This sounds promising! I am interested in at least running some experiments to find out how well it all works, and to learn enough to configure my future little patch of DECnet. I'd like to have fun participating, but not if it causes problems for my upstream node or HECnet in general. I know that my connectivity situation is unusual. I am in southern California as my callsign suggests, just outside Riverside. Is there any possibility of connecting through something like an SSH tunnel, and relying on that for a higher level of security and authentication for the link than the DECnet protocols natively provide? If I'd like to eventually connect several nodes on my local network into HECnet, would I need to consume a whole area number? Or is there a means for further routing within an existing HECnet area? I still have much to learn about DECnet. Looking through the HECnet node list, it looks like I could stake a claim on a batch of dog-related host names without treading in somebody else's naming conventions. I agree that dynamic DNS is probably not quick enough for my dysfunctional use case. Sometimes my connection flaps up and down rapidly. I'm not sure if I could even accept an incoming connection, or if my cell provider is doing NAT and/or firewalling that would preclude that. I don't understand yet what makes my cellular connection quality so variable. Maybe loading on the local cell towers? Neutrino flux? Demonic corruption? Oddly, when my cell phone reports a "5GE" or "LTE" connection, I expect a poor experience. Everything works best when it's reporting "4G". My connection quality at any given time unpredictably ranges from "doesn't work at all" to "play World of Warcraft for hours with no issues". It's frustrating, but that's the price I pay for living on a nice, big patch of dirt instead of in suburbia. My current physical hardware includes my VAX-11/730, a PDP-11/34A, and a planned future PDP-11/44 project. All of the physical hardware is waiting in varying states of disrepair for me to give it more attention, so I'll use SIMH simulations for semi-regular activity. My VAX-11/730 has formerly run, but it's presently waiting for me to fix some mouse damage. My PDP-11/34A is a work in progress; I've gotten the CPU working, but the big Fujitsu SMD drive it came with has a dead power supply that I haven't debugged yet. I don't know what's on its hard drive. I'll probably run RSX-11M on it, at least part-time? My PDP-11/44 is still a pile of dusty boards, an even dustier pile of chassis to Frankenstein together, and a twinkle in my eye. Has anybody else been on HECnet with dialup-like, or even (gasp!) mobile connections?

------------------------------------------------------------------------ On 3/2/20 12:55 PM, Mark J. Blair wrote: DECnet-over-AX25 appeals to my sense of whimsy, even if it's not remotely practical! If Bob and I have a clear line of sight (extremely unlikely), then a microwave link is also possible. I do have a half-finished project to put in a microwave link between my folks' house and mine, but their connectivity isn't much better. They have DSL and a cellular hotspot; at least one of those links is up frequently, but not always. :) > ------------------------------------------------------------------------ > On Mar 2, 2020, at 9:45 AM, Thomas DeBellis <tommytimesharing at gmail.com> wrote: > > Since you and Bob are both HAM's; what is the distance between your stations? Is IP over radio a ridiculous idea?

On Mar 2, 2020, at 1:10 PM, Dave McGuire <mcguire at neurotica.com> wrote: On 3/2/20 1:06 PM, Thomas DeBellis wrote: I did something very similar about ten years ago. I was stuck in rural West Virginia for several months dealing with a family estate. It was a huge house on a gigantic piece of land, and was awesome in just about every way, except unbelievably isolated. No network access of any kind available there, except Hughes satellite, which is unusable for really much of anything. I put DDWRT on a WRT-54G, paired it with a small Yagi, and scanned around. I ended up pointing it at the lights on a distant mountaintop on the other side of the valley, and, as you said, bingo. We ran with that for about nine months. I still have no idea of whose connectivity that was. -Dave

On Mar 2, 2020, at 9:45 AM, Thomas DeBellis <tommytimesharing at gmail.com>

Since you and Bob are both HAM's; what is the distance between your

On Mar 2, 2020, at 10:26 AM, Robert Armstrong <bob at jfcl.com> wrote: You haven't said how you plan to connect on your end (or if you did, I missed it).

And you'll have only one connection to me for all your machines. Whatever you connect to me will serve as a DECnet router for the rest of your machines, and you'll have to set up some kind of local network at your location to interconnect all of them.

And the idea of a Ham radio AX.25 connection is cute but it's not realistic. A) we're too far apart (I live in Silicon Valley) for anything other than an HF connection and that'd be agonizingly slow, b) can't really leave an HF transceiver running unattended all the time, and c) the legalities of using ham radio for Internet connections are tricky at best (the FCC doesn't allow amateur radio to be used for any commercial purposes). Sorry.

