Hi,
I've been running the Panda distribution of TOPS-20 on the KLH10 simulator on a Raspberry Pi 4 for some years now. It's WALACH on HECnet. The version of the Pi's OS had been the 32 bit implementation but as some of these were being replaced by 8Gb versions, I recently took the plunge and updated (by re-imaging) everything to the current 64 bit Raspberry Pi OS.
In rebuilding SIMH and KLH10, I sync'd with the latest sources. With SIMH, there were no issues but the KLH10 updates, seemingly post ~2021 have broken networking. The line in my klt20.ini file is:
devdef ni0 564 ni20 dedic=true ifc=tapX ifmeth=tap+bridge ipaddr=192.168.2.46 enaddr=aa:00:04:00:6c:74
With the dpni20 executable running as root, the 'old' behaviour causes it to use the default bridge name 'bridge0' and create a tap device 'tap0' (it disregards the ifc=tapX part of the line).
The new behaviour results in a whole load of IOCTL errors while attempting to create the tap device.
I avoided the problem by restoring the original version of the KLH10 simulator sources and compiling.
The whole thing has reminded me that I wanted to fix the code a while back so that it would make use of an existing tap device, created via /etc/network/interfaces, with full control granted to the non-privileged user the rest of KLH10 runs under, then dpni20 would not need to be privileged. Once I've finished with the upgrades I might have a go.... along with all of the other half-done/half-forgotten projects in this area.
Keith
Hi,
During my postgraduate days, I ran ANU NEWS on our VAX/VMS systems and we peered large parts of the hierarchy. I think even back then we were exclusively NNTP over TCP/IP. I have a vague memory that I wrote a multi-threaded NNTP server for ANU but it was using the CMUTEK TCP/IP product so it was all done with QIO and ASTs - none of this weird select stuff of Unix sockets.
With Google groups being hit by 1000s of spam messages a day - astrology, voodoo, death spells, etc. - I'm tempted to get ANU NEWS installed and running again, probably SIMH-hosted VAX OpenVMS and, as above, peer some of the groups. Disc space and bandwidth aren't an issue and I'd like to see NNTP running over DECnet.... And I'm more familiar with ANU NEWS's SMG interface than I ever was with the likes of tin and rm.
Actually, wanting to do this has less to do with Google's mess and more about nostalgia. If/when it's all up and running, I'll consider getting it running on some real Alpha or IA64 hardware, probably based mostly on which has the lower power consumption.
Just wondering if anybody else is doing/done something similar.
Keith
Version 5.3(235)-5 incorporates all fixes since the major release of
5.3, Edit 230 in January of this year. These are as follows and can be
identified in the code with the edit number as the prefix of a comment.
<about:blank?compose#_Toc137380940>
[231] Fix RECEIVE with no file name
[232] 36 bit byte file sizes
[233] Transaction Log and Debug log fixes and enhancements
Transaction logging
Garbled Text
Write-Protection Failures
Enhancements
Debug Log Decode Fix
[234] Error messages may not be seen if displayed in remote server
[235] Properly signal and handle file errors in server mode
Additional Batch Tests
K2036P: Kermit-20 36 Bit Mode via pseudo-terminal
K2036C: Kermit-20 36 Bit Mode via pseudo-terminal with parity
Updated Help
All source, documentation, executables, control files and test data are
available for Anonymous NFT on HECnet from VENTI2::PS:<OINKY.K20MIT> and
associated subdirectories. Documentation includes more detail on the
above list. All regression tests have passed except for Tops-10 because
of some local networking issues.
Be aware that VENTI2:: is running a FAL alpha candidate, so let me know
if you get unexpected behavior.
Hi,
Is anyone testing this mailing list's recipients?
