[H390-VM] VM/XA

[W Mainframe [e-mail] [H390-VM]]

Guys,
Does anyone is running a VM/XA system under Hercules? Is it possible?
Thank youDan

Sent from Yahoo Mail for iPhone

[Ivan Warren [e-mail] [H390-VM]]

Le 5/21/2018 à 5:18 PM, Fran Hensler ***@gmail.com [H390-VM] a écrit :
>
>
> A from my journal:
>
> “05/10/92 (Mother’s Day) Installed IBM 9221-170 w/96M (64M usable)
> running VM/ESA 1.0 - 370 feature.   The 370 Feature is a functional
> equivalent to VM/SP R6 in that programs can only address 32M but an
> additional 32M can be used for paging.”
>
>
Very surprised !!!!

A virtual machine running under VM/ESA 370 feature (aka VM/SP 6+ HPO)
can still only have 16MB of virtual storage (but the whole 64M (max) of
physical storage of the system can be used - thanks to S/370 Extended
DAT[1])

--Ivan

[1] Extended DAT has nothing to do with DAT Extension features.. It's
just the possiblilty to address 26 bits of real storage in the DAT
tables in S/370 mode.

[Drew Derbyshire [e-mail] [H390-VM]]

On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
[H390-VM] wrote:
> If IBM would allow to run a modern
> VM version, from VM/SP to Z/VM, would
> be possible to rebuild an hobbyist VNET
> like network?

TLDR: Why would you need a modern VM?

IBM had peer to peer connections for 239 systems in 1979, per
https://en.wikipedia.org/wiki/RSCS#Background.  Those would have been
VM/370 R6 systems or older. The basic concept of store and forward via
the RSCS and the VM spool really didn't change in 20 years.

The Wiki link above has a comment about how the public released version
of RSCS had stubs for store & forward functionality. That makes me
wonder how hard it would be to add peer-to-peer store & forward to the
VM/370 R6 RSCS version.

The thing about VM was how CP was always distinct from the VM instances
which ran under it. Even now you could probably take VM/SP RSCS Version
1 R3 and get run it running on from anything from VM/370 R6 up through
z/VM.  (Note that I have /not/ tried this.)

A useful reference for RSCS behavior under later VM versions would be
https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemNetworkingPro_13960888.

I can't seem to find a definition of the actual RSCS protocols, but
foreign programs like JNET (VMS) and UREP (UNIX)  implemented them, so
it must be possible.  Or implement a Hercules specific BSC driver (VMH?).

        * * *
Now, to answer the obvious question before it's asked ... how does one
invite MVS 3.8 to the party:

That's harder.

The NJE protocol */is/* documented:
https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1.hasa600/toc.htm

I have to admit, doing MVS 3.8 NJE support would make me nervous/.*

*/Doing Store & Forward under VM requires a single virtual machine that
supports RSCS protocols and a spool that already supports tagging. There
is not anything break outside the reader queue of the RSCS machine,
Compare this to NJE store & forward support on MVS, which requires you
get into the guts of JES2. If (when?) you screw up and your spool is toast.

(That reminds of an old story of a HASP spool that was reported at 110%
utilization.  The next thing that happened, not surprisingly, was a
system crash.)

-ahd-

p.s. Is the non-store & forward RSCS on the six pack?

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

The six stages of a project:
1. Unbounded enthusiasm 4. Frantic Search for the Guilty
2. Total disillusionment 5. Punishment of the Innocent
3. Panic 6. Reward of the non-participants

['Dave Wade' [e-mail] [H390-VM]]

Drew,



Yes the non-program-product version of RSCS is on the 6-pack. I have not
looked at it in detail, but I have used it with some of the RJE tools that
were written for MVS.

Volume 3 of the VM Program Logic Manual covers RSCS and there is a copy of
the Release 5 version on bitsavers.



Dave



From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
Sent: 22 May 2018 00:34
To: H390-***@yahoogroups.com
Subject: [H390-VM] You're in a maze of twisty little peer-to-peer network
links, all different (was Re: VM/XA)





On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
<mailto:***@vitillaro.org> [H390-VM] wrote:
> If IBM would allow to run a modern
> VM version, from VM/SP to Z/VM, would
> be possible to rebuild an hobbyist VNET
> like network?

TLDR: Why would you need a modern VM?

IBM had peer to peer connections for 239 systems in 1979, per
https://en.wikipedia.org/wiki/RSCS#Background. Those would have been VM/370
R6 systems or older. The basic concept of store and forward via the RSCS and
the VM spool really didn't change in 20 years.

The Wiki link above has a comment about how the public released version of
RSCS had stubs for store & forward functionality. That makes me wonder how
hard it would be to add peer-to-peer store & forward to the VM/370 R6 RSCS
version.

The thing about VM was how CP was always distinct from the VM instances
which ran under it. Even now you could probably take VM/SP RSCS Version 1 R3
and get run it running on from anything from VM/370 R6 up through z/VM.
(Note that I have not tried this.)

A useful reference for RSCS behavior under later VM versions would be
https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemN
etworkingPro_13960888.

I can't seem to find a definition of the actual RSCS protocols, but foreign
programs like JNET (VMS) and UREP (UNIX) implemented them, so it must be
possible. Or implement a Hercules specific BSC driver (VMH?).

* *
*
Now, to answer the obvious question before it's asked ... how does one
invite MVS 3.8 to the party:

That's harder.

The NJE protocol is documented:
https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1
..hasa600/toc.htm

I have to admit, doing MVS 3.8 NJE support would make me nervous.

Doing Store & Forward under VM requires a single virtual machine that
supports RSCS protocols and a spool that already supports tagging. There is
not anything break outside the reader queue of the RSCS machine, Compare
this to NJE store & forward support on MVS, which requires you get into the
guts of JES2. If (when?) you screw up and your spool is toast.

(That reminds of an old story of a HASP spool that was reported at 110%
utilization. The next thing that happened, not surprisingly, was a system
crash.)

-ahd-

p.s. Is the non-store & forward RSCS on the six pack?

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

The six stages of a project:
1. Unbounded enthusiasm 4. Frantic Search for the Guilty
2. Total disillusionment 5. Punishment of the Innocent
3. Panic 6. Reward of the non-participants

[Drew Derbyshire [e-mail] [H390-VM]]

On 5/21/18 4:46 PM, 'Dave Wade' ***@gmail.com [H390-VM] wrote:
>
> Yes the non-program-product version of RSCS is on the 6-pack. I have
> not looked at it in detail, but I have used it with some of the RJE
> tools that were written for MVS.
>
> Volume 3 of the VM Program Logic Manual covers RSCS and there is a
> copy of the Release 5 version on bitsavers.
>
Thanks.

Unfortunately but not surprising, it doesn't include the VMC, VMB or NJE
drivers mentioned in the R3M0 manual.  So no Store & Forward protocols
in it.

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

Know thyself. If you need help, call the C.I.A.

['Dave Wade' [e-mail] [H390-VM]]

Drew,



It is likely that an NJE driver will be included in an upcoming release of
the six-pack. Do you need the NJE driver to get TAG data preserved across
links. Whilst I have USED RSCS I have never managed a machine on which it
was installed, so whilst I know what it does I have very little idea about
how it does it. The VM systems I managed were all connected by X.25
networks. They were accessible via an X.25 to Bitnet gateway but that was
not on a machine I managed.



Dave



From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
Sent: 22 May 2018 02:21
To: H390-***@yahoogroups.com
Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer
network links, all different (was Re: VM/XA)








On 5/21/18 4:46 PM, 'Dave Wade' ***@gmail.com
<mailto:***@gmail.com> [H390-VM] wrote:

Yes the non-program-product version of RSCS is on the 6-pack. I have not
looked at it in detail, but I have used it with some of the RJE tools that
were written for MVS.

Volume 3 of the VM Program Logic Manual covers RSCS and there is a copy of
the Release 5 version on bitsavers.

Thanks.

Unfortunately but not surprising, it doesn't include the VMC, VMB or NJE
drivers mentioned in the R3M0 manual. So no Store & Forward protocols in
it.

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

Know thyself. If you need help, call the C.I.A.