Be aware that multiple simh instances, or pyDECnet, on the same host can't talk to each other on the same physical NI. You'll have to set up a tap device to connect them (or have multiple NIs, but that's usually overkill). Not a problem, really - just an extra step.

circuit tap-0 Ethernet tap:/dev/net/tap0

On Mar 9, 2020, at 12:47 PM, Mark J. Blair <nf6x at nf6x.net> wrote: I made good progress yesterday on updating my home network setup. I

I've started trying to figure out how to run PyDECnet on the same router,

circuit tap-0 Ethernet tap:/dev/net/tap0

2020-03-09 09:43:27.098: DOGRTR: No hardware address for Ethernet TAP-0 2020-03-09 09:43:27.099: DOGRTR: Event type 2.0, Local node state

From node 1.42 (DOGRTR), occurred 09-Mar-2020 09:43:27.099 Reason = Operator command Old state = Off New state = On Does that look right?

circuit tap-0 Ethernet tap:/dev/net/tap0

2020-03-09 09:45:38.533: Thread-1: Exception handling http GET of / Traceback (most recent call last): File

nodeidx, tnode, parts = self.findnode () File

return 0, nodelist[0], parts IndexError: list index out of range and this error in my browser: Error response Error code: 500 Message: Internal Server Error. Error code explanation: 500 - Exception during server

FWIW, the "tap:" option in pyDECnet never worked for me. I use pcap

access tap devices - circuit TAP-0 Ethernet pcap:tap0 --single-address --cost=2 works for me with Ubuntu 18.04...

------------------------------------------------------------------------ On 3/10/20 9:40 AM, Paul Koning wrote: Thanks, sorry for leaving that dangling. I'll test that and merge it into my code. paul > ------------------------------------------------------------------------ > On Mar 10, 2020, at 4:28 AM, Keith Halewood <Keith.Halewood at pitbulluk.org> wrote: > > Hi, > > These are the changes: > > In Ethernet.py: > > class _TapEth (_Ethernet): > def open (self): > fd = os.open('/dev/net/tun', os.O_RDWR | os.O_NONBLOCK) > ifr = struct.pack('16sH', self.dev.encode('utf-8'), IFF_TAP | IFF_NO_PI) > ioctl(fd, TUNSETIFF, ifr) > self.tap = fd > self.sellist = ( fd, ) > # Turn the interface on -- needed only on Mac OS > > With the definitions of those constants up near the top: > > TUNSETIFF = 0x400454ca > TUNSETOWNER = TUNSETIFF + 2 > IFF_TUN = 0x0001 > IFF_TAP = 0x0002 > IFF_NO_PI = 0x1000 > > > > > Keith > > ------------------------------------------------------------------------ > From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of David Moylan > Sent: 10 March 2020 07:46 > To: hecnet at Update.UU.SE > Subject: RE: [HECnet] Intermittent Connection with PyDECnet? > > Robert Armstrong said on Tuesday, 10 March 2020 4:17 AM >> FWIW, the "tap:" option in pyDECnet never worked for me. I use pcap > to >> access tap devices - >> >> circuit TAP-0 Ethernet pcap:tap0 --single-address --cost=2 >> >> works for me with Ubuntu 18.04... > When I first went to play with pyDECnet under Ubuntu 18, I also discovered issues with the tap adapter. Fundamentally the issue comes down to newer kernels (such as in Ubuntu 18) that use a /dev/net/tun interface to communicate instead of /dev/tapx > > I heard on the list that Keith Halewood had modified Paul's code to work with the newer interface standard, so I reached out to him and he provided me with the changes. It's only a few lines of code that need modification. > > I believe Keith has provided Paul with the changes. I'd love to see them become mainstream in pyDECnet, as at the moment if I pull down a new build, I have to make the modifications by hand again. > > With the code changes, tap works a dream for me. > > circuit tap-1022 Ethernet tap:tap1022 --random-address > > cheers, Wiz!! >

