19 Oct
2020
19 Oct
'20
11:01 p.m.
L.S.
Currently I am finishing testing for the release of the last member of the Pdp8 series
computers: Simh Pdp8a with some new devices (and some from old not present in Ppd8).
Current state of affairs is, that it is running Os8 V3S 128k monitor and the Kt8
diagnostics and is stable.
Also F4 runs and basic Rts8 V2b and V3.
The last hurdle is to activate Decnet8 as the last test station to really load a Pdp8 with
some realistic real time work.
However ?.
Activating Decnet8 leads to a lot of funny but nasty surprises and is tedious at best.
Basic software can be found on the internet with only 2 flavors of Nip and Nsp which
differ only slightly.
The setup, assembling and loading is easy enough, though Nip02 uses a thing that is only
available in Rts8 V3 exec although it was released for V2b.
That should be catched with prerelease testing.
The initial point is already, that the Ddcmp lines are not starting up automatically.
Looking in the code there are pointers to things with temp status or not implemented, like
the STRTUP function code in Nsp which might be needed to startup the lines.
There are quite some discrepancies between current available Decnet8 and the documentation
in Decnet8.pdf as can be seen with the Nip report:
According to the Decnet8 doc copyrighted in 1977, the example display of status looks as
follows:
4.5 SAMPLE NIP PRINTOUT
The following is a typical NIP printout:
NETWORK INFORMATION PROGRAM V1
PHYSICAL LINE STATUS [DDCMP V4B]
LINE NODNAM (#) STATE DRIVER VRSN UP SINCE
0 ACTON (3) ON-LINE KL8J 3A 8:03:35 28-DEC-76
1 MAYNRD (2) OFF-LINE DKC8 2B
2 CONCRD (5) STARTING LOCAL 15A
LOGICAL LINK STATUS [NSP VllE]
CHAN STATE LINE TASK SRCE-NAME[PPN] TYPE LINK DEST-NAME[PPN] TYPE LINK
1 IN-USE 0 12 TARZAN[6,0] 10 4 JANE [3,1] 7 6
2 WAKING 1 3 DONALD[2,2] 3 3
3 UNUSED 1 3 DUMMY [4,4] 29 14
4 BOOT 2 6 MASS [1,23] 3 44
PHYSICAL LINE ERROR STATUS
LINE 0:
3 NAKS WITH HEADER CRC ERRORS
5 NAKS WITH DATA CRC ERRORS
83 NAKS WITH REP ERRORS
1 NAK WITH BUFFER UNAVAILABLE
2 NAKS WITH HEADER FORMAT ERRORS
192 NAKS WITH MESSAGES TOO LONG
100 MSGS WITH HEADER CRC ERRORS
1 MSG WITH REP ERROR
1 MSG WITH DATA OVERRUN
10 MSGS WITH HEADER FORMAT ERRORS
9 MSGS WITH MESSAGES TOO LONG
LINE 1: NO ERRORS
LINE 2:
3 NAKS WITH DATA OVERRUNS
1 MSG WITH REP ERROR
A real printout on the current generated system is as follows:
NETWORK INFORMATION PROGRAM V1C
PHYSICAL LINE STATUS [DDCMP V1A]
LINE NODNAM (#) STATE DRIVER VRSN UP SINCE
0 UNKNOWN***** OFF-LINE KL8J 1A
LOGICAL LINK STATUS [NSP V1C]
CHAN STATE LINE TASK SRCE-NAME[PPN] TYPE LINK DEST-NAME[PPN] TYPE LINK
1 UNUSED 0 36 TLK [1,2] 0 0
PHYSICAL LINE ERROR STATUS
LINE 0: NO ERRORS
0 MESSAGES TRANSMITTED
0 MESSAGES RECEIVED
The following oddities ? amongst more - can be seen:
1. The date is 28-dec-76, the various edit dates in the current software are all
centered on april/may-77.
2. Ddcmp is dated 7-apr-77 with version 1A, the doc mentiones already 4B
3. KL8 ISR is versioned 3A while the current available version is 1A and dated at
8-apr-77
4. And then the Nip version V1 as documented is older than the V1C version of the
internet release!!
And there are a lot more things going on. So the bottom question remains: has anyone seen
the internet versions working or is there a clobber up between an advanced laboratory
(Phase-II??) version as documented (look at the Ddcmp versions) and a preleased or prior
old Decnet version. And if so, can anything be retrieved or is it lost forever.
Best regards,
RV