HECnet November 2021

hecnet@lists.dfupdate.se

42 participants
48 discussions

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 --> Cryptic --> Rsx-11S

by R.Voorhorst＠swabhawat.com

Hi Mark, The Dn20 is Rsx-11S with Decnet layered on it in phase-III/IV forms; however the Decnet layer is only offered on pdp10 software and mainly without source code. It will not run (yet). Maybe the Cex layer could be tinkered onto an Rsx-11M 3.1 base system that has (some) source code available. Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Mark Matlock Sent: Friday, 19 November, 2021 19:06 To: hecnet at Update.UU.SE Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 --> Cryptic All, This discussion has been quite interesting! Back in the 1982, I did programming for a large RSX11M/DECnet based system that input commodity trading data from around eight different commodity exchanges (CBOT, COMEX, Mercantile, etc.) as well as commodity news service and distributed it to over 200 VT100s located all over the U.S. The host system was an 11/44 connected to the development 11/44 via DMCs and they shared a DT07 bus with box. In the DT07 was a couple DZ11s for data input from the exchanges and a DV11 with 16 synchronous serial ports that were connected to twelve remotely located 11/23s running RSX11S & DECnet. The 11/23s were downloaded RSX11S via their DUV11s and the re-entrant RSX tasks received DECnet task to task data packets about trades and then displayed individualized data to ~20 VT100s on DZV11s and DLV11-Js. A DUP11 on one of the 11/44s also did RJE jobs to an IBM mainframe. A watch-dog timer ran between the 11/44s so that a semi-hot fail over could occur and the DT07 would move the I/O to the development system if the primary had hardware issues. Later I had an 11/24 that connected into this network via the CPUs 2nd DL11 port that was often used for a TU58. I don?t remember exactly what it connected to on the 11/44 side. All the network connections were 9600 baud point to point lease lines from AT&T. If anyone out there has a DECnet distribution for RSX11S, I?d love to get a copy to experiment with today. Best, Mark > On Nov 19, 2021, at 10:50 AM, Johnny Billquist <bqt at softjar.se> wrote: > > Well, I meant this in the context of a DUP11, but sure, it's also true with DMC. > > But basically, in RSX, DECnet is just a component of something called the communications executive, or CEX. CEX provides its own environment in which multiple processes run, and provide services. > > CEX then provides an interface to its drivers from RSX, which means that any RSX program can make use of the services and protocols that CEX have, without have DECnet in the picture. > > So for example, CEX provides communication lines, which are packet based. If that is over asynch lines it utlizes DDCMP over async serial. Any application in RSX can make use of this. If you instead have a synch line, like a DUP11, or DMC, that is another interface inside CEX. From the outside it makes no difference if it is a DUP or a DMC. Bottom side is synch serial, using DDCMP. With DMC this is all done in the hardware. With a DUP11, CEX is doing the DDCMP layer in software. > > For the user program this is all invisible. You just communicate over the DUP11, and it will be doing DDCMP. > > If you actually want to use the DUP11 *whithout* DDCMP, then you'd go for the DUP11 device driver instead of accessing it through CEX. > > And talking to a device using CEX in RSX means you talk to the "direct line access" interface, or DLX. That is just one interface, independent of what underlying hardware you are going to access. In the DLX, you start by opening the link, and for that you use the name, which is the same as the line name you see for DECnet. But you are in fact then just talking at the line level, using whatever protocol layer is used for that line. DDCMP, for example. > > Anything using DDCMP is always in packets. Don't matter how the hardware looks. > > DLX is also used to access ethernet, by the way. And that is how I coexist with DECnet in RSX, since ethernet can allow multiple channels on the same hardware. > > I hope this makes sense. Else I can try to explain it more. > > Johnny