I chose the code change I made simply because I don?t use pcap. I also don?t like layers for layers sake. *From:*owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] *On Behalf Of *David Moylan *Sent:* 10 March 2020 21:19 *To:* hecnet at Update.UU.SE *Subject:* RE: [HECnet] Intermittent Connection with PyDECnet? I?m all for backwards compatibility, however the /dev/net/tun interface has been in linux since the 2.6 days. A quick google shows that people back in 2010 have been using it, and browsing on kernel.org for the oldest archived kernel entry of 2.6.11 shows code examples for /dev/net/tun dated 2005. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/Docum… I think the starting date for support of /dev/net/tun is earlier than that of the discontinuation of support for /dev/tapx. Given that the whole goal of pyDECnet is to permit usage of decnet on more modern environments, I think having code that supports back to 2005 is a pretty solid move. Cheers, Wiz!! *From:*owner-hecnet at Update.UU.SE <mailto:owner-hecnet at Update.UU.SE> [mailto:owner-hecnet at Update.UU.SE] *On Behalf Of *Thomas DeBellis *Sent:* Wednesday, 11 March 2020 2:08 AM *To:* hecnet at Update.UU.SE <mailto:hecnet at Update.UU.SE> *Subject:* Re: [HECnet] Intermittent Connection with PyDECnet? Since you may be running both on newer and older kernels, does it make sense to test for which tunnel type interface exists and simply use that? You don't even need to test versions, but rather try /dev/net/tun and if it works, use that or if it fails, use /dev/tapx. I'm a big fan of dynamic configuration, also known in the trade as 'auto magic' or 'automatic nice things'. ------------------------------------------------------------------------ On 3/10/20 9:40 AM, Paul Koning wrote: Thanks, sorry for leaving that dangling.? I'll test that and merge it into my code. paul ------------------------------------------------------------------------ On Mar 10, 2020, at 4:28 AM, Keith Halewood <Keith.Halewood at pitbulluk.org> <mailto:Keith.Halewood at pitbulluk.org> wrote: Hi, These are the changes: In? Ethernet.py: ?? class _TapEth (_Ethernet): ?????? def open (self): ?????????? fd = os.open('/dev/net/tun', os.O_RDWR | os.O_NONBLOCK) ?????????? ifr = struct.pack('16sH', self.dev.encode('utf-8'), IFF_TAP | IFF_NO_PI) ?????????? ioctl(fd, TUNSETIFF, ifr) ?????????? self.tap = fd ?????????? self.sellist = ( fd, ) ?????????? # Turn the interface on -- needed only on Mac OS With the definitions of those constants up near the top: TUNSETIFF = 0x400454ca TUNSETOWNER = TUNSETIFF + 2 IFF_TUN = 0x0001 IFF_TAP = 0x0002 IFF_NO_PI = 0x1000 Keith ------------------------------------------------------------------------ From: owner-hecnet at Update.UU.SE <mailto:owner-hecnet at Update.UU.SE> [mailto:owner-hecnet at Update.UU.SE] On Behalf Of David Moylan Sent: 10 March 2020 07:46 To: hecnet at Update.UU.SE <mailto:hecnet at Update.UU.SE> Subject: RE: [HECnet] Intermittent Connection with PyDECnet? Robert Armstrong said on Tuesday, 10 March 2020 4:17 AM FWIW, the "tap:" option in pyDECnet never worked for me.? I use pcap to access tap devices - ??? circuit TAP-0 Ethernet pcap:tap0 --single-address --cost=2 works for me with Ubuntu 18.04... When I first went to play with pyDECnet under Ubuntu 18, I also discovered issues with the tap adapter. Fundamentally the issue comes down to newer kernels (such as in Ubuntu 18) that use a /dev/net/tun interface to communicate instead of /dev/tapx I heard on the list that Keith Halewood had modified Paul's code to work with the newer interface standard, so I reached out to him and he provided me with the changes. It's only a few lines of code that need modification. I believe Keith has provided Paul with the changes. I'd love to see them become mainstream in pyDECnet, as at the moment if I pull down a new build, I have to make the modifications by hand again. With the code changes, tap works a dream for me. circuit tap-1022 Ethernet tap:tap1022 --random-address cheers, Wiz!!

On Mar 10, 2020, at 5:18 PM, David Moylan <djm at wiz.net.au> wrote: I?m all for backwards compatibility, however the /dev/net/tun interface has been in linux since the 2.6 days. A quick google shows that people back in 2010 have been using it, and browsing on kernel.org for the oldest archived kernel entry of 2.6.11 shows code examples for /dev/net/tun dated 2005. https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/Docum… I think the starting date for support of /dev/net/tun is earlier than that of the discontinuation of support for /dev/tapx. Given that the whole goal of pyDECnet is to permit usage of decnet on more modern environments, I think having code that supports back to 2005 is a pretty solid move.

On Mar 10, 2020, at 7:54 PM, Mark J. Blair <nf6x at nf6x.net> wrote: Last night, I incorporated the suggestions from Paul and Robert, and got pydecnet to launch happily and display the web browser status interface. However, it doesn't appear to be seeing packets from either the physical LAN or the other tap interface which I have a simh instance connected to, using either the tap: or pcap: modes. I have not tried the tap: code changes discussed here yet, and I might just wait for Paul to incorporate those in the main release. But I'm surprised that using pcap: mode as Robert suggested didn't work for me. I may have configured something incorrectly. I'll try some more later.

On Mar 10, 2020, at 5:12 PM, Robert Armstrong <bob at jfcl.com> wrote: Dumb question - you did ifconfig the tap device as "up", right?

Dumb question #2 - you are running pyDECnet as root, right? And note that, for me at least, sudo doesn't work with pyDECnet; somehow it forgets that it's root. I have to do "sudo -i" and the type the pyDECnet command line.

On Mar 10, 2020, at 8:50 PM, Supratim Sanyal <supratim at riseup.net> wrote: KLH-10 TOPS-10 noob questions: 2) At the monitor dot prompt, I can login as 1,2 and R OPR and then ENTER NCP etc. What exactly does the SET EXECUTOR NODE xx.yyy do?