[Tony Harminc [e-mail] [H390-VM]]

On 22 May 2018 at 03:32, 'Dave Wade' ***@gmail.com wrote:

> It is likely that an NJE driver will be included in an upcoming release of the six-pack. Do you need the NJE driver to get TAG data preserved across links. Whilst I have USED RSCS I have never managed a machine on which it was installed, so whilst I know what it does I have very little idea about how it does it. The VM systems I managed were all connected by X.25 networks. They were accessible via an X.25 to Bitnet gateway but that was not on a machine I managed.

Does VM/370 *have* tag data (or a CP TAG command)? I don't think it
had any major use other than for RSCS to decide where to route things,
though in theory you could tag SPOOL files for any reason you liked.

Tony H.

['Dave Wade' [e-mail] [H390-VM]]

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 24 May 2018 00:28
> To: H390-***@yahoogroups.com
> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer network
> links, all different (was Re: VM/XA)
>
> On 22 May 2018 at 03:32, 'Dave Wade' ***@gmail.com wrote:
>
> > It is likely that an NJE driver will be included in an upcoming release of the
> six-pack. Do you need the NJE driver to get TAG data preserved across links.
> Whilst I have USED RSCS I have never managed a machine on which it was
> installed, so whilst I know what it does I have very little idea about how it
> does it. The VM systems I managed were all connected by X.25 networks.
> They were accessible via an X.25 to Bitnet gateway but that was not on a
> machine I managed.
>
> Does VM/370 *have* tag data (or a CP TAG command)? I don't think it had
> any major use other than for RSCS to decide where to route things, though in
> theory you could tag SPOOL files for any reason you liked.
>

The manual says it has the TAG command. Up to 136 characters can be entered..
I think even when I was involved at NERC LISTSERV was not freely available as we could have used it internally.
I have several workshop tapes from the 80's and 90's that I purloined from George Shedlock's site

http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html

but I don't think any of these has an old listserv on them.


> Tony H.
>
Dave W.

[Tony Harminc [e-mail] [H390-VM]]

.On 24 May 2018 at 02:26, 'Dave Wade' ***@gmail.com wrote:

>> Does VM/370 *have* tag data (or a CP TAG command)? I don't think it had
>> any major use other than for RSCS to decide where to route things, though in
>> theory you could tag SPOOL files for any reason you liked.
>>
>
> The manual says it has the TAG command. Up to 136 characters can be entered.

Makes sense if RSCS Networking Program Product was actually supported
on VM/370. TAG is pretty much necessary for store & forward.

> I think even when I was involved at NERC LISTSERV was not freely available as we could have used it internally.
> I have several workshop tapes from the 80's and 90's that I purloined from George Shedlock's site
>
> http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html
>
> but I don't think any of these has an old listserv on them.

Unfortunately they all seem to be missing from that site. :-(

As are the older ones are from the official VM Workshop site at
http://www.vmworkshop.org/pastTape.shtml

I fear that these are going to be lost, as so much has been already.

Tony H.

['\'Fish\' (David B. Trout)' [e-mail] [H390-VM]]

Tony Harminc wrote:
> Dave Wade wrote:

[...]
> > I have several workshop tapes from the 80's and 90's
> > that I purloined from George Shedlock's site
> >
> > http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html
>
> Unfortunately they all seem to be missing from that site. :-(
[...]
> I fear that these are going to be lost, as so much has been
> already.

The web page is badly formatted as well.

I was able to download about half of the files however, by simply replacing upper case ".ZIP" with lower case ".zip".

Unfortunately even after doing that, some of the files were still not found.. The ones which appear to be missing (which I was unable to download(*)) are:


90PCTL.AWS.zip

91WT.AWS.zip

92WTPC.AWS.zip
92VM.AWS.zip
92WT.AWS.zip

93TL.AWS.zip
93PCWT.AWS.zip
93VM.AWS.zip

94TL.AWS.zip
94PC.AWS.zip
94VM.AWS.zip


All of the others I *WAS* able to download by using lower case ".zip".

Dave? Can you fix the URLs on your web page so we can download the missing files?

I plan to upload them to upload them to my SoftDevLabs FTP site for preservation if that's okay? Yes? No? Maybe?

Thanks!


------------
(*) Either because the filename in the download URL is wrong or the file itself is truly missing ("not found").

--
"Fish" (David B. Trout)
Software Development Laboratories
http://www.softdevlabs.com
mail: ***@softdevlabs.com

['Dave Wade' [e-mail] [H390-VM]]

I

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 24 May 2018 22:58
> To: H390-***@yahoogroups.com
> Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer network
> links, all different (was Re: VM/XA)
>
> Tony Harminc wrote:
> > Dave Wade wrote:
>
> [...]
> > > I have several workshop tapes from the 80's and 90's that I
> > > purloined from George Shedlock's site
> > >
> > > http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html
> >
> > Unfortunately they all seem to be missing from that site. :-(
> [...]
> > I fear that these are going to be lost, as so much has been already.
>
> The web page is badly formatted as well.

Sorry about that. It was never really meant for public consumption. The web site is being moved from random HTML to Druple.
Its now less worse

>
> I was able to download about half of the files however, by simply replacing
> upper case ".ZIP" with lower case ".zip".
>
> Unfortunately even after doing that, some of the files were still not found.
> The ones which appear to be missing (which I was unable to download(*))
> are:
>
>
> 90PCTL.AWS.zip
>
> 91WT.AWS.zip
>
> 92WTPC.AWS.zip
> 92VM.AWS.zip
> 92WT.AWS.zip
>
> 93TL.AWS.zip
> 93PCWT.AWS.zip
> 93VM.AWS.zip
>
> 94TL.AWS.zip
> 94PC.AWS.zip
> 94VM.AWS.zip
>

Geroges site appears to be up today. I often find it down. Perhaps he is reading this and has brought it up..

http://www.shedlock.org/Shedlock/Hercules/VM_Workshop_Tapes/vmwkshop.html

I have copied the ones I am missing and they should be available now. I am supposed to have unlimited bandwidth, not sure about George.

>
> All of the others I *WAS* able to download by using lower case ".zip".
>
> Dave? Can you fix the URLs on your web page so we can download the
> missing files?
>
> I plan to upload them to upload them to my SoftDevLabs FTP site for
> preservation if that's okay? Yes? No? Maybe?
>

I took them from George. I think that having them on several sites would be a good idea.

> Thanks!
>

Dave

>
> ------------
> (*) Either because the filename in the download URL is wrong or the file itself
> is truly missing ("not found").
>
> --
> "Fish" (David B. Trout)
> Software Development Laboratories
> http://www.softdevlabs.com
> mail: ***@softdevlabs.com
>
>
>
>
>
>
> ------------------------------------
> Posted by: "\"Fish\" \(David B. Trout\)" <***@gmail.com>
> ------------------------------------
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>

['\'Fish\' (David B. Trout)' [e-mail] [H390-VM]]

Dave Wade wrote:

[...]
> I have copied the ones I am missing and they should be
> available now. I am supposed to have unlimited bandwidth,
> not sure about George.

Then that could be a problem, since all of the links on your newly formatted web page point to George's site, not yours.

Also, the list of 1992 (Tulane University) tapes appear to have changed, and is now missing the 'WT' tape. Previously, the list was:

. 1992 TL
. 1992 WTPC
. 1992 VM
. 1992 WT

Now however, the list is:

. 1992 TL
. 1992 WTPC
. 1992 VM
. 1992 TL

The last entry is now a duplicate of the first entry: both are for 'TL', and 'WT' is now missing. Is that correct?

And should all of your links be pointing to George's site? Or your site?

Thanks.