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 --> Cryptic -->Dup11 simh only for Ddcmp

by R.Voorhorst＠swabhawat.com

Yes, that is how Pdp10 KS10 and Vax and Rsx and Rsts all interface to the Dmc/Dmr offered layer and for Anf10 on Pdp10 as well. But trying Decnet on bare Dup11 will reveal all. However this will fail on simh: look at the comment in Pdp11_dup.c from Bob: dup DUP11 Unibus/DPV11 Qbus bit synchronous interface This module implements a bit synchronous interface to support DDCMP. Other synchronous protocols which may have been supported on the DUP11/DPV11 bit synchronous interface are explicitly not supported. So the 0,01 % rises to 100% that it will not work as bare sync device -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Paul Koning Sent: Friday, 19 November, 2021 17:23 To: hecnet at update.uu.se Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 --> Cryptic I assume it means that RSX (and RSTS too) lets you use a DMC directly from an application as an I/O device you can open. If so, you get a packet service that lets you transmit whatever packets you want, which will be sent encapsulated in DDCMP protocol. In the RSX case I'd assume that includes support for maintenance mode (in RSTS that isn't supported). paul > On Nov 19, 2021, at 11:14 AM, R. Voorhorst <R.Voorhorst at swabhawat.com> wrote: > > Please can you explain your second sentence somewhat? I read: "... the Ddcmp layer ... it does present it as a device ...". I do not perceive the connotation. > > > Reindert > > -----Original Message----- > From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Johnny Billquist > Sent: Friday, 19 November, 2021 14:46 > To: hecnet at Update.UU.SE > Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 > > RSX definitely can use a DUP11 standalone through DECnet. In which case it's software side DDCMP. > > Heck, you can also use the DDCMP layer yourself, without involving DECnet. It does present it as a device to the system on which you can send/receive packets. > > Johnny

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re3 --> DNxx

by R.Voorhorst＠swabhawat.com

Yes, that is what I am busy with: get that system running in this case for Decnet-III; it that runs DN60 will also run on Rsx-11S but then the interfacing to the Ibm world is the next thing to tackle. Not that I will devote too much time to the DN60, but DN20 I probably will. Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Robert Armstrong Sent: Friday, 19 November, 2021 17:33 To: hecnet at Update.UU.SE Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re3 > <dave.g4ugm at gmail.com> <dave.g4ugm at gmail.com> wrote: > It might be worth looking at how Hercules encapsulates SDLC as that > will do TOPS-10 had an IBM HASP communication FE product; I think it was called a DN60. I used it long ago - you could submit card decks to JES2 under OS/VS2 from files on the -10 and get the resulting print out back as another disk file on the -10. Beats a keypunch and a line printer by miles... It'd be kind of neat, although a total waste of time, to hook up a simh TOPS10 to a Hercules and try to get that working. Bob

3 years

VMS progtram for real tape -> TAP

by bob＠jfcl.com

Can someone point me to a program for VMS that will read a real, physical, 9-track tape into a simh compatible TAP or TPC file? I was sure I already had one, but I can't seem to find it anymore. Thanks, Bob