KLH-10 TOPS-10 noob questions: 1) At the TOPS-10 boot startup option prompt, I can type in CHANGE and then set the DECnet address. How do I make it persist across reboots and not have to do this every time?

2) At the monitor dot prompt, I can login as 1,2 and R OPR and then ENTER NCP etc. What exactly does the SET EXECUTOR NODE xx.yyy do?

I think you're right, but it has been _decades_ since I last used Tops-10.? At WPI, we had a KA-10 running a much modified 6.03 series monitor that we were quite proud of.? At Marlboro, the project that I was working on (FILE-FINDER, a database for DUMPER tapes) was quite Tops-20 centric; we depended on files with holes in them. I'm unaware of any systems level structured data store in either Tops-10 or Tops-20 with the exception of the Quasar failsoft file (QSRFSS, holds queue, print, batch requests across crashes).? I don't find this surprising; if you crash and corrupt a file with confuration information in it, a flat ASCII file is whaaay easier to recover than an specially engineered database.? The binary accounting and error files are sequential and don't count, IMHO. Under Tops-20, we used the following 'trick' for start-up speed and persisted configuration.? The configuration file was 'compiled' into binary and directly mapped into memory on start-up. 1. This was necessary for LPTSPL as it is started up for jobs, but shut down and put into a quiescent state when there is nothing left to print.? When you have a lot of printers, reparsing LPFORM.INI can be a real dog.? Very noticeable. 2. I got the idea from the mailer, which does the same thing for mailing-list.txt 3. The EXEC will also do it; you can restore a binary environment with all your special scripts really fast (like on PUSH or LOGIN) 4. I had been thinking about doing this for the Extended Mode FTP server, but I'm not sure it's worth it.? I instrumented the start up time and it's in the milliseconds.? Probably would be necessary for a couple hundred simultaneous small requests. If I ever get truly serious about supporting Galaxy again, then probably I'll bite the bullet and put up Tops-10 so I can validate execution. > ------------------------------------------------------------------------ > On 3/10/20 9:25 PM, Johnny Billquist wrote: > > I've never used Tops-10 as an operator, so I can't answer most of > this, but one question I think I can... > > My understanding is that neither Tops-10, nor TOPS-20 have a > persistent database. Instead you need to have a script that does all > the definitions, and you need to run it at every boot. But I could be > confused about that one. >> ------------------------------------------------------------------------ >> On 2020-03-11 01:50, Supratim Sanyal wrote: >> >> KLH-10 TOPS-10 noob questions: >> >> 1) At the TOPS-10 boot startup option prompt, I can type in CHANGE >> and then set the DECnet address. How do I make it persist across >> reboots and not have to do this every time?

I don't think it would be a terrible idea to understand just a little bit more to know what start up files to edit, but I certainly do understand not wanting to be bothered with something.? If I remember any more, I'll let you know.

The new SIMH port appears to allow for a slaved PDP-6, which I remember seeing on the 9th floor (It was connected to the MIT AI KA-10).? I think they only shared one moby.? The last I heard was that the KL sources to ITS (MC) were lost.? Anyway, if the SIMH KL simulator allows multiple CPU's, then you could run Tops-10 SMP, which really was a tour de force.? Extremely cool.

> ------------------------------------------------------------------------ > > On 3/10/20 10:41 PM, Supratim Sanyal wrote: > > 31.37 (TWONKY) is just a straight TWONKY distribution on KLH-10. All > required keyboard interactions to get it to boot up are consistently > the same; so I might be able to wrap it up around an expect script > ... worth a shot. > >> ------------------------------------------------------------------------ >> >> On 3/10/20 9:46 PM, Thomas DeBellis wrote: >> >> I think you're right, but it has been _decades_ since I last used >> Tops-10.? At WPI, we had a KA-10 running a much modified 6.03 series >> monitor that we were quite proud of. At Marlboro, the project that I >> was working on (FILE-FINDER, a database for DUMPER tapes) was quite >> Tops-20 centric; we depended on files with holes in them. >> >> I'm unaware of any systems level structured data store in either >> Tops-10 or Tops-20 with the exception of the Quasar failsoft file >> (QSRFSS, holds queue, print, batch requests across crashes).? I >> don't find this surprising; if you crash and corrupt a file with >> confuration information in it, a flat ASCII file is whaaay easier to >> recover than an specially engineered database.? The binary >> accounting and error files are sequential and don't count, IMHO. >> >> Under Tops-20, we used the following 'trick' for start-up speed and >> persisted configuration.? The configuration file was 'compiled' into >> binary and directly mapped into memory on start-up. >> >> 1. This was necessary for LPTSPL as it is started up for jobs, but >> shut down and put into a quiescent state when there is nothing >> left to print.? When you have a lot of printers, reparsing >> LPFORM.INI can be a real dog.? Very noticeable. >> 2. I got the idea from the mailer, which does the same thing for >> mailing-list.txt >> 3. The EXEC will also do it; you can restore a binary environment >> with all your special scripts really fast (like on PUSH or LOGIN) >> 4. I had been thinking about doing this for the Extended Mode FTP >> server, but I'm not sure it's worth it.? I instrumented the >> start up time and it's in the milliseconds.? Probably would be >> necessary for a couple hundred simultaneous small requests. >> >> If I ever get truly serious about supporting Galaxy again, then >> probably I'll bite the bullet and put up Tops-10 so I can validate >> execution. >> >>> ------------------------------------------------------------------------ >>> On 3/10/20 9:25 PM, Johnny Billquist wrote: >>> >>> I've never used Tops-10 as an operator, so I can't answer most of >>> this, but one question I think I can... >>> >>> My understanding is that neither Tops-10, nor TOPS-20 have a >>> persistent database. Instead you need to have a script that does >>> all the definitions, and you need to run it at every boot. But I >>> could be confused about that one. >>>> ------------------------------------------------------------------------ >>>> On 2020-03-11 01:50, Supratim Sanyal wrote: >>>> >>>> KLH-10 TOPS-10 noob questions: >>>> >>>> 1) At the TOPS-10 boot startup option prompt, I can type in CHANGE >>>> and then set the DECnet address. How do I make it persist across >>>> reboots and not have to do this every time? > -- > Supratim Sanyal, W1XMT > 39.19151 N, 77.23432 W > QCOCAL::SANYAL via HECnet

