Re: HPE OpenVMS Hobbyist license program is closing
On Wed, Mar 11, 2020 at 08:08:43AM -0700, Al Kossow via cctalk wrote: > > > Years ago us SGI hobbyists were able to talk to SGI about this and a huge > > problem SGI had with any kind of hobbyist > > license for IRIX or turning it free is it's fully of licensed 3rd party > > stuff. But maybe now that's it's expired, or all > > the companies things were licensed from are gone. > > Release of Classic HP3000 died for the same reason from the same company. > > HP has no motivation to spend time/money to release this. Also, the only way > CHM > was able to release what we did (HP1000, 68K 9000 and Apollo) was having > us release it only for non-commercial research/hobby purposes. Which would be great - anyone still running OpenVMS on VAX for business purposes presumably has the appropriate (non-expiring) licenses anyway. So having OpenVMS VAX (and maybe even Alpha) released with an eternal hobbyist license (strictly non-commercial research/hobby purposes) would not only be a great win, but all we can really ask for. Kind regards, Alex. -- "Opportunity is missed by most people because it is dressed in overalls and looks like work." -- Thomas A. Edison
Re: Prime 50-Series emulator announcement
> > So, how do I login a regular user? Does the emulator come up with > > an open port that I can telnet into to get a regular serial line? > Use the emulator -tport command line option to set its incoming > terminals port. Bill, The inittab snippet from Jim's message that I reposted here shows you how the runem script expectes to be invoked, and run shows you the various options used to run the public emulators. De
Re: Prime 50-Series emulator announcement
Bill, On Thu, Mar 12, 2020 at 02:50:46PM -0400, Bill Gunshannon via cctalk wrote: > So, how do I login a regular user? Does the emulator come up with an > open port that I can telnet into to get a regular serial line? Use the emulator -tport command line option to set its incoming terminals port. -- Kevin http://www.RawFedDogs.net http://www.Lassie.xyz http://www.WacoAgilityGroup.org Bruceville, TX What's the definition of a legacy system? One that works! Errare humanum est, ignoscere caninum.
Re: Prime 50-Series emulator announcement
All I can say is Wow!! I have wanted to see this for quite some time. It took no time at all to build the emulator and I fired it up with Rev 19.2 which was the one I had the most experience with in the past. Came up very nicely. Really sorry I got rid of all that Prime documentation. :-( Luckily I did keep all my Pocket References so I am not completely out of the picture. Now the questions start... OK, just one for now. When I started up the emulator it ended with the Supervisor on the console. So, how do I login a regular user? Does the emulator come up with an open port that I can telnet into to get a regular serial line? I can't wait to play with all the compilers I used to use. One more question, I guess. Is there any likelihood that any of the third party software like EDV or any of the Salford or Shefield packages still exist in someone's storage unit? I wonder if there are any chances people like Minitab or SPSS could be convinced to let the old Prime versions of their software loose? bill
Re: Prime documentation
Dennis, On Thu, Mar 12, 2020 at 10:51:52AM -0400, Dennis Boone via cctalk wrote: > To continue through to the end of the startup process, tell it: > > CO CONTINUE 6 The system appears to have a few more issues. After the above: CO CONTINUE 6 OK,CO SYSTEM>DSM.SHARE.COMI 7 /* Share DSM facility Not found. (CO) ER! CO CONTINUE 6 OK,START_DSM /* Start DSM phantoms [START_DSM Rev. 22.1.4.R48 Copyright (c) 1991, Prime Computer, Inc.] [Serial #IDBR-G5RKKW-J6BW (ALBERTUS MAGNUS COLLEGE)] [12 Mar 20 10:06:44 Thursday] DSM initialization started. OK, /* CO SYSTEM>SPOOL.SHARE.COMI 7 /* Share SPOOL database OK, /* OK, /* CO SYSTEM>SPL.SHARE.COMI 7 /* Share SPL compiler OK, /* CO SYSTEM>PLP.SHARE.COMI 7 /* Share PLP compiler OK, /* CO SYSTEM>PMED.SHARE.COMI 7/* Share EDS OK, /* OK,PROP PR0 -START/* Start PR0 printer [PROP Rev. 22.1.4 Copyright (c) 1990, Prime Computer, Inc.] [Serial #IDB8-G5RKKX-K90X (ALBERTUS MAGNUS COLLEGE)] Error from Spooler (PROCESS_ASYNC-88): Not found.: Opening environment file ER! *** DSMSR (user 17) at 10:06 Unable to open journal Insufficient access rights.. ( *** DSMSR (user 17) at 10:06 Error: Unable to write loaded configuration file. () Phantom 18: Normal logout at 10:06 Time use *** SYSTEM_MANAGER (user 19) at 10:06 Unable to open journal DSM*>JOURNALS>SYSTEM_MANAGER.2. (Insufficient access rig Warning: SYSTEM_MANAGER terminated. Event logging will not take place! User 19: Phantom requested terminal input. d: 00h 00m connect, 00m 00s CPU, 00m 00s I/O. Phantom 19: Abnormal logout at 10:06 Time used: 00h 00m connect, 00m 01s CPU, 00m 00s I/O. Phantom 17: Normal logout at 10:06 Time used: 00h 00m connect, 00m 01s CPU, 00m 00s I/O. -- Kevin http://www.RawFedDogs.net http://www.Lassie.xyz http://www.WacoAgilityGroup.org Bruceville, TX What's the definition of a legacy system? One that works! Errare humanum est, ignoscere caninum.