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 --> Cryptic

by R.Voorhorst＠swabhawat.com

Please can you explain your second sentence somewhat? I read: "... the Ddcmp layer ... it does present it as a device ...". I do not perceive the connotation. Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Johnny Billquist Sent: Friday, 19 November, 2021 14:46 To: hecnet at Update.UU.SE Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4 RSX definitely can use a DUP11 standalone through DECnet. In which case it's software side DDCMP. Heck, you can also use the DDCMP layer yourself, without involving DECnet. It does present it as a device to the system on which you can send/receive packets. Johnny On 2021-11-19 12:22, R. Voorhorst wrote: > Well, that leaves 0,01% chance to see that assumption defeated so why not test it? > I'll see next week if that can be tested if a software testbed can be found to support the single Dup11 that wise: VMS at least does not in the Decnet corner but Vax Psi may be different. Maybe Rsx-11M+ would be a suitable testbed? > > Reindert > > -----Original Message----- > From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On > Behalf Of Mark Pizzolato - Info Comm > Sent: Friday, 19 November, 2021 05:19 > To: hecnet at Update.UU.SE > Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should > be Ddcmp --Re3 > > I have not followed the details of this conversation, BUT I'm 99.99% sure that a PDP11 can be configured with just a DUP with DECnet using that interface, and that DUP can be "wired" to another simulator which has a DUP, KDP/DUP or a DMC and DECnet can be spoken just fine. > > - Mark > > On Thursday, November 18, 2021 at 3:35 PM, Reindert wrote: >> >> No, the Dup is only simulated as Kdp/Dup combo, hence the Ddcmp >> characteristics of the pair. >> >> >> Reindert >> >> -----Original Message----- >> From: owner-hecnet at Update.UU.SE [mailto:owner- hecnet at Update.UU.SE] >> On Behalf Of Johnny Billquist >> Sent: Friday, 19 November, 2021 00:31 >> To: hecnet at Update.UU.SE >> Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> >> should be Ddcmp --Re2 >> >> Fair enough. >> >> But DUP11 is a sync character based interface. So that one too >> requires that you do DDCMP in software. But I didn't think simh simulated a DUP11? >> >> Johnny >> >> On 2021-11-18 23:33, R. Voorhorst wrote: >>> Yes, but they are async lines and these are indeed character based >>> with >> software determining the protocol, not so the sync liness. Async >> decnet ddcmp behaves slightly differently from sync decnet ddcmp in >> number of characters used. Look at the packet contents in used >> characters and in between. >>> >>> Reindert >>> >>> -----Original Message----- >>> From: owner-hecnet at Update.UU.SE [mailto:owner- >> hecnet at Update.UU.SE] On >>> Behalf Of Johnny Billquist >>> Sent: Thursday, 18 November, 2021 23:23 >>> To: hecnet at Update.UU.SE >>> Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> >>> should be Ddcmp --Re >>> >>> Well... RSX for example can talk DDCMP over DZ11 or DL11 lines just >>> as well, >> which are plain simple serial lines existing in simh. >>> Hopefully such lines can be connected with interfaces doing DDCMP in >> hardware (simulated in simh of course), and it should all work. >>> >>> And any software running in RSX just have an interface that gives >>> the >> DDCMP packet layer. The actual serial lines are not accessible as >> such by any other software than the DDCMP driver. >>> >>> Johnny >>> >>> On 2021-11-18 23:01, R. Voorhorst wrote: >>>> Point in this case is that the ddcmp is handled in Simh and not in >>>> system >> software so software can only deal with packets/messages; there is >> need for a simh sync line where the software itself will be able to >> communicate in bytes/characters and synthesize the packets/messages >> as the next higher organization. >>>> >>>> Reindert >>>> >>>> -----Original Message----- >>>> From: owner-hecnet at Update.UU.SE [mailto:owner- >> hecnet at Update.UU.SE] On >>>> Behalf Of Johnny Billquist >>>> Sent: Thursday, 18 November, 2021 22:39 >>>> To: hecnet at Update.UU.SE >>>> Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> >>>> should be Ddcmp >>>> >>>> Well, a serial line over which DDCMP is running is purely a packet >>>> based >> interface. >>>> >>>> Johnny >>>> >>>> On 2021-11-18 21:46, R. Voorhorst wrote: >>>>> Indeed it is Ddcmp, but the meaning is the difference between >>>>> packet/message oriented and character oriented and the latter >>>>> should >> also exist. >>>>> >>>>> Reindert >>>>> >>>>> -----Original Message----- >>>>> From: owner-hecnet at Update.UU.SE [mailto:owner- >> hecnet at Update.UU.SE] >>>>> On Behalf Of Robert Armstrong >>>>> Sent: Thursday, 18 November, 2021 21:11 >>>>> To: hecnet at Update.UU.SE >>>>> Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) >>>>> >>>>>> Paul Koning <paulkoning at comcast.net> wrote: >>>>>> They speak DDCMP, yes, but not DECnet. Any protocol layered over >>>>>> DDCMP >>>>> should work. >>>>> >>>>> Indeed, and we just proved that with ANF10 which uses the >>>>> DMCs and DUPs but is not DECnet. >>>>> >>>>> Bob >>>>> >>>>> >>>>> >>>> >>> >> >> -- >> Johnny Billquist || "I'm on a bus >> || on a psychedelic trip >> email: bqt at softjar.se || Reading murder books >> pdp is alive! || tryin' to stay hip" - B. Idol > > -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