The reason I ask is:
06-02 00:07:58 NOREC SMTP OKBL DE 65.108.62.138:39112 cassini.dfupdate.se. 65.108.0.0/16 24940 HETZNER-AS, DE
06-02 00:07:59 NOREC SMTP OKBL DE 65.108.62.138:39112 cassini.dfupdate.se. 65.108.0.0/16 24940 HETZNER-AS, DE
06-02 00:33:06 NOREC SMTP OKBL DE 136.243.112.216:42384 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:33:06 NOREC SMTP OKBL DE 136.243.112.216:42384 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:34:26 NOREC SMTP OKBL DE 65.108.62.138:40994 cassini.dfupdate.se. 65.108.0.0/16 24940 HETZNER-AS, DE
06-02 00:35:02 NOREC SMTP OKBL DE 136.243.112.216:36476 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:35:12 NOREC SMTP OKBL DE 136.243.112.216:36908 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:40:02 NOREC SMTP OKBL DE 136.243.112.216:36476 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:40:12 NOREC SMTP OKBL DE 136.243.112.216:36908 static.216.112.243.136.clients.your-server.de. 136.243.0.0/16 24940 HETZNER-AS, DE
06-02 00:45:56 NOREC SMTP OKBL DE 65.108.62.138:41946 cassini.dfupdate.se. 65.108.0.0/16 24940 HETZNER-AS, DE
06-02 00:57:11 NOREC SMTP OKBL DE 65.108.62.138:43082 cassini.dfupdate.se. 65.108.0.0/16 24940 HETZNER-AS, DE
Activity from cassini.dfupdate.se is valid and expected.
The other attempt(s) from static.216.112.243.136.clients.your-server.de was an attempted non-local relay to jessindewinter(a)gmail.com<mailto:jessindewinter@gmail.com>
It wasn't exactly a problem and the attempt was denied with some tar-pitting along the way.
Times in GMT+01:00 (BST) by the way.
Keith
Hey folks. Does anyone have images of the RSX-11M v4.8 RK06/RK07
kit? I have the RL kit, but I would like to have the RK06/RK07 kit.
I think Oleg may have it, but my mail server doesn't seem to be
talking to his mail server.
Thanks,
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
Hey folks, would anyone here happen to have the firmware files for
the Emulex Performance 2500 and Performance 4000 LAT terminal servers?
I really regret not having snagged a copy of that stuff when I
managed a bunch of those in the early 1990s. Wasn't thinking ahead I guess.
Thanks,
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I’m trying to get mail working from my latest Linux DECnet release to OopenVMS with the intention of eventually upgrading to the mail11v3 protocol. I’m running an up to date version of VAX OpenVMS V7.3 with all patches I’ve found on a recent download of OpenSimh on a really slow ARM system. The system has a default DECnet account as well as a MAIL$SERVER account. Whenever I try to send e-mail, the connection is rejected with a “No such user” error:
Auditable event: Network login failure
Event time: 29-NOV-2023 10:14:32.96
PID: 00000098
Process name: MAIL_8198
Username: DECNET
Remote node id: 42186 (41.202)
Remote username: Linux0000
Status: %LOGIN-F-NOSUCHUSER, no such user
I also get the same rejection error if I try to connect to FAL with no access control information. However, if I connect to FAL using the explicit nonpriviledged userid/password returned by NCP, everything works as expected.
Any ideas?
Thanks,
John.
Version 5.3(248)-5 incorporates additional features, enhancements and
fixes since the minor release of 5.3(235)-5, in June of this year
(11-Jun-2023).These are as follows and can be identified in the code
with the edit number as the prefix of a comment.
