| Newsgroups |
gmane.comp.emulators.hercules390.general |
| Organization |
TNet Consulting |
| Message-ID |
<[email protected]> |
On 4/17/19 9:13 PM, Ivan Warren [email protected] [hercules-390] wrote:
> Never be afraid to ask
I'm not afraid to ask.
I'm not afraid to wipe egg of my face either.
But I don't want to disturb or annoy people.
> CP (the VM Hypervisor) maintains 3 queues for each defined virtual
> machine : The Reader Queue, The Printer Queue, The Punch Queue. Each of
> these queues contain sets of cards and/or printed lines. They are each
> grouped in spool file which contain the cards/prints and some metadata.
Do the reader, printer, and punch have their own spool file? Or do they
share one spool file? (Per guest VM.)
> View each "spool file" as a box - with a box ID (Spool ID), a label(Tag),
> a class (a 1 character information), and is located in someone's drawer
> (spool queue).
Is it correct to say that each guest VM has a spool queue with the three
aforementioned spool files therein?
> You use the CP SPOOL <PRT|PUN> TO <userid> to indicate the disposition
> of the next "punch" or "print" output. <userid> can be yourself or
> "*" in which case it is sent to your own reader queue.
Is "*" a synonym for the current user? (In this context.)
I can see how something like that could make the JCL user agnostic. The
user that runs the JCL job gets the output sent to the "*" user ID.
Hypothetically.
> <userid> can be "SYSTEM" in which case it becomes eligible for actual
> physical punching (and put on the Punch queue) or printing (put on the
> print queue) - under certain conditions. When sending files via RSCS
> you simply do a 'CP SPOOL PUN TO RSCS' then punch your file.
ACK
I understand the difference between a printer and a card punch in the
physical world. But my mind is having trouble differentiating the two
in the virtual world.
I guess a printer is outputing for human consumption while punching is
outputing for computer consumption.
> You use the spool file Tag (Using the CP TAG Command) to give that
> information. RSCS looks at the spool file "TAG" metadata and then
> reacts accordingly. The TAG is an arbitrary string. When using RSCS when
> transfering a file through an NJE link, the tag basically contains the
> remote node to send it to followed by information to be relayed to the
> remote node for disposition.
>
> Specification of disposition of a transfered file using NJE is included
> in the NJE protocol. RSCS then knows what to do with the received file
> (to which VM User to send it to).
So for VNET / BITNET / other RSCS based networks, does each intermediary
system have to know how to get to the destination systems? Routing of
some sort?
If this is the case, I'd think that the SYSA could use RSCS to send a
file / job to SYSC via SYSB. As in SYSA would have a route to SYSC
through SYSB.
My mind is harking back to UUCP bang paths. As I understand it, you
either need source routing (bang paths) or intermediate routing (SYSA
knowing that SYSC is via SYSB).
> So basically, RSCS uses CP spool files metadata, no need for additional
> cards ;)
Cool.
I wasn't aware that such out-of-band data went with the spooled data. I
naively assumed it needed to be in-band like legacy job cards.
> On a side note, the CP Virtual reader is a bit magic and can turn
> "virtually printed files" into magical 132 columns cards and read them !
> This allows transfering printer listings to other users, other nodes,
> RSCS managed printers, RSCS managed RJE printers, etc..
Interesting.
Can those magical 132 column cards then be executed like normal 80
column cards?
Thank you for the explanation Ivan. :-)
--
Grant. . . .
unix || die
[Non-text portions of this message have been removed]