3 years

PyDECnet setup

by wonko＠4amlunch.net

That other thread is getting out of hand so I'll start a new one. I've started getting this error: Exception in thread https: Traceback (most recent call last): ?File "/usr/lib/python3.8/threading.py", line 932, in _bootstrap_inner ???self.run() ?File "/usr/lib/python3.8/threading.py", line 870, in run ???self._target(*self._args, **self._kwargs) ?File "/usr/local/lib/python3.8/dist-packages/decnet-1.0.592-py3.8.egg/decnet/http.py", line 103, in serverstart ???httpd = DECnetMonitor (server_addr, DECnetMonitorRequest, ?File "/usr/local/lib/python3.8/dist-packages/decnet-1.0.592-py3.8.egg/decnet/http.py", line 126, in __init__ ???self.socket = source_addr.create_server () ?File "/usr/local/lib/python3.8/dist-packages/decnet-1.0.592-py3.8.egg/decnet/host.py", line 399, in create_server ???sock = self.bind_socket (self.listen_family) ?File "/usr/local/lib/python3.8/dist-packages/decnet-1.0.592-py3.8.egg/decnet/host.py", line 387, in bind_socket ???sock.bind (self.sockaddr) OSError: [Errno 98] Address already in use Neither lsof nor netstat show that port in use. Doesn't matter if I use the default 8000 or set something different, I get that every time. I didn't always though. This was running fine before I tried adding the circuit for ethernet access. Thoughts? -brian

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re:4

by R.Voorhorst＠swabhawat.com

Well, that leaves 0,01% chance to see that assumption defeated so why not test it? I'll see next week if that can be tested if a software testbed can be found to support the single Dup11 that wise: VMS at least does not in the Decnet corner but Vax Psi may be different. Maybe Rsx-11M+ would be a suitable testbed? Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Mark Pizzolato - Info Comm Sent: Friday, 19 November, 2021 05:19 To: hecnet at Update.UU.SE Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re3 I have not followed the details of this conversation, BUT I'm 99.99% sure that a PDP11 can be configured with just a DUP with DECnet using that interface, and that DUP can be "wired" to another simulator which has a DUP, KDP/DUP or a DMC and DECnet can be spoken just fine. - Mark On Thursday, November 18, 2021 at 3:35 PM, Reindert wrote: > > No, the Dup is only simulated as Kdp/Dup combo, hence the Ddcmp > characteristics of the pair. > > > Reindert > > -----Original Message----- > From: owner-hecnet at Update.UU.SE [mailto:owner- hecnet at Update.UU.SE] On > Behalf Of Johnny Billquist > Sent: Friday, 19 November, 2021 00:31 > To: hecnet at Update.UU.SE > Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should > be Ddcmp --Re2 > > Fair enough. > > But DUP11 is a sync character based interface. So that one too > requires that you do DDCMP in software. But I didn't think simh simulated a DUP11? > > Johnny > > On 2021-11-18 23:33, R. Voorhorst wrote: > > Yes, but they are async lines and these are indeed character based > > with > software determining the protocol, not so the sync liness. Async > decnet ddcmp behaves slightly differently from sync decnet ddcmp in > number of characters used. Look at the packet contents in used > characters and in between. > > > > Reindert > > > > -----Original Message----- > > From: owner-hecnet at Update.UU.SE [mailto:owner- > hecnet at Update.UU.SE] On > > Behalf Of Johnny Billquist > > Sent: Thursday, 18 November, 2021 23:23 > > To: hecnet at Update.UU.SE > > Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> > > should be Ddcmp --Re > > > > Well... RSX for example can talk DDCMP over DZ11 or DL11 lines just > > as well, > which are plain simple serial lines existing in simh. > > Hopefully such lines can be connected with interfaces doing DDCMP in > hardware (simulated in simh of course), and it should all work. > > > > And any software running in RSX just have an interface that gives > > the > DDCMP packet layer. The actual serial lines are not accessible as such > by any other software than the DDCMP driver. > > > > Johnny > > > > On 2021-11-18 23:01, R. Voorhorst wrote: > >> Point in this case is that the ddcmp is handled in Simh and not in > >> system > software so software can only deal with packets/messages; there is > need for a simh sync line where the software itself will be able to > communicate in bytes/characters and synthesize the packets/messages as > the next higher organization. > >> > >> Reindert > >> > >> -----Original Message----- > >> From: owner-hecnet at Update.UU.SE [mailto:owner- > hecnet at Update.UU.SE] On > >> Behalf Of Johnny Billquist > >> Sent: Thursday, 18 November, 2021 22:39 > >> To: hecnet at Update.UU.SE > >> Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> > >> should be Ddcmp > >> > >> Well, a serial line over which DDCMP is running is purely a packet > >> based > interface. > >> > >> Johnny > >> > >> On 2021-11-18 21:46, R. Voorhorst wrote: > >>> Indeed it is Ddcmp, but the meaning is the difference between > >>> packet/message oriented and character oriented and the latter > >>> should > also exist. > >>> > >>> Reindert > >>> > >>> -----Original Message----- > >>> From: owner-hecnet at Update.UU.SE [mailto:owner- > hecnet at Update.UU.SE] > >>> On Behalf Of Robert Armstrong > >>> Sent: Thursday, 18 November, 2021 21:11 > >>> To: hecnet at Update.UU.SE > >>> Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) > >>> > >>>> Paul Koning <paulkoning at comcast.net> wrote: > >>>> They speak DDCMP, yes, but not DECnet. Any protocol layered over > >>>> DDCMP > >>> should work. > >>> > >>> Indeed, and we just proved that with ANF10 which uses the > >>> DMCs and DUPs but is not DECnet. > >>> > >>> Bob > >>> > >>> > >>> > >> > > > > -- > Johnny Billquist || "I'm on a bus > || on a psychedelic trip > email: bqt at softjar.se || Reading murder books > pdp is alive! || tryin' to stay hip" - B. Idol

