange overpunches required)
put in our illicit messages.

When Mr. Payne fired it up the following morning it greeted him with
MULTI-APING INITIATED and then REMOTE-9 GO APE GO; noise on the line
resulted in APE ERR. Furious, he roared off to his boss and the two of
them took a close look at the ipl deck which now consisted of about 100
old dirty worn cards and three suspiciously new ones. They duplicated
the new cards with printing turned on hoping to see the offending
messages but only got random chicken tracks. It didn't take a genius to
figure out who had done it so they roared off to OUR boss who in due
course called us out on the carpet, told us to put back the original
cards and to quit playing pranks on the operator. We did what we were
told but I couldn't help noticing that our boss looked to be suppressing
a smile while he bawled us out.
Continue reading on narkive:
Loading...