------------------------------------------------------------------------ On 3/11/20 4:05 AM, G. wrote: .R OPR OPR> SET KSYS NOW Wait for '-- KSYS processing completed --' then give the 'shut' command at the 'KLH10>>' emulator prompt (which will return you to the 'BOOT>' prompt) and finally 'halt' and 'quit' the emulator. > ------------------------------------------------------------------------ > On Tue, 10 Mar 2020 20:50:15 -0400, Supratim Sanyal wrote: > > 3) What is the proper way to shut a KLH-10 PDP-10 running TOPS-10 system > down?

Yes please. It hogs 100% of Xeon core. I did add a "devdef idler 700 host" to the .ini - it does not help, so again I need a smarter 704dcn.

On 3/11/2020 09:38, Thomas DeBellis wrote: .CONO 700,1 ?CONO? -- Supratim Sanyal, W1XMT 39.19151 N, 77.23432 W QCOCAL::SANYAL via HECnet

On Wed, 11 Mar 2020 09:29:02 -0400, Supratim Sanyal wrote: Here we go, for you and everyone else :) 1. First of all make sure you have a complete set of monitor sources in [10,7,MON]. If you followed either the Software Installation Guide or some online tutorial you should be okay. In particular make sure that you have the following: [10,7,MON]S.MAC [10,7,MON]DEVPRM.MAC [10,7,MON]DTEPRM.MAC [10,7,MON]NETPRM.MAC [10,7,MON]D36PAR.MAC [10,7,MON]COMMON.MAC [10,7,MON]COMDEV.MAC [10,7,MON]COMMOD.MAC [10,7,MON]CLOCK1.MAC 2. Also make sure you have REL:TOPS10.REL and [10,7,MON]MONGEN.EXE 3. Make a backup copy of the original TOPS10.REL. I like to preserve the timestamp of original files, therefore I do a rename followed by a copy of the renamed file (I have adopted .DEC as the extension for original files) .RENAME REL:TOPS10.DEC=REL:TOPS10.REL .COPY REL:TOPS10.REL=REL:TOPS10.DEC (destination=source) 4. Change your default to [10,7,MON] .R SETSRC *CP [10,7,MON] *^Z 5. Make a backup copy of CLOCK1.MAC as you did for TOPS10.REL 6. Open CLOCK1.MAC with any editor of your choice (SOS, EDIT, etc.) and modify it as follows: Before: SOJG 6,1 ;AC1 - DON'T CAUSE EXCESSIVE MEMORY INTERFERENCE After: CONO 740,1 ;AC1 - DON'T CAUSE EXCESSIVE MEMORY INTERFERENCE (I have removed some white space to fit either line in 78 characters) The address in the above CONO instruction must match the address of the idler in the emulator configuration. Several online tutorials suggest 700 but for some reason 740 should be a safer choice (I forgot the specifics, nonetheless some post in alt.sys.pdp10 may have some explanation) 7. Compile CLOCK1.MAC (it will create CLOCK1.REL) .COMPIL/COMPIL CLOCK1 8. Replace the CLOCK1 module from CLOCK1.REL in TOPS10.REL .R MAKLIB *REL:TOPS10.REL=REL:TOPS10.REL/MASTER:CLOCK1,DSK:CLOCK1.REL/REPLACE:CLOCK1 *REL:TOPS10.REL=REL:TOPS10.REL/INDEX *^Z (Indexing is optional but will speed up relinks) 9. Run the monitor configuration utility. It will ask lots of questions, note that most defaults will be okay. I have attached to this message a sample record of a generation of mine. Obviously some parameters such as the system name and address will have to be customized. Values in the decimal and octal "symbol,value" sections are specific for my installation and could be safely skipped. If you answer YES to the first question the utility will generate a MONGEN.MIC file with a record of your input for later reference: .R MONGEN Write MONGEN.MIC (YES,NO) : YES [...] The latest Software installation Guide contains a guide of the whole procedure (chapter 9). You can find it at the following address: http://pdp-10.trailing-edge.com/tops10v704_docc/01/10,7/docupd/sig.mem.html 10. With any editor of your choice create a MONBLD.MIC procedure file containing the following lines: .DAYTIME .COMPILE/COMPILE F,S .COMPILE/COMPILE DEVPRM,DTEPRM,NETPRM,D36PAR .COMPILE/COMPILE SYSCNF+<COMMON,COMDEV,COMMOD> .R LINK *SYSTEM/SAVE/NOINITIAL/HASH:13K = /LOCALS - *COMMON,COMDEV,COMMOD,REL:TOPS10/SEARCH- */PATCH:200/COUNTERS/GO .DIRECT/DETAIL SYSTEM.EXE .DAYTIME Initial dots and asterisks are required. It's important that a space be present between /LOCALS and the following hyphen or it will not work. 11. Execute the relink procedure you created in the previous step: .DO MONBLD It will take from a few seconds to a few minutes and will create a new SYSTEM.EXE monitor file in the current directory. 12. Copy the new monitor to the system directory giving it some unique name: .COPY SYS:TRYMON.EXE=DSK:SYSTEM.EXE 13. Reboot your system and type the name of your new monitor at the BOOT> prompt. Unless you have some strange configuration it should be enough for it to find it, otherwise type its full path: BOOT>TRYMON 14. If everything works as intended you could either copy or rename your new monitor as SYS:SYSTEM.EXE so that it would be loaded by default during boot by just pressing the return key. Whenever you will need a new monitor it will be enough to just change your default to [10,7,MON] and then run MONGEN and MONBLD as above. If you want to avoid running the whole MONGEN every time you could also directly edit either SYSCNF.MAC and/or F.MAC and then run MONBLD. Let me know if it works or if I have forgot something, :) G.