[236] Implement /FORKLESS connections for DECnet NRT’s
[237] Implement the RETURN keyword to stop processing a TAKE file
[238] Enhanced NRT disconnect messages
[239] Use .HPTOD (High Precision Time of Day) if the monitor has it
HPTIM% .HPTOD
[240] When logging to a disk file, display the number of bytes written
[241] Handle potential case of incorrect data being sent for CWD
[242] ECHO /TRANSACTION /DEBUG /SESSION Switches
[243] Fix off by one error which appended NUL characters in packets
[244] Fix server FINISH does not close all logs
[245] Rewrite MOVSTU to use EXTEND instead of loop
[246] Additional C escape sequences: \O and \U
[247] Rewrite IAC handling to use EXTEND instead of loop
[248] Fix incorrect string lengths being returned by C constant
expansion
Kermit-20-Testing-Battery-5.3(248)-5 Updates and new Batch Tests
K20DPD: Updated Kermit-20 Packet Decoding Example
K20NRV: Updated Kermit-20 Server functions via Tops-20 DECnet NRT to
local host
K20NRT: Updated Kermit-20 Server functions via Tops-20 DECnet NRT to
remote host
K20DPD: Updated Kermit-20 Packet Decoding Example
K20P8R: New Regression test against Kermit-20 4.2(174) [2-May-85] 8
bit files with parity
K20RDC: New Communications comparison Kermit-20 4.2(174) [2-May-85]
and 5.3(248)-5
K1036N: Kermit-10 36 bit Regression Tests via DECnet NRT
Updated Help
It is available for anonymous NFT from VENTI2:: in the K20MIT
sub-directory. The full announcement may be find in text format in:
VENTI2::TOMMYT:<OINKY.K20MIT.DOCUMENTATION>KERMIT-20_V5_3(248)-5-ANNOUNCEMENT.TXT.
Note that the left and right parenthesis will need to be quoted with a
Control-V. Alternatively, you can just grab the whole documentation
directory and have your local NFT do the naming.
Sometime around mid-2022 the Linux kernel developers decided to remove the DECnet code from the Linux
kernel. Kernel 6.0.x was the last release which included source code for a DECnet implementation. More
recently, there has been some discussion around removing the code from the Long Term Support kernels.
These changes mean that the repository I have been maintaining at
<https://github.com/JohnForecast/RaspbianDECnet> can no longer be installed on recent releases.
Over the past year or so, I have been working on a replacement for this repository with the following
characteristics:
- Designed to be built as an external kernel module
This simplifies and speeds up the installation since we no longer need to rebuild the entire
kernel.
- Can only be built as an ethernet endnode
Again this substantially simplifies the kernel code. The routing code was alway marked as
“experimental” and I never tried to get it running. If you need a DECnet router, pyDECnet or
Route20 are much better solutions.
- Minimize the use of Linux kernel frameworks
Many of the problems with keeping RaspbianDECnet running between Linux versions
were changes to the kernel framework APIs. By limiting the use of these APIs I am hoping
that the kernel module will need fewer changes to keep up with kernel changes. So far,
I have had to make one additional conditional code change at kernel 6.5 and that was in
the socket layer so all networking code would need to be changed.
In addition to a new kernel module there have been a number of changes to the userland code:
- What used to be “fal2” is now the default file access listener. The old “fal” is still available in the
“fal-old” directory
- What used to be “nml2” is now the default and only network management listener.
- There is now a subset implementation of “ncp” which is sufficient to support all of the requests
available from the Linux network management listener. It does implement a “tell” prefix so all
of these commands may be issued to remote systems. In addition, it implements
“ncp copy known nodes from <node name/address>” to update the local node name database.
- The new kernel module now supports node counters as defined in the Network Management
Specification and the ncp/nml combination are able to display and zero them.
- The installation procedure is no longer targeted solely at Raspberry Pi releases. The current
procedure supports installation on systems derived from Debian or Fedora although only a
limited number of distributions have been tested (see README.DECnet).
This new release is available at:
<https://github.com/JohnForecast/LinuxDECnet>
and should be considered to be in Beta test at this time. See README.DECnet in the top level directory
for more information and installation instructions.
Hey folks. I've gotten a Pro380 running at LSSM, and have just
gotten a DECNA board installed. Can anyone point me to any docs for
Pro/DECnet? I don't see much online. I'd like to get it talking to the
rest of the DECnet machines at the museum and on HECnet. And, ideally,
LAT if possible.
Thanks,
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
