You can obviously get into problems trying to figure out what to remove
for the local executor before the fact because you don't know who is
picking up the node file.
I believe that the correct thing to do when getting and processing the
node list is for the insertion application to 'politely' decline to
insert the local Executor.? Tops-20 will actually let you re-insert the
local Executor with no complaints, /provided/ that neither the local
Executor node name or node number have changed since boot.? However my
SETNODE rewrite is too 'polite' to even attempt that.? The below is from
the GETNOD weekly batch job log of 25-Sep-2021.
21:05:28 USER?? SETNODE>*Verbosity Terse
21:05:28 USER?? Verbosity level: TERSE
21:05:28 USER?? SETNODE>*Take SYSTEM:NODE-DATA.TXT.0 /NO-ACCESS
21:05:28 USER
21:05:28 USER?? [Fork SETND2 opening ? for reading]
21:05:28 USER *% Ignoring any modification to Executor node at line
number 849*
21:05:28 USER
21:05:28 USER?? SETNODE>*Verbosity Normal
21:05:28 USER?? Verbosity level: NORMAL
21:05:28 USER?? SETNODE>*Build Incremental
21:05:28 USER?? % No nodes in this table
21:05:28 USER?? Nodes processed: 7
21:05:28 USER?? Total nodes inserted: 7
21:05:28 USER?? SETNODE>*Build Full
21:05:28 USER?? Nodes processed: 899
21:05:28 USER?? Nodes processed: 7
21:05:28 USER?? Total nodes inserted: 906
21:05:28 USER?? SETNODE>*Status
21:05:28 USER?????? Total Unchanged nodes:??? 899
21:05:28 USER??? Total updates to perform:????? 7
21:05:28 USER?????? Total nodes to delete:?? NONE
21:05:28 USER??? Total words, shared text:?? 1632
21:05:28 USER?? Total incremental changes:????? 7
21:05:28 USER???????? Total nodes to save:??? 906
21:05:28 USER?? SETNODE>*Show Delete
21:05:28 USER?? % No nodes in this table
21:05:28 USER?? SETNODE>*Show Update
21:05:28 USER?? DURAL?? 29.801? IRNBRU? 29.806? KATANA 29.803? NOISY??
29.804? NOVA??? 29.800
21:05:28 USER?? PRLVAX? 29.805? TOOTH?? 29.802
21:05:28 USER?? Total nodes in table: 7
21:05:28 USER?? SETNODE>*Insert Updates
21:05:28 USER?? [Inserted 7 nodes]
21:05:28 USER?? SETNODE>*Save /NO-ACCESS /RESTRICTED
21:05:28 USER
21:05:28 USER?? [Fork SETND2 opening ? for writing restricted]
21:05:28 USER?? [Saving to binary file: TOMMYT:<SYSTEM>NODE-DATA.BIN.1
;PROCESS-ONLY-JFN:2? ;RD ;WR ;BSZ:36 ]
21:05:28 USER?? SETNODE>*Exit
------------------------------------------------------------------------
On 11/11/21 6:30 AM, Johnny Billquist wrote:
Tony,
On 2021-11-11 10:30, Tony Nicholson wrote:
Johnny,
Under RSTS/E I?ve found I can use FIX.CMD, but I have to comment the
definition for the current node and add a missing definition for MIM.
Ah. Yes, similar "problem" as with RSX in that the own node won't
work. That's the reason MIM is not in there. I should create a
different one for others. FIX.CMD is where I started, and it is very
tailored to my own use.
Then duplicate the definitions changing the ?def
nod? to ?set nod?
and feed it into NCP.
Maybe you should look at FIX.COM?
When I update pydecnet using a new nodenames.dat
I also use
$ ncp copy known nodes from TNRTR to both
(TNRTR Is my pydecnet router)
That definitely works in VMS. Does that also work in RSTS/E?
With that said, while generated command files are all nice and good,
for at least VMS and RSX, I would actually recommend people use the
provided DECnet tools instead.
In RSX:
RUN LB:[5,54]NNCFSL
and collect the nodes from MIM, and then:
SCP STOP
SCP START
and you have the new nodenames in there.
In VMS (as mentioned above):
$ NCP COPY KNOWN NODES FROM MIM TO BOTH
If the VMS solution also works in RSTS/E that is really nice.
? Johnny
Tony
On Thu, 11 Nov 2021 at 19:54, Johnny Billquist <bqt at softjar.se
<mailto:bqt at softjar.se>> wrote:
??? If a specific format is needed for RSTS/E nodes, I can create that
??? format file whenever the nodename database is updated. Just let
me know
??? what the format should be.
??? At the moment, the following files are created:
??? FIX.T20 (for TOPS-20)
???
FIX.COM <http://FIX.COM> (for VMS)
??? HECNET.PY (for PyDECnet)
??? FIX.CMD (for RSX)
??? All can be found at MIM::HECNET:
???? ? ?Johnny
--
Tony Nicholson <tony.nicholson at
computer.org
<mailto:tony.nicholson at computer.org>>