For links to the manuals on bitsavers, and other things, you can go here. I think that most of the PDP-10 links are still valid, I really need to get back to verifying the links. http://www.avanthar.com/healyzh/decemulation/pdp10emu.html Zane

issues with the tap adapter.

On Mar 9, 2020, at 1:07 PM, Paul Koning <paulkoning at comcast.net> wrote: ... The bad error handling is that PyDECnet should complain if you didn't define at least one node or bridge.

On Mar 9, 2020, at 2:34 PM, Mark J. Blair <nf6x at nf6x.net> wrote: Hmm. I also got that error when I set the correct hardware address on the tap0 interface before launching pydecnet.

On Mar 9, 2020, at 2:34 PM, Mark J. Blair <nf6x at nf6x.net> wrote: Dumb question time, because I struggle with VMS: Is there a DECnet equivalent to "ping" that I can use to verify that my emulations can talk to pydecnet? I gather that's what the mirror server is for, but I don't know how to invoke it on the VMS side.

On Mar 2, 2020, at 12:27 PM, Mark J. Blair <nf6x at nf6x.net> wrote: This sounds promising! I am interested in at least running some experiments to find out how well it all works, and to learn enough to configure my future little patch of DECnet. I'd like to have fun participating, but not if it causes problems for my upstream node or HECnet in general. I know that my connectivity situation is unusual. I am in southern California as my callsign suggests, just outside Riverside. Is there any possibility of connecting through something like an SSH tunnel, and relying on that for a higher level of security and authentication for the link than the DECnet protocols natively provide?

If I'd like to eventually connect several nodes on my local network into HECnet, would I need to consume a whole area number? Or is there a means for further routing within an existing HECnet area? I still have much to learn about DECnet. Looking through the HECnet node list, it looks like I could stake a claim on a batch of dog-related host names without treading in somebody else's naming conventions.

I agree that dynamic DNS is probably not quick enough for my dysfunctional use case. Sometimes my connection flaps up and down rapidly. I'm not sure if I could even accept an incoming connection, or if my cell provider is doing NAT and/or firewalling that would preclude that.

On Mar 2, 2020, at 1:05 PM, Mark J. Blair <nf6x at nf6x.net> wrote: I would think that SSL would be fine for the link security. Would that also provide a mechanism for the caller to authenticate themself to the upstream link?