--
"Fish" (David B. Trout)
Software Development Laboratories
http://www.softdevlabs.com
mail: ***@softdevlabs.com

['Dave Wade' [e-mail] [H390-VM]]

Fish,
Can you just check that your cache has cleared. When I look here:-

http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html

I see

1992 (Tulane University)

1992 TL
1992 WTPC
1992 VM
1992 WT

And they all seem to point back to my site

Dave

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 25 May 2018 16:26
> To: H390-***@yahoogroups.com
> Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer network
> links, all different (was Re: VM/XA)
>
> Dave Wade wrote:
>
> [...]
> > I have copied the ones I am missing and they should be available now.
> > I am supposed to have unlimited bandwidth, not sure about George.
>
> Then that could be a problem, since all of the links on your newly formatted
> web page point to George's site, not yours.
>
> Also, the list of 1992 (Tulane University) tapes appear to have changed, and is
> now missing the 'WT' tape. Previously, the list was:
>
> . 1992 TL
> . 1992 WTPC
> . 1992 VM
> . 1992 WT
>
> Now however, the list is:
>
> . 1992 TL
> . 1992 WTPC
> . 1992 VM
> . 1992 TL
>
> The last entry is now a duplicate of the first entry: both are for 'TL', and 'WT' is
> now missing. Is that correct?
>
> And should all of your links be pointing to George's site? Or your site?
>
> Thanks.
>
> --
> "Fish" (David B. Trout)
> Software Development Laboratories
> http://www.softdevlabs.com
> mail: ***@softdevlabs.com
>
>
>
>
>
>
> ------------------------------------
> Posted by: "\"Fish\" \(David B. Trout\)" <***@gmail.com>
> ------------------------------------
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>

[[e-mail] [H390-VM]]

I had not visited that page before, but I get custom 404 pages on the following tapes:

1990 PCTL
1991 WT
1992 WTPC
1992 WT
1993 VM
1994 TL



All others downloaded fine.


Sorry if I'm rehashing already-known info.


Scott

['Dave Wade' [e-mail] [H390-VM]]

Scott,



Not at all, sorry I am being slow, I am useless at the detailed niggly stuff. Also I am fiddling with this while copying files on my P390. Looks like George does not have 1992 wt either. The other bad ones look like plain AWS files not ZIP. You may have to right click and do “save as”.



Dave



From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
Sent: 25 May 2018 17:46
To: H390-***@yahoogroups.com
Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer network links, all different (was Re: VM/XA)





I had not visited that page before, but I get custom 404 pages on the following tapes:



1990 PCTL

1991 WT

1992 WTPC

1992 WT

1993 VM

1994 TL



All others downloaded fine.



Sorry if I'm rehashing already-known info.



Scott

['\'Fish\' (David B. Trout)' [e-mail] [H390-VM]]

Dave Wade wrote:

> Fish,
> Can you just check that your cache has cleared.

Duh!


> When I look here:-
>
> http://www.smrcc.org.uk/members/g4ugm/shedlock/vmwkshop.html
>
> I see
>
> 1992 (Tulane University)
>
> 1992 TL
> 1992 WTPC
> 1992 VM
> 1992 WT
>
> And they all seem to point back to my site

Yes, I see that -- now. But that wasn't what I was seeing before.

But there's still a problem. The following links are still "Page not found":

90PCTL.AWS.ZIP
91WT.AWS.ZIP
92WTPC.AWS.ZIP
92WT.AWS.ZIP
93VM.AWS.ZIP
94TL.AWS.ZIP

--
"Fish" (David B. Trout)
Software Development Laboratories
http://www.softdevlabs.com
mail: ***@softdevlabs.com

['Dave Wade' [e-mail] [H390-VM]]

Its mostly me. I think these are fixed.

Dave

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 25 May 2018 16:26
> To: H390-***@yahoogroups.com
> Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer network
> links, all different (was Re: VM/XA)
>
> Dave Wade wrote:
>
> [...]
> > I have copied the ones I am missing and they should be available now.
> > I am supposed to have unlimited bandwidth, not sure about George.
>
> Then that could be a problem, since all of the links on your newly formatted
> web page point to George's site, not yours.
>
> Also, the list of 1992 (Tulane University) tapes appear to have changed, and is
> now missing the 'WT' tape. Previously, the list was:
>
> . 1992 TL
> . 1992 WTPC
> . 1992 VM
> . 1992 WT
>
> Now however, the list is:
>
> . 1992 TL
> . 1992 WTPC
> . 1992 VM
> . 1992 TL
>
> The last entry is now a duplicate of the first entry: both are for 'TL', and 'WT' is
> now missing. Is that correct?
>
> And should all of your links be pointing to George's site? Or your site?
>
> Thanks.
>
> --
> "Fish" (David B. Trout)
> Software Development Laboratories
> http://www.softdevlabs.com
> mail: ***@softdevlabs.com
>
>
>
>
>
>
> ------------------------------------
> Posted by: "\"Fish\" \(David B. Trout\)" <***@gmail.com>
> ------------------------------------
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>

[Tony Harminc [e-mail] [H390-VM]]

On 21 May 2018 at 21:20, Drew Derbyshire ***@kew.com wrote:
>
> On 5/21/18 4:46 PM, 'Dave Wade' ***@gmail.com [H390-VM] wrote:
>>
>> Yes the non-program-product version of RSCS is on the 6-pack. I have not looked at it in detail, but I have used it with some of the RJE tools that were written for MVS.
>>
>> Volume 3 of the VM Program Logic Manual covers RSCS and there is a copy of the Release 5 version on bitsavers.
>>
> Thanks.
>
> Unfortunately but not surprising, it doesn't include the VMC, VMB or NJE drivers mentioned in the R3M0 manual. So no Store & Forward protocols in it.

It's worth keeping in mind what these drivers do. VMB and VMC drive
BSC and CTC connections, respectively. The content is the same, and
it's essentially just a 4K CP SPOOL block. So to this extent the
communication is somewhat dependent (more than a modern designer would
choose) on the layout of a CP SPOOL file record.

The NJI (not NJE) driver was written to roughly the JES2 NJE specs,
though (as has been pointed out many many times by Lynne Wheeler) it
was far more robust than JES2's NJE, and was used in IBM's internal
VNET network to allow otherwise incompatible or malfunctioning end
nodes to interoperate. This is the protocl documented much later in
the Network Job Entry Formats and Protocols book, currently
SA32-0988-30. (Dash-30 must be telling us something...)

SML (Simple Multi-Leaving?) is interesting in that it can act as host
or terminal, i.e. it can look like HASP/JES2 to a remote 360/20 or the
like running HRTPB360, *or* it can look like HRTPB360 to another host.
So to some extent the roots for NJE are in SML, certainly at the lower
level or multi-leaving, which is built on BSC.

In the late 1970s I had a version of HRTPB360 from some mods tape that
had been converted to run on OS (mostly to use WAIT instead of a loop,
and EXCP instead of SIO/TIO), and I modified it further to run on VS1
and interact with the console operator to do file transfers to and
from JES2.

All the above are from RSCS V1, which ran standalone in a virtual
machine. RSCS V2 ran under the Group Control System (GCS) of later
VMs, which also ran a version of VTAM, probably modified from VS1
VTAM.

Well, I digress...

Tony H.

[Giuseppe Vitillaro [e-mail] [H390-VM]]

On Mon, 21 May 2018, Drew Derbyshire ***@kew.com [H390-VM] wrote:

> On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org [H390-VM]
> wrote:
>> If IBM would allow to run a modern
>> VM version, from VM/SP to Z/VM, would
>> be possible to rebuild an hobbyist VNET
>> like network?
>
> TLDR: Why would you need a modern VM?

Because I'm as naive as possible about VM? ;-)

I'm slowing realizing VM/370 is enough ;-)

Still I dream about at least VM/SP or VM/XA, the
VM versions I had been a user for ten years ;-)

Peppe.

['Dave Wade' [e-mail] [H390-VM]]

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 22 May 2018 10:02
> To: Drew Derbyshire ***@kew.com [H390-VM] <H390-
> ***@yahoogroups.com>
> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer
network
> links, all different (was Re: VM/XA)
>
> On Mon, 21 May 2018, Drew Derbyshire ***@kew.com [H390-VM] wrote:
>
> > On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
> > [H390-VM]
> > wrote:
> >> If IBM would allow to run a modern
> >> VM version, from VM/SP to Z/VM, would be possible to rebuild an
> >> hobbyist VNET like network?
> >
> > TLDR: Why would you need a modern VM?
>
> Because I'm as naive as possible about VM? ;-)
>
> I'm slowing realizing VM/370 is enough ;-)
>
> Still I dream about at least VM/SP or VM/XA, the VM versions I had been a
> user for ten years ;-)
>

Me too. Even with all its tweaks and add-ons the six-pack is not quite as
nice as the VM/SP5 I love and remember..