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --> Re5

by R.Voorhorst＠swabhawat.com

Yes, my idea as well and sync idle could probably be simulated with a 1 character sync packet (Udp?) every 10 sec. or so to determine (dead) line timeouts. So a certain simh standard for simulation for this should first be defined. Nice job for the creative programmers amongst us ... Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Paul Koning Sent: Friday, 19 November, 2021 01:12 To: hecnet at update.uu.se Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re3 It seems like a very simple matter to simulate it as a basic character device, separate from the KMC pairing. The only real question is how you represent a sync data stream in a simullated device. Is it just the byte stream over a TCP connection? That would be easy for synchronous character mode. If you also want to simulate HDLC sync devices, that's a different matter entirely; there isn't an obvious way to do that. In other words, if someone wants this, just design how it communicates and do the code, it doesn't sound hard. paul > On Nov 18, 2021, at 6:35 PM, R. Voorhorst <R.Voorhorst at swabhawat.com> wrote: > > No, the Dup is only simulated as Kdp/Dup combo, hence the Ddcmp characteristics of the pair. > > > Reindert > > -----Original Message----- > From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Johnny Billquist > Sent: Friday, 19 November, 2021 00:31 > To: hecnet at Update.UU.SE > Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re2 > > Fair enough. > > But DUP11 is a sync character based interface. So that one too requires that you do DDCMP in software. But I didn't think simh simulated a DUP11? > > Johnny