Are we actually talking about encrypting all the DECnet traffic, or just authentication? I'm a little worried about the amount of computational overhead involved in encrypting all the DECnet traffic, although I suppose that given the trivial volume of HECnet traffic it's not a major concern. Bob -----Original Message----- From: Hecnet-list [mailto:hecnet-list-bounces+bob=jfcl.com at lists.sonic.net] On Behalf Of Paul Koning Sent: Monday, March 2, 2020 10:27 AM To: hecnet at update.uu.se Subject: Re: [HECnet] Intermittent Connection with PyDECnet? SSL library that exists in Python. SSH tunnel not quite so much. Would SSL be sufficient? also provide a mechanism for the caller to authenticate themself to the upstream link? The Python SSL library has a pile of support for certificates (in both directions, as far as I can tell, so mutual authentication is possible). Also ways to query the certificates used. This is stuff I haven't used before so it will take some study to understand it. paul _______________________________________________ Hecnet-list mailing list Hecnet-list at lists.sonic.net https://lists.sonic.net/mailman/listinfo/hecnet-list

That said, I wouldn't really worry too much. First of all, few people would understand DECnet unless they explicitly look for it. Second, this is all just hobbyist fun. It's not as if something actually depends on this running. Third, anyone (more or less) is free to connect if they want to anyway, so it's not a secret cabal with entrance rites. In the end, if someone really is worried, they should not connect to HECnet to start with. DECnet itself was never designed for the security levels or issues people know about today. If security is keeping you awake at night, then you should not run DECnet at all. Don't fool yourself into thinking that because you are running some link over SSH or whatever, you are now safe.

I don't see how you can have multiple Multinet TCP connections all listening on port 700 at the same time. Yes, you can have multiple outgoing connections all to port 700 on some remote machine, but that's not the same. And for UDP it's no problem for everybody to share the same port, but again that's not the same.

On Mar 2, 2020, at 1:56 PM, Robert Armstrong <bob at jfcl.com> wrote: I don't see how you can have multiple Multinet TCP connections all listening on port 700 at the same time. Yes, you can have multiple outgoing connections all to port 700 on some remote machine, but that's not the same. And for UDP it's no problem for everybody to share the same port, but again that's not the same. Bob

------------------------------------------------------------------------ On 3/2/20 2:42 PM, Johnny Billquist wrote: Yeah. And this is where I'm not at all happy with RSX. Because the node verification in RSX is set on the whole system, and not per circuit. ? Johnny

? Don?t know about TOPS-20, but in NCP/NICE/NML it?s ?????????????? SET CIRCUIT ? VERIFICATION ENABLED and ?????????????? SET NODE ? TRANSMIT PASSWORD ? The password the remote node sends has to match the one that?s associated with that node in the local DECnet database.

I don't see where I would set an id and password in 7-1-CONFIG or in NCP.