> Peppe.
>

Dave

[Giuseppe Vitillaro [e-mail] [H390-VM]]

On Tue, 22 May 2018, 'Dave Wade' ***@gmail.com [H390-VM] wrote:

>
>
>> -----Original Message-----
>> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
>> Sent: 22 May 2018 10:02
>> To: Drew Derbyshire ***@kew.com [H390-VM] <H390-
>> ***@yahoogroups.com>
>> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer
> network
>> links, all different (was Re: VM/XA)
>>
>> On Mon, 21 May 2018, Drew Derbyshire ***@kew.com [H390-VM] wrote:
>>
>>> On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
>>> [H390-VM]
>>> wrote:
>>>> If IBM would allow to run a modern
>>>> VM version, from VM/SP to Z/VM, would be possible to rebuild an
>>>> hobbyist VNET like network?
>>>
>>> TLDR: Why would you need a modern VM?
>>
>> Because I'm as naive as possible about VM? ;-)
>>
>> I'm slowing realizing VM/370 is enough ;-)
>>
>> Still I dream about at least VM/SP or VM/XA, the VM versions I had been a
>> user for ten years ;-)
>>
>
> Me too. Even with all its tweaks and add-ons the six-pack is not quite as
> nice as the VM/SP5 I love and remember..
>
>> Peppe.
>>
>
> Dave
>
>
Me too ;-)

Definitely Dave. I tried VM/370 just for a while and
just to get MVS3.8j running in a virtual machine (beside
that boring problem of 2k pages) and it should be enough
for an RSCS VNET-like network.

What I mostly lacked had been XEDIT and REXX, but
this is a problem which may be overcome, if I remember
correctly.

Peppe.

['Dave Wade' [e-mail] [H390-VM]]

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 22 May 2018 16:25
> To: 'Dave Wade' ***@gmail.com [H390-VM] <H390-
> ***@yahoogroups.com>
> Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer
network
> links, all different (was Re: VM/XA)
>
> On Tue, 22 May 2018, 'Dave Wade' ***@gmail.com [H390-VM]
> wrote:
>
> >
> >
> >> -----Original Message-----
> >> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> >> Sent: 22 May 2018 10:02
> >> To: Drew Derbyshire ***@kew.com [H390-VM] <H390-
> ***@yahoogroups.com>
> >> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer
> > network
> >> links, all different (was Re: VM/XA)
> >>
> >> On Mon, 21 May 2018, Drew Derbyshire ***@kew.com [H390-VM] wrote:
> >>
> >>> On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
> >>> [H390-VM]
> >>> wrote:
> >>>> If IBM would allow to run a modern VM version, from VM/SP to
> >>>> Z/VM, would be possible to rebuild an hobbyist VNET like network?
> >>>
> >>> TLDR: Why would you need a modern VM?
> >>
> >> Because I'm as naive as possible about VM? ;-)
> >>
> >> I'm slowing realizing VM/370 is enough ;-)
> >>
> >> Still I dream about at least VM/SP or VM/XA, the VM versions I had
> >> been a user for ten years ;-)
> >>
> >
> > Me too. Even with all its tweaks and add-ons the six-pack is not quite
> > as nice as the VM/SP5 I love and remember..
> >
> >> Peppe.
> >>
> >
> > Dave
> >
> >
> Me too ;-)
>
> Definitely Dave. I tried VM/370 just for a while and just to get MVS3.8j
> running in a virtual machine (beside that boring problem of 2k pages) and
it
> should be enough for an RSCS VNET-like network.
>
> What I mostly lacked had been XEDIT and REXX, but this is a problem which
> may be overcome, if I remember correctly.
>

Not quite. There is now a full screen editor, and an implementation of REXX
but whilst they allow some things to run they are not the same as the
originals...

> Peppe.
>
Dave

>
> ------------------------------------
> Posted by: Giuseppe Vitillaro <***@vitillaro.org>
> ------------------------------------
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>

[Giuseppe Vitillaro [e-mail] [H390-VM]]

On Tue, 22 May 2018, 'Dave Wade' ***@gmail.com [H390-VM] wrote:

>
>
>> -----Original Message-----
>> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
>> Sent: 22 May 2018 16:25
>> To: 'Dave Wade' ***@gmail.com [H390-VM] <H390-
>> ***@yahoogroups.com>
>> Subject: RE: [H390-VM] You're in a maze of twisty little peer-to-peer
> network
>> links, all different (was Re: VM/XA)
>>
>> On Tue, 22 May 2018, 'Dave Wade' ***@gmail.com [H390-VM]
>> wrote:
>>
>>>
>>>
>>>> -----Original Message-----
>>>> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
>>>> Sent: 22 May 2018 10:02
>>>> To: Drew Derbyshire ***@kew.com [H390-VM] <H390-
>> ***@yahoogroups.com>
>>>> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer
>>> network
>>>> links, all different (was Re: VM/XA)
>>>>
>>>> On Mon, 21 May 2018, Drew Derbyshire ***@kew.com [H390-VM] wrote:
>>>>
>>>>> On 05/20/2018 01:25 PM, Giuseppe Vitillaro ***@vitillaro.org
>>>>> [H390-VM]
>>>>> wrote:
>>>>>> If IBM would allow to run a modern VM version, from VM/SP to
>>>>>> Z/VM, would be possible to rebuild an hobbyist VNET like network?
>>>>>
>>>>> TLDR: Why would you need a modern VM?
>>>>
>>>> Because I'm as naive as possible about VM? ;-)
>>>>
>>>> I'm slowing realizing VM/370 is enough ;-)
>>>>
>>>> Still I dream about at least VM/SP or VM/XA, the VM versions I had
>>>> been a user for ten years ;-)
>>>>
>>>
>>> Me too. Even with all its tweaks and add-ons the six-pack is not quite
>>> as nice as the VM/SP5 I love and remember..
>>>
>>>> Peppe.
>>>>
>>>
>>> Dave
>>>
>>>
>> Me too ;-)
>>
>> Definitely Dave. I tried VM/370 just for a while and just to get MVS3.8j
>> running in a virtual machine (beside that boring problem of 2k pages) and
> it
>> should be enough for an RSCS VNET-like network.
>>
>> What I mostly lacked had been XEDIT and REXX, but this is a problem which
>> may be overcome, if I remember correctly.
>>
>
> Not quite. There is now a full screen editor, and an implementation of REXX
> but whilst they allow some things to run they are not the same as the
> originals...
>
>> Peppe.
>>

My rough english is not enough to cope with
my intentions :-) "overcome" isn't apparently
the right word. The integration of XEDIT and REXX
is one of my best memories of the VM world.

By the way, I laso lack LPEX, if I remember
the name of that wonderful editor.

Peppe.

['Derbyshire, Drew' [e-mail] [H390-VM]]

To paraphrase an old Hertz car rental campaign -- "There's VM/SP ... and
not exactly."

> _._,___
>



--
Drew Derbyshire
*Assistant Cat Valet (Probationary)*
Telephone [phone]

[[e-mail] [H390-VM]]

"In America they haven't spoken it in years!" -- Professor Henry Higgins, "My Fair Lady".

btw, I like the word "overcome" in this context. That's how I feel using VM/370 R6.

As for your English in general, you're fine. I was born hearing it, and I've been screwing it up since before CP was written. :-)

If hunting for a word, consider http://www.thesaurus.com/ http://www.thesaurus.com/ -- I keep the link on my bookmarks bar. (Not that it always helps -- I fed it overcome and I had to scroll for the correct meaning.)

-ahd-

Drew Derbyshire

[Giuseppe Vitillaro [e-mail] [H390-VM]]

On Wed, 23 May 2018, ***@kew.com [H390-VM] wrote:

> "In America they haven't spoken it in years!" -- Professor Henry Higgins, "My Fair Lady".
>
> btw, I like the word "overcome" in this context. That's how I feel using VM/370 R6.

That's definitely nice ;-)

>
> As for your English in general, you're fine. I was born hearing it, and I've been screwing it up since before CP was written. :-)

Well, I know I can read, I can write, I can speak and ... occasionally
.... I may even understand spoken english, especially newyorkers american
english (tongue in cheek ;-) ).

But my feeling, speaking or writing in english, is likely
being in the need to upload a 1Gb file using a 9600 baud
modem ... I think you understand what I mean ;-)

>
> If hunting for a word, consider http://www.thesaurus.com/ http://www.thesaurus.com/ -- I keep the link on my bookmarks bar. (Not that it always helps -- I fed it overcome and I had to scroll for the correct meaning.)