3 years

ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re2

by R.Voorhorst＠swabhawat.com

Yes, but they are async lines and these are indeed character based with software determining the protocol, not so the sync liness. Async decnet ddcmp behaves slightly differently from sync decnet ddcmp in number of characters used. Look at the packet contents in used characters and in between. Reindert -----Original Message----- From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Johnny Billquist Sent: Thursday, 18 November, 2021 23:23 To: hecnet at Update.UU.SE Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should be Ddcmp --Re Well... RSX for example can talk DDCMP over DZ11 or DL11 lines just as well, which are plain simple serial lines existing in simh. Hopefully such lines can be connected with interfaces doing DDCMP in hardware (simulated in simh of course), and it should all work. And any software running in RSX just have an interface that gives the DDCMP packet layer. The actual serial lines are not accessible as such by any other software than the DDCMP driver. Johnny On 2021-11-18 23:01, R. Voorhorst wrote: > Point in this case is that the ddcmp is handled in Simh and not in system software so software can only deal with packets/messages; there is need for a simh sync line where the software itself will be able to communicate in bytes/characters and synthesize the packets/messages as the next higher organization. > > Reindert > > -----Original Message----- > From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On > Behalf Of Johnny Billquist > Sent: Thursday, 18 November, 2021 22:39 > To: hecnet at Update.UU.SE > Subject: Re: [HECnet] ANF10 network --> DN200 (RJE, Decnet) --> should > be Ddcmp > > Well, a serial line over which DDCMP is running is purely a packet based interface. > > Johnny > > On 2021-11-18 21:46, R. Voorhorst wrote: >> Indeed it is Ddcmp, but the meaning is the difference between >> packet/message oriented and character oriented and the latter should also exist. >> >> Reindert >> >> -----Original Message----- >> From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On >> Behalf Of Robert Armstrong >> Sent: Thursday, 18 November, 2021 21:11 >> To: hecnet at Update.UU.SE >> Subject: RE: [HECnet] ANF10 network --> DN200 (RJE, Decnet) >> >>> Paul Koning <paulkoning at comcast.net> wrote: >>> They speak DDCMP, yes, but not DECnet. Any protocol layered over >>> DDCMP >> should work. >> >> Indeed, and we just proved that with ANF10 which uses the DMCs >> and DUPs but is not DECnet. >> >> Bob >> >> >> > -- Johnny Billquist || "I'm on a bus || on a psychedelic trip email: bqt at softjar.se || Reading murder books pdp is alive! || tryin' to stay hip" - B. Idol

3 years

ANF10 network --> DN200/DN20 --> instant gratification ...? -->throttling

by R.Voorhorst＠swabhawat.com

Nope: 2000/1 means 2000 instructions and wat 1 msec which is the resultion of the sleep timer. Reindert From: owner-hecnet at Update.UU.SE [mailto:owner-hecnet at Update.UU.SE] On Behalf Of Supratim Sanyal Sent: Thursday, 18 November, 2021 21:06 To: hecnet at Update.UU.SE Subject: Re: [HECnet] ANF10 network --> DN200/DN20 --> instant gratification ...? On 11/18/21 1:35 PM, Paul Koning wrote: On Nov 18, 2021, at 1:26 PM, Supratim Sanyal <mailto:supratim at riseup.net> <supratim at riseup.net> wrote: On 11/18/21 11:10 AM, R. Voorhorst wrote: Hi, Hi. ... 2. The simh config file for configuring a working simh Pdp11/34 with the proper throttling to make it work; the proper setting is important because of Chk11 init phase which is instruction timing based!!! How does one arrive at a realistic argument for throttle? Did you put 2000/1 because you reckon your host processor does 2000 instructions in the same time as the guest processor does 1? Or is it more pure clock speed based? SET THROTTLE 2000/1 Check the docs ("help set throttle"). The simplest is to say what instruction rate you want: "set throttle 300k" is probably a pretty realistic setting to make it look like an 11/20; "set throttle 2m" is a somewhat optimistic 11/70. I am interpreting the "2000/1" as "wait 2 seconds after one instruction" which cannot be right, hence my question. BTW, throttle is dicey; my MicroVAX 3800 (the SIMH "vax") machines are set to 1.9 mips (3.8 times of the 0.5 mip 11/780), but still they grab 100% of a core running *BSD. But this is a discussion for the SIMH forum, not here. paul --

3 years

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

HECnet November 2021