Yes, this is /exactly/ what I was talking about below, "I sometimes get the impression that it parses for more than it can actually do."? You can put any kind of circuit id you want, but you get the exact same pick list, even when the parameter makes no sense.? NMLT20 does the semantic processing and decides whether the request passes muster.? So it looks like I don't have it, but I'll have a look at the sources maybe later. It's not a limitation of Tops-20 but rather that the OPR user mode parser does not have the appropriate semantic knowledge. You can get some really fancy parsing with COMND%.? I guess my favorite was a program I wrote that built its parameter list on the fly based on what you were able to set and what you had already set.? So, you could type a "?" and be assured that what was listed was in fact settable.? As you set each parameter, it got taken off the parse list, so you could get real time feed back on what you had left to do. Very nice and a bit overboard, but that's me, I guess.? I have never seen COMND%'s equal anywhere on Unix or Windows or anywhere else except for C-MM and C-KERMIT, both of which used Columbia C based COMND% package. > ------------------------------------------------------------------------ > > ?On 3/2/20 7:36 PM, Robert Armstrong wrote: > > FWIW, you can?t/couldn?t do verification on an Ethernet circuit > anyway.? It only works on point to point links. ?Command completion > may be smart enough to only show you the options that are applicable > to your input so far, or maybe TOPS20 just doesn?t have it? > > Bob > > *From:*owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] > *On Behalf Of *Thomas DeBellis > *Sent:* Monday, March 2, 2020 3:55 PM > *To:* hecnet at Update.UU.SE > *Subject:* Re: [HECnet] Intermittent Connection with PyDECnet? > > I appear to have neither CIRCUIT VERIFICATIONnor NODE TRANSMIT > PASSWORD, viz: > > NCP>set cirCUIT ni-0 ? ALL > ? or one of the following: > ?ACTIVE??? ???? BABBLE??? ? BLOCKING???? CHANNEL COST > ?COUNTER????? DEAD > ?DTE??? ??? ? ? DYING??? ?? HELLO??? ??? INACTIVE LINE > ?LISTEN??? ?? MAXIMUM > ?NUMBER??? ???? OWNER??? ?? POLLING????? RECALL ROUTER > SERVICE????? STATE > ?TRANSMIT?????? TRIBUTARY?? TYPE??? ???? USAGE > > NCP>set noDE venti2:: ? ALL > ? or one of the following: > ?ADDRESS?? CIRCUIT > ? or one of the following: > ?AREA??? ??? ? BROADCAST??? ??? BUFFER??? ????? CONSOLE ?? COUNTER > ?CPU??? ??? ?? DELAY??? ??? ??? DIAGNOSTIC??? ? DUMP ????? HARDWARE > ?HOST??? ??? ? IDENTIFICATION?? INACTIVITY INCOMING??? ? LOAD > ?MAXIMUM??? ?? OUTGOING??? ???? RETRANSMIT??? ? ROUTING ?? SECONDARY > ?SEGMENT??? ?? SERVICE??? ????? SOFTWARE??????? STATE > SUBADDRESSES > ?TERTIARY????? TYPE > > I may be mis-remembering how our DN200 was configured.? I believe the > Top-20 NCP parser may have shared keyword tables with other DECnet > implementations via BLISS at one point. Anyway, I recall at least some > of it was BLISS.? I haven't look in that particular box lately, but > the parse tables appear to assembler now, with heavy uses of GLXMAC > macros. > > The Tops-20 operator interface has two 'sub-modes', as it were: NCP > and LCP.? NCP parses DECnet related keywords.? LCP is used for LAT > configuration.? The operator subsystem is kind of neat; OPR parsers > send packets to Orion, a communications server and router.? It notes > an NCP subset and routes that to NMLT20 which contains just what you > think it would. > > Early versions of NMLT20 (about 1980) appeared insufficiently > productized.? It could hang in a loop or (more frequently) crash for > no readily apparent reason.? We submitted a rather large number of > SPR's...? However, it's quite stable now; I don't believe I've had a > single crash.? I have yet to come to a fully informed conclusion about > the lexical interface; I sometimes get the impression that it parses > for more than it can actually do. > > Elsewhere in Tops-20, DECnet implementation is a mixed bag; most of > the monitor code appears finished; I go months without reboots and > that is with active development.? The user mode code is another matter > entirely.? I had my hands full fixing a number of issues with DAP and > FAL.? It was /invaluable/ having heterogeneous HECnet nodes to test > against, RSX 11M+ and VMS.? SETNOD needs some finishing and perhaps > some additional functionality.? I have yet to look at FTS; I don't see > how it can get or generate any traffic at all. The monitor interface > (NFT:) does not appear to be operational.? Elsewhere (perhaps Galaxy) > appears to have no submission interface.? Johnny and I have some minor > tweaks to flush out with MAIL11 (in addition to the changes that I've > already put in). > > On 3/2/20 3:25 PM, Robert Armstrong wrote: > > >I don't see where I would set an id and password in 7-1-CONFIG or > in NCP. > > Don?t know about TOPS-20, but in NCP/NICE/NML it?s > > SET CIRCUIT ? VERIFICATION ENABLED > > and > > SET NODE ? TRANSMIT PASSWORD ? > > The password the remote node sends has to match the one that?s > associated with that node in the local DECnet database. > > Bob >

Did you see any SET EXEC TRANSMIT PASSWORD maybe? Anyway, the NCP parsing in RSX is all written in MACRO-11, so it was never shared with anything written in BLISS. ? Johnny On 2020-03-03 00:55, Thomas DeBellis wrote:

On 2020-03-03 02:16, Thomas DeBellis wrote: RSX don't even authenticate on a node basis. As I said, it's not "SET EXECUTOR NODE ..." it is "SET EXECUTOR TRANSMIT PASSWORD ...". In RSX, it's global. Enable it, and you have it everywhere, and the same password everywhere. No way of any finer granularity. Rather unsatisfying... Oh, I know DEC did move a lot to BLISS. But DECnet-11M wasn't among them, ever. There are only a couple of things in RSX written in BLISS. On the DECnet side, that is PHONE, and then some Professional specific modules that were tailored for P/OS. In RSX itself, some file system related stuff is in BLISS, but not much else. ? Johnny

On Mar 2, 2020, at 2:43 PM, Robert Armstrong <bob at jfcl.com> wrote: connection on that ... So it's correct that in the current implementation a site would need a unique port for every Multinet TCP listen connection? Just want to make sure I'm not imposing needless restrictions...

Hmm.... On Multinet on VMS that just works. Don't know how or why (never really thought about it), but it does.

On Mar 2, 2020, at 6:35 AM, Johnny Billquist <bqt at softjar.se> wrote: Hi. On 2020-03-02 06:34, Mark J. Blair wrote: Yes. It is possible. If you use TCP as the transport, it can work with you having only intermittent connectivity, and different addresses. Paul Koning will have to tell if PyDECnet can also act as a listener from some arbitrary remote system. But with RSX I can definitely do that. So if nothing else, you can always set it up so that you connect your PyDECnet to an RSX host...