Thanks for the link!

Peppe.

[[e-mail] [H390-VM]]

> The integration of XEDIT and REXX
> is one of my best memories of the VM world.

See Rich Alderson's announcement about the LCM+L 4361. The 4361 is running VM/SP 5, which has REXX and XEDIT.

[Peter Coghlan [e-mail] [H390-VM]]

> > If IBM would allow to run a modern
> > VM version, from VM/SP to Z/VM, would
> > be possible to rebuild an hobbyist VNET
> > like network?
>
> TLDR: Why would you need a modern VM?
>
> IBM had peer to peer connections for 239 systems in 1979, per
> https://en.wikipedia.org/wiki/RSCS#Background.  Those would have been
> VM/370 R6 systems or older. The basic concept of store and forward via
> the RSCS and the VM spool really didn't change in 20 years.
>
> The Wiki link above has a comment about how the public released version
> of RSCS had stubs for store & forward functionality. That makes me
> wonder how hard it would be to add peer-to-peer store & forward to the
> VM/370 R6 RSCS version.
>

I can confirm that it is possible and it has been done (mostly). The
peer-to-peer store and forward of files part is pretty much complete but
the processing of messages and commands is not quite there yet.

>
> The thing about VM was how CP was always distinct from the VM instances
> which ran under it. Even now you could probably take VM/SP RSCS Version
> 1 R3 and get run it running on from anything from VM/370 R6 up through
> z/VM.  (Note that I have /not/ tried this.)
>

Quite possibly. However, I am not in a position to try this either.

>
> A useful reference for RSCS behavior under later VM versions would be
> https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemNetworkingPro_13960888.
>
> I can't seem to find a definition of the actual RSCS protocols, but
> foreign programs like JNET (VMS) and UREP (UNIX)  implemented them, so
> it must be possible.  Or implement a Hercules specific BSC driver (VMH?).
>

There are various publications available such as SA22-7539-02 and SC23-0070-03,
both titled "Network Job Entry Formats and Protocols". I think I downloaded
them from the IBM website.

>
>         * * *
> Now, to answer the obvious question before it's asked ... how does one
> invite MVS 3.8 to the party:
>
> That's harder.
>

Someone who knows MVS much better that I do looked into this and agrees that
it is harder.

> The NJE protocol */is/* documented:
> https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1.hasa600/toc.htm
>
> I have to admit, doing MVS 3.8 NJE support would make me nervous/.*
>
> */Doing Store & Forward under VM requires a single virtual machine that
> supports RSCS protocols and a spool that already supports tagging. There
> is not anything break outside the reader queue of the RSCS machine,
> Compare this to NJE store & forward support on MVS, which requires you
> get into the guts of JES2. If (when?) you screw up and your spool is toast.
>
> (That reminds of an old story of a HASP spool that was reported at 110%
> utilization.  The next thing that happened, not surprisingly, was a
> system crash.)
>

There are two other possibilities.

Firstly, RSCS can exchange VM spool files with MVS running in a virtual
machine. Probably not very satisfactory but should work as a last resort.

Secondly, RSCS can do RJE to a BSC line using the DMTSML driver and I am told
MVS can also do RJE to a BSC line. It should be possible to use a BSC line
emulated by Hercules as the last leg to link MVS to an RSCS NJE store and
forward network. Some horrible kludges may be necessary to allow MVS to
communicate with nodes more than one hop away but it should be a lot easier
than implementing full NJE in MVS. It can also be used as a stepping stone
to doing it "properly".

> -ahd-
>
> p.s. Is the non-store & forward RSCS on the six pack?
>

Yes. All the source is there. It includes the DMTSML driver which can do
some variants of RJE over a BSC line.

From reading "NJE Formats and Protocols", it becomes clear that NJE is mostly
just another layer of protocol placed on top of RJE. The lower levels of both
are pretty much identical. I have written an update for the DMTSML link
driver that turns it into an NJE link driver. The latent support for routing
present in the base RSCS allows store and forward to work using the VM spool.
It is less clear to me how to handle routing of NMRs (NJE messages and
commands) in a way that is compatible with handling RJE messages and commands
so I am stalled on that part at the moment.

It turns out that the BSC lines emulated by Hercules are not quite ideal
for NJE links between RSCS nodes. I have a modified Hercules device which
looks like a 2703 on the inside and implements NJE over IP to the outside
world. The NJE over IP protocol (also known as VMNET but not to be confused
with something different in Hercules which was also called VMNET) was
developed to allow the BITNET network to operate over TCP/IP. IBM adopted
the same protocol for it's implementation of NJE over TCP/IP so using this
protocol has the added advantage that it is possible to interoperate with
later versions of RSCS and MVS which are natively capable of NJE over IP,
as well as JNET and UREP as mentioned previously. There is also the freely
available HUJI-NJE which was designed to implement NJE on VMS and Unix.
With a fair bit of work (by someone other than me), it should be possible
to use this to implement NJE on Linux.

Regards,
Peter Coghlan

Ps: I think this project is somewhere around state 1b on the list below :-)

>
> --
> Drew Derbyshire
>
> Google Voice Telephone: [phone] (NOT for text messages!)
>
> The six stages of a project:
> 1. Unbounded enthusiasm 4. Frantic Search for the Guilty
> 2. Total disillusionment 5. Punishment of the Innocent
> 3. Panic 6. Reward of the non-participants
>


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

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


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

Yahoo Groups Links

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/H390-VM/

<*> Your email settings:
Individual Email | Traditional

<*> To change settings online go to:
http://groups.yahoo.com/group/H390-VM/join
(Yahoo! ID required)

<*> To change settings via email:
H390-VM-***@yahoogroups.com
H390-VM-***@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
H390-VM-***@yahoogroups.com

<*> Your use of Yahoo Groups is subject to:
https://info.yahoo.com/legal/us/yahoo/utos/terms/

[Ivan Warren [e-mail] [H390-VM]]

Le 5/22/2018 à 1:35 PM, Peter Coghlan ***@beyondthepale.ie
[H390-VM] a écrit :
>> > If IBM would allow to run a modern
>> > VM version, from VM/SP to Z/VM, would
>> > be possible to rebuild an hobbyist VNET
>> > like network?
>>
I'd be willing to participate in a NJE network of some sort !

--Ivan

[Mark Morgan Lloyd [e-mail] [H390-VM]]

On 22/05/18 12:45, Ivan Warren ***@vmfacility.fr [H390-VM] wrote:
>
>
> Le 5/22/2018 à 1:35 PM, Peter Coghlan ***@beyondthepale.ie
> [H390-VM] a écrit :
>>>   > If IBM would allow to run a modern
>>>   > VM version, from VM/SP to Z/VM, would
>>>   > be possible to rebuild an hobbyist VNET
>>>   > like network?
>>>
> I'd be willing to participate in a NJE network of some sort !

But could it replace Yahoo?

--
Mark Morgan Lloyd
markMLl .AT. telemetry.co .DOT. uk

[Opinions above are the author's, not those of his employers or colleagues]

['Dave Wade' [e-mail] [H390-VM]]

> -----Original Message-----
> From: H390-***@yahoogroups.com <H390-***@yahoogroups.com>
> Sent: 23 May 2018 10:42
> To: h390-***@yahoogroups.com
> Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer network
> links, all different (was Re: VM/XA)
>
> On 22/05/18 12:45, Ivan Warren ***@vmfacility.fr [H390-VM] wrote:
> >
> >
> > Le 5/22/2018 à 1:35 PM, Peter Coghlan ***@beyondthepale.ie
> > [H390-VM] a écrit :
> >>> > If IBM would allow to run a modern
> >>> > VM version, from VM/SP to Z/VM, would
> >>> > be possible to rebuild an hobbyist VNET
> >>> > like network?
> >>>
> > I'd be willing to participate in a NJE network of some sort !
>
> But could it replace Yahoo?
>


I was going to say a couple of stone masons with chisels and UPS could replace Yahoo, but I guess that UPS could break tablets of stone into gravel.....
... sadly the listserve software isn't free either

> --
> Mark Morgan Lloyd
> markMLl .AT. telemetry.co .DOT. uk
>

Dave

> [Opinions above are the author's, not those of his employers or colleagues]
>
>
> ------------------------------------
> Posted by: Mark Morgan Lloyd <markMLl.h390-***@telemetry.co.uk>
> ------------------------------------
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>

[Guy Sotomayor Jr [e-mail] [H390-VM]]

> On May 23, 2018, at 6:08 AM, 'Dave Wade' ***@gmail.com [H390-VM] <H390-***@yahoogroups.com> wrote:
>
>
>
> > -----Original Message-----
> > From: H390-***@yahoogroups.com <mailto:H390-***@yahoogroups.com> <H390-***@yahoogroups.com <mailto:H390-***@yahoogroups.com>>
> > Sent: 23 May 2018 10:42
> > To: h390-***@yahoogroups.com <mailto:h390-***@yahoogroups.com>
> > Subject: Re: [H390-VM] You're in a maze of twisty little peer-to-peer network
> > links, all different (was Re: VM/XA)
> >
> > On 22/05/18 12:45, Ivan Warren ***@vmfacility.fr <mailto:***@vmfacility.fr> [H390-VM] wrote:
> > >
> > >
> > > Le 5/22/2018 à 1:35 PM, Peter Coghlan ***@beyondthepale..ie <mailto:***@beyondthepale.ie>
> > > [H390-VM] a écrit :
> > >>> > If IBM would allow to run a modern
> > >>> > VM version, from VM/SP to Z/VM, would
> > >>> > be possible to rebuild an hobbyist VNET
> > >>> > like network?
> > >>>
> > > I'd be willing to participate in a NJE network of some sort !
> >
> > But could it replace Yahoo?
> >
>
> I was going to say a couple of stone masons with chisels and UPS could replace Yahoo, but I guess that UPS could break tablets of stone into gravel......
> ... sadly the listserve software isn't free either
>
>
A friend of mine who owns/runs a hosting service (it’s the one where the bitsavers.org <http://bitsavers.org/> master is located) is willing to host the various Hercules lists for free. All we need to do is for the appropriate folks to contact him (contact me off list) and get it done.

TTFN - Guy

[Tony Harminc [e-mail] [H390-VM]]

On 23 May 2018 at 09:08, 'Dave Wade' ***@gmail.com wrote:
>
>> On 22/05/18 12:45, Ivan Warren ***@vmfacility.fr [H390-VM] wrote:
>> >
>> > Le 5/22/2018 à 1:35 PM, Peter Coghlan ***@beyondthepale.ie a écrit :
>> >>> > If IBM would allow to run a modern VM version, from VM/SP to Z/VM, would
>> >>> > be possible to rebuild an hobbyist VNET like network?
>> >>>
>> > I'd be willing to participate in a NJE network of some sort !
>>
>> But could it replace Yahoo?
>
> I was going to say a couple of stone masons with chisels and UPS could replace Yahoo, but I guess that UPS could break tablets of stone into gravel......
> ... sadly the listserve software isn't free either

Hmmm... The mainstream LISTSERV that was popular in the 1980s was the
Eric Thomas version. That, doubtless much rewritten, is the current
LSOFT product. But... Eric built that on something called Revised
Listserv, written by I'm not sure who, but certainly in REXX. And
before that there was (not Revised) Listserv, and surely that was not
anyone's program product. Maybe on an old VM Workshop tape? Sadly,
most of those old tapes are no longer around, despite my periodic
nagging of the tape maintainers. :-(

Tony H.

[Drew Derbyshire [e-mail] [H390-VM]]

On 5/22/18 4:35 AM, Peter Coghlan ***@beyondthepale.ie
[H390-VM] wrote:
>> > If IBM would allow to run a modern
>> > VM version, from VM/SP to Z/VM, would
>> > be possible to rebuild an hobbyist VNET
>> > like network?
>>
>> TLDR: Why would you need a modern VM?
>>
>> IBM had peer to peer connections for 239 systems in 1979, per
>> https://en.wikipedia.org/wiki/RSCS#Background.  Those would have been
>> VM/370 R6 systems or older. The basic concept of store and forward via
>> the RSCS and the VM spool really didn't change in 20 years.
>>
>> The Wiki link above has a comment about how the public released version
>> of RSCS had stubs for store & forward functionality. That makes me
>> wonder how hard it would be to add peer-to-peer store & forward to the
>> VM/370 R6 RSCS version.
>>
> I can confirm that it is possible and it has been done (mostly). The
> peer-to-peer store and forward of files part is pretty much complete but
> the processing of messages and commands is not quite there yet.
Messages are sort of important because they also provide file forward
status.
>> Even now you could probably take VM/SP RSCS Version
>> 1 R3 and get run it running on from anything from VM/370 R6 up through
>> z/VM.  (Note that I have /not/ tried this.)
>>
> Quite possibly. However, I am not in a position to try this either.
>
>> A useful reference for RSCS behavior under later VM versions would be
>> https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemNetworkingPro_13960888.
>>
>> I can't seem to find a definition of the actual RSCS protocols, but
>> foreign programs like JNET (VMS) and UREP (UNIX)  implemented them, so
>> it must be possible.  Or implement a Hercules specific BSC driver (VMH?).
>>
> There are various publications available such as SA22-7539-02 and SC23-0070-03,
> both titled "Network Job Entry Formats and Protocols". I think I downloaded
> them from the IBM website.
SC23-0070-03 looks more applicable.
>> Now, to answer the obvious question before it's asked ... how does one
>> invite MVS 3.8 to the party:
>>
>> That's harder.
>>
> Someone who knows MVS much better that I do looked into this and agrees that
> it is harder.
>
>> The NJE protocol */is/* documented:
>> https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.1.0/com.ibm.zos.v2r1.hasa600/toc.htm
>>
>> Doing Store & Forward under VM requires a single virtual machine that
>> supports RSCS protocols and a spool that already supports tagging. There
>> is not anything break outside the reader queue of the RSCS machine,
>> Compare this to NJE store & forward support on MVS, which requires you
>> get into the guts of JES2. If (when?) you screw up and your spool is toast.
>>
> There are two other possibilities.
>
> Firstly, RSCS can exchange VM spool files with MVS running in a virtual
> machine. Probably not very satisfactory but should work as a last resort.
>
> Secondly, RSCS can do RJE to a BSC line using the DMTSML driver and I am told
> MVS can also do RJE to a BSC line. It should be possible to use a BSC line
> emulated by Hercules as the last leg to link MVS to an RSCS NJE store and
> forward network. Some horrible kludges may be necessary to allow MVS to
> communicate with nodes more than one hop away but it should be a lot easier
> than implementing full NJE in MVS. It can also be used as a stepping stone
> to doing it "properly".
It comes down to so long is MVS 3.8 is leaf node, it's easier.  Then it
can say "ever node I don't know, go HERE".  This would actually be
better than the NJE enabled MVS/SP, which simply dropped all unknown
nodes on the floor.

Of course, NJE enabled MVS/SP had other issues, like no native support
to message off node, and hacks to support it tended to generate your
entire chat transcript in the system log -- not good for young college
students in love in the BitNET era.  (TUCC's VM/XA box did this).

I think in summary, it is best to put simply MVS aside until VM works.
>> p.s. Is the non-store & forward RSCS on the six pack?
>>
> Yes. All the source is there. It includes the DMTSML driver which can do
> some variants of RJE over a BSC line.
>
> From reading "NJE Formats and Protocols", it becomes clear that NJE is mostly
> just another layer of protocol placed on top of RJE. The lower levels of both
> are pretty much identical. I have written an update for the DMTSML link
> driver that turns it into an NJE link driver. The latent support for routing
> present in the base RSCS allows store and forward to work using the VM spool.
> It is less clear to me how to handle routing of NMRs (NJE messages and
> commands) in a way that is compatible with handling RJE messages and commands
> so I am stalled on that part at the moment.
>
> It turns out that the BSC lines emulated by Hercules are not quite ideal
> for NJE links between RSCS nodes. I have a modified Hercules device which
> looks like a 2703 on the inside and implements NJE over IP to the outside
> world. The NJE over IP protocol (also known as VMNET but not to be confused
> with something different in Hercules which was also called VMNET) was
> developed to allow the BITNET network to operate over TCP/IP. IBM adopted
> the same protocol for it's implementation of NJE over TCP/IP so using this
> protocol has the added advantage that it is possible to interoperate with
> later versions of RSCS and MVS which are natively capable of NJE over IP,
> as well as JNET and UREP as mentioned previously. There is also the freely
> available HUJI-NJE which was designed to implement NJE on VMS and Unix.
> With a fair bit of work (by someone other than me), it should be possible
> to use this to implement NJE on Linux.
I hope you also kept the original device (or that the new one is
compatible with the original on the other end), as there will be people
like me who need to rely on the stock behavior.
> Ps: I think this project is somewhere around state 1b on the list below :-)
That was a random selection, but seemed accurate.  I wondered who would
notice.

Today's is less random.
>> The six stages of a project:
>> 1. Unbounded enthusiasm 4. Frantic Search for the Guilty
>> 2. Total disillusionment 5. Punishment of the Innocent
>> 3. Panic 6. Reward of the non-participants

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

P-1 CUR ALLOC 20193.....5804M CALL GREGORY

[Rhialto [e-mail] [H390-VM]]

On Tue 22 May 2018 at 10:33:28 -0700, Drew Derbyshire ***@kew.com [H390-VM] wrote:
> Messages are sort of important because they also provide file forward
> status.

And for the Interchat Relay Network of course!

-Olaf.
--
___ Olaf 'Rhialto' Seibert -- Wayland: Those who don't understand X
\X/ rhialto/at/falu.nl -- are condemned to reinvent it. Poorly.

[Drew Derbyshire [e-mail] [H390-VM]]

On 5/19/18 1:53 AM, 'Dave Wade' ***@gmail.com [H390-VM] wrote:
> If you are interested in XA and the problems and issues it caused this
> is a
>
> good read:-
>
> http://vm.marist.edu/~vmshare/browse?fn=VMXASP&ft=MEMO
>
> I believe that VM/SP6 was supported as long as XA. IBM was still
> announcing
> enhancements to SP/6 AFTER XA/SP was announced
>
> http://www-01.ibm.com/common/ssi/rep_ca/2/877/ENUSZP89-0342/
>
> and still supporting folks on old hardware...
>

Given that IBM was selling 24-bit machines (9370, 43xx) into the late
1980's and the long tail of older systems, VM/SP 6 being supported after
VM/XA was announced was to be expected.

IBM has always supported its older machines reasonably well. Clarkson
University was running a S/360-65J through 1980.  And after they
decommissioned it, the local Alcoa plant had a machine with 2314's  they
could use to /still /access data. The IBM S/370-165 was legendary for
being announced without virtual memory and then being too expensive to
upgrade.  So 165's running MVT were common well after MVS was available.

Finally, the spring 1988 decommit of IBM from the general availability
of VM/XA SP 1 did not encourage migration to it.

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

"Beware of bugs in the above code; I have only proved it correct, not
tried it." -- Donald Knuth

['John P. Hartmann' [e-mail] [H390-VM]]

VNET was exclusively for IBM employees.

There is still an RSCS network, Bitnet.

On 05/20/2018 07:25 PM, Giuseppe Vitillaro ***@vitillaro.org
[H390-VM] wrote:
> If IBM would allow to run a modern
> VM version, from VM/SP to Z/VM, would
> be possible to rebuild an hobbyst VNET
> like network?


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

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


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

Yahoo Groups Links

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/H390-VM/

<*> Your email settings:
Individual Email | Traditional

<*> To change settings online go to:
http://groups.yahoo.com/group/H390-VM/join
(Yahoo! ID required)

<*> To change settings via email:
H390-VM-***@yahoogroups.com
H390-VM-***@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
H390-VM-***@yahoogroups.com

<*> Your use of Yahoo Groups is subject to:
https://info.yahoo.com/legal/us/yahoo/utos/terms/

[Ivan Warren [e-mail] [H390-VM]]

Le 7/17/2018 à 7:11 PM, 'John P. Hartmann' ***@gmail.com
[H390-VM] a écrit :
> VNET was exclusively for IBM employees.
>
> There is still an RSCS network, Bitnet.
>
>
Isn't VNET the TCPNJE protocol ?

--Ivan

[Beth [e-mail] [H390-VM]]

Bitnet is still alive??

On 7/17/18 1:11 PM, 'John P. Hartmann' ***@gmail.com [H390-VM] wrote:
> VNET was exclusively for IBM employees.
>
> There is still an RSCS network, Bitnet.
>


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

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


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

Yahoo Groups Links

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/H390-VM/

<*> Your email settings:
Individual Email | Traditional

<*> To change settings online go to:
http://groups.yahoo.com/group/H390-VM/join
(Yahoo! ID required)

<*> To change settings via email:
H390-VM-***@yahoogroups.com
H390-VM-***@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
H390-VM-***@yahoogroups.com

<*> Your use of Yahoo Groups is subject to:
https://info.yahoo.com/legal/us/yahoo/utos/terms/

[Peter Coghlan [e-mail] [H390-VM]]

Ivan Warren wrote:
> Le 7/17/2018 à 7:11 PM, 'John P. Hartmann' ***@gmail.com [H390-VM] a écrit :
> > VNET was exclusively for IBM employees.
> >
> > There is still an RSCS network, Bitnet.
> >
> >
> Isn't VNET the TCPNJE protocol ?
>

The US Princeton University came up with TCPNJE to implement the BITNET II
network in the late 1980s:

ftp://ftp.funet.fi/pub/netinfo/CREN/brfc0002.text

I suspect IBM's VNET predated this and was probably implemented using bisync
lines.

Regards,
Peter Coghlan.


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

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


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

Yahoo Groups Links

<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/H390-VM/

<*> Your email settings:
Individual Email | Traditional

<*> To change settings online go to:
http://groups.yahoo.com/group/H390-VM/join
(Yahoo! ID required)

<*> To change settings via email:
H390-VM-***@yahoogroups.com
H390-VM-***@yahoogroups.com

<*> To unsubscribe from this group, send an email to:
H390-VM-***@yahoogroups.com

<*> Your use of Yahoo Groups is subject to:
https://info.yahoo.com/legal/us/yahoo/utos/terms/

[Drew Derbyshire [e-mail] [H390-VM]]

Per Wikipedia, the first BitNET connections actually date back to 1981,
which was probably RSCS V1R2 (which appears to have come out in 1980).
Being interstate connections, no doubt they were bisync links. RSCS V1R2
supported both bisync and CTC connections, and a multi-hop network topology.

VM/SP Peer-to-peer connectivity in RSCS V1R3 was publicly available by
1982, which did not seem to add major changes.

This RSCS is not (but may the child of) the earlier more limited beast
under VM/370 R6 which was also called RSCS.

Please refer to
https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemNetworkingPro_13960888
for the state of the RSCS art as it was in 1982.

One last note, quoting Wikipedia:
> The migration of the ARPANET to TCP/IP was officially completed onflag
> dayJanuary 1, 1983, when the new protocols were permanently activated
So 1981-1982 was too early for people to think of RSCS protocols over
TCP/IP.  :-)

-ahd-

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

Harris's Lament: All the good ones are taken

[Michael Short [e-mail] [H390-VM]]

I believe that ARPANET completed the change to TCP/IP in 1983, but there
were already
other TCP/IP activity before then. When I was at Clarkson University, we
were already using
FAL and BITNET II in 1982 and had been doing so for quite a while. We
linked to Princeton
and Yale for our gateways.

Mike

On Tue, Jul 17, 2018 at 5:12 PM Drew Derbyshire ***@kew.com [H390-VM] <
H390-***@yahoogroups.com> wrote:

>
>
> Per Wikipedia, the first BitNET connections actually date back to 1981,
> which was probably RSCS V1R2 (which appears to have come out in 1980).
> Being interstate connections, no doubt they were bisync links. RSCS V1R2
> supported both bisync and CTC connections, and a multi-hop network topology.
>
> VM/SP Peer-to-peer connectivity in RSCS V1R3 was publicly available by
> 1982, which did not seem to add major changes.
>
> This RSCS is not (but may the child of) the earlier more limited beast
> under VM/370 R6 which was also called RSCS.
>
> Please refer to
> https://archive.org/details/bitsavers_ibm370RSCSlingCommunicationsSubsystemNetworkingPro_13960888
> for the state of the RSCS art as it was in 1982.
>
> One last note, quoting Wikipedia:
>
> The migration of the ARPANET to TCP/IP was officially completed on flag
> day January 1, 1983, when the new protocols were permanently activated
>
> So 1981-1982 was too early for people to think of RSCS protocols over
> TCP/IP. :-)
>
> -ahd-
>
> --
> Drew Derbyshire
>
> Google Voice Telephone: [phone] (NOT for text messages!)
>
> Harris's Lament: All the good ones are taken
>
>
>

[Drew Derbyshire [e-mail] [H390-VM]]

On 7/17/18 6:45 PM, Michael Short ***@gmail.com [H390-VM]
wrote:
> I believe that ARPANET completed the change to TCP/IP in 1983, but
> there were already
> other TCP/IP activity before then. When I was at Clarkson University,
> we were already using
> FAL and BITNET II in 1982 and had been doing so for quite a while.
I believe that you have the wrong decade. Sorry.
> We linked to Princeton and Yale for our gateways.
When I returned to Clarkson in fall 1985, the only WAN connection was to
UCONNVM.  In fact, we lost the BITNet link when Gloria (late September
1985) caused U Conn to shutdown.

We didn't get any Internet at CU until 1986; Jeff Honig installed
omnigate (a MicroVax) as the gateway.

I moved away from Potsdam for the last time in 1987, so I can't say when
BITNet II came, but it was later.

-ahd-

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

Backups? We don't *NEED* no steenking baX%^~,VbKx
General Failure on D: Abort, Retry, Fail?

[Michael Short [e-mail] [H390-VM]]

After doing a little research on BITNET II, I see that its inception was in
1987. We obtained a
copy shortly after it was announced. If I remember correctly, it was fairly
easy to install and
integrate. We ran this configuration until BITNET's board decided to end
BITNET.

On Tue, Jul 17, 2018 at 10:32 PM Drew Derbyshire ***@kew.com [H390-VM] <
H390-***@yahoogroups.com> wrote:

>
>
> On 7/17/18 6:45 PM, Michael Short ***@gmail.com [H390-VM]
> wrote:
>
>
> I believe that ARPANET completed the change to TCP/IP in 1983, but there
> were already
> other TCP/IP activity before then. When I was at Clarkson University, we
> were already using
> FAL and BITNET II in 1982 and had been doing so for quite a while.
>
> I believe that you have the wrong decade. Sorry.
>
> We linked to Princeton and Yale for our gateways.
>
> When I returned to Clarkson in fall 1985, the only WAN connection was to
> UCONNVM. In fact, we lost the BITNet link when Gloria (late September
> 1985) caused U Conn to shutdown.
>
> We didn't get any Internet at CU until 1986; Jeff Honig installed omnigate
> (a MicroVax) as the gateway.
>
> I moved away from Potsdam for the last time in 1987, so I can't say when
> BITNet II came, but it was later.
>
> -ahd-
>
> --
> Drew Derbyshire
>
> Google Voice Telephone: [phone] (NOT for text messages!)
>
> Backups? We don't *NEED* no steenking baX%^~,VbKx
> General Failure on D: Abort, Retry, Fail?
>
>
>

[Gregg Levine [e-mail] [H390-VM]]

Hello!
I've come to a decision.
I'm going to allow this discussion to continue unless some sort of a
consensus is reached. And when that happens we're done. Now I imagine
that the one who reads this with an audience will want to ask why
about that, and so does the one who works from underwater, and
definitely the one who works with a statue staring at him. (Don't
blink!) And of course the even bigger question will surface as to who
I mean. That's already been covered.

Oddly enough Clarkson, which is based here in upper New York State,
was the distribution home for Packet Drivers for the early Internet. I
once met the writer behind most of them, now dedicates himself to
better open sourced map making.
-----
Gregg C Levine ***@gmail.com
"This signature fought the Time Wars, time and again."

On Tue, Jul 17, 2018 at 11:44 PM, Michael Short
***@gmail.com [H390-VM] <H390-***@yahoogroups.com> wrote:
>
>
>
> After doing a little research on BITNET II, I see that its inception was in 1987. We obtained a
> copy shortly after it was announced. If I remember correctly, it was fairly easy to install and
> integrate. We ran this configuration until BITNET's board decided to end BITNET.
>
> On Tue, Jul 17, 2018 at 10:32 PM Drew Derbyshire ***@kew.com [H390-VM] <H390-***@yahoogroups.com> wrote:
>>
>>
>>
>> On 7/17/18 6:45 PM, Michael Short ***@gmail.com [H390-VM] wrote:
>>
>>
>> I believe that ARPANET completed the change to TCP/IP in 1983, but there were already
>> other TCP/IP activity before then. When I was at Clarkson University, we were already using
>> FAL and BITNET II in 1982 and had been doing so for quite a while.
>>
>> I believe that you have the wrong decade. Sorry.
>>
>> We linked to Princeton and Yale for our gateways.
>>
>> When I returned to Clarkson in fall 1985, the only WAN connection was to UCONNVM. In fact, we lost the BITNet link when Gloria (late September 1985) caused U Conn to shutdown.
>>
>> We didn't get any Internet at CU until 1986; Jeff Honig installed omnigate (a MicroVax) as the gateway.
>>
>> I moved away from Potsdam for the last time in 1987, so I can't say when BITNet II came, but it was later.
>>
>> -ahd-
>>
>> --
>> Drew Derbyshire
>>
>> Google Voice Telephone: [phone] (NOT for text messages!)
>>
>> Backups? We don't *NEED* no steenking baX%^~,VbKx
>> General Failure on D: Abort, Retry, Fail?
> ________________________________
> Posted by: Michael Short <***@gmail.com>

[Drew Derbyshire [e-mail] [H390-VM]]

On 7/17/18 9:58 PM, Gregg Levine ***@gmail.com [H390-VM] wrote:
>
> Hello!
> I've come to a decision.
> I'm going to allow this discussion to continue unless some sort of a
> consensus is reached. And when that happens we're done. Now I imagine
> that the one who reads this with an audience will want to ask why
> about that, and so does the one who works from underwater, and
> definitely the one who works with a statue staring at him. (Don't
> blink!) And of course the even bigger question will surface as to who
> I mean. That's already been covered.
>
This is maybe the most confusing post I've seen on this list in a while.

The previous (RSCS dates) conversation was already over -- I didn't see
a reason to comment on Mike's revised dates.

--
Drew Derbyshire

Google Voice Telephone: [phone] (NOT for text messages!)

"Is this going to be a standup fight, sir, or another bughunt?"
-- Aliens

[Giuseppe Vitillaro [e-mail] [H390-VM]]

On Tue, 17 Jul 2018, Peter Coghlan ***@beyondthepale.ie [H390-VM] wrote:

> Ivan Warren wrote:
>> Le 7/17/2018 à 7:11 PM, 'John P. Hartmann' ***@gmail.com [H390-VM] a écrit :
>>> VNET was exclusively for IBM employees.
>>>
>>> There is still an RSCS network, Bitnet.
>>>
>>>
>> Isn't VNET the TCPNJE protocol ?
>>
>
> The US Princeton University came up with TCPNJE to implement the BITNET II
> network in the late 1980s:
>
> ftp://ftp.funet.fi/pub/netinfo/CREN/brfc0002.text
>
> I suspect IBM's VNET predated this and was probably implemented using bisync
> lines.
>

The timing match, Peter.

I joined IBM in the late 1985 and the VNET was a rather recent
innovation at that time, the conversion wasn't even completed
I think, and probably began somewhere between 1980 and 1983.

In 1991, when I left the IBM Rome Scientific Center (Rome, Italy)
to work here in Perugia (still as an IBMer), with
the local University as my main customer, VNET and BITNET
were actually the only large networks I had available
from Italy.

Only few italian Universities were connected to a "small"
Internet, at that time, and RSCS/DECNET networks were
actually the standards for my country, in the academic/scientific
world.

Guess who actually connected University of Perugia to Internet,
in the late 1991, with a first successfull "ping"
command to "ftp.mit.edu" from the IBM AIX/370 system
(running on a 3090/VEC under VM/XA) "ipgaix.unipg.it"? ;-)

Peppe.