Re: Prime documentation
> OK,CO SYSTEM>ED.SHARE.COMI 7 /* Share ED editor > Not found. (CO) > ER! To continue through to the end of the startup process, tell it: CO CONTINUE 6 Until this is fixed, you'll have to use the non-shared editor, NSED, instead of the shared editor ED. > Thank you for all the work you've put in to making these systems and > this wealth of PRIME information and resources available to hobbyists!!! Most of the credit goes to Jim, who invested a stupendous number of hours in writing the emulator, and a bunch more in scanning a lot of those manuals. De
Re: Prime documentation
Dennis, On Wed, Mar 11, 2020 at 09:20:08PM -0400, Dennis Boone via cctalk wrote: > I was able to reproduce the issue. When I repacked these using the > Rev. 22 MAGSAV, the install worked correctly. Here's the repacked tape > set: > > https://yagi.h-net.org/m2214repack.tar.gz I was able to restore from the above tapes, and have a bootable system. It looks like the system might have an editor issue. At the tail end of the boot, I'm getting: OK,CO SYSTEM>ED.SHARE.COMI 7 /* Share ED editor Not found. (CO) ER! I just downloaded the sample system images. I'll use those to start getting familiar with PRIMOS, and get back to installing a system from scratch at a later date. Thank you for all the work you've put in to making these systems and this wealth of PRIME information and resources available to hobbyists!!! -- Kevin http://www.RawFedDogs.net http://www.Lassie.xyz http://www.WacoAgilityGroup.org Bruceville, TX What's the definition of a legacy system? One that works! Errare humanum est, ignoscere caninum.
Re: Prime 50-Series emulator announcement
A set of sample system images derived from the public emulators can be downloaded from https://yagi.h-net.org/p50em_samplemachines.tar.gz to get you started. The tarball is 309078820 bytes, and its sha256sum is 8fe261f7a9f19e9fab2814371387f9cef5c64161fe7cf4bdc542144202678ca9. We may rebuild these later and distribute them individually in a more formal fashion. De
Re: Make behavior on TOPS20 Panda distribution
On Thu, Mar 12, 2020, 2:43 AM Rich Alderson via cctalk < cctalk@classiccmp.org> wrote: > From: David Griffith > Sent: Wednesday, March 11, 2020 2:41 AM > > > I've successfully managed to get a simulated TOPS20 environment going > with > > networking. Now there's a new strange hitch: the behavior of make. > Would > > someone please look at this and tell me what I'm doing wrong? > > > For a simple Makefile, it works fine if I just type "make", but if I type > > "make ", I get the commands sent to the command line, but no > > responses. The display then stalls at the eighth source file. > Strangely, if > > I scroll up in the xterm, it gets dragged back down as if more stuff is > being > > printed. > > "make" is not native to TOPS-20, and is unknown to the vast majority of > TOPS-20 > programmers. The best thing to do is to get the source for make and run it > under DDT to see where it's failing to perform as you expect, and fix it > yourself. > > It is unfortunate that Unix make was ever ported to TOPS-20, because there > was > already a MAKE command which invoked TECO with an empty buffer. This is > the > source of the joke response to "MAKE LOVE"; it has nothing to do with your > issue with the Unix program. > Maybe he's running the native one? The error message suggests that may be the case. Warner Instead of using a badly ported Unix program, you ought to learn to build > TOPS-20 control files and do your compiles that way. > > Just my $0.02. > > Rich >
DDCMP over serial lines in RSTS V10.1
I thought that I had sent a message to the list a few months ago showing the
procedure to use the software DDCMP driver in RSTS V10.1. That's a standard
feature, but undocumented and probably unsupported. It works well and
interoperates with PyDECnet and with the DDCMP implementation in SIMH.
Presumably it will talk to the software DDCMP in RSX, though I have not tried
that.
Attached are the instructions and the program mentioned.
paul
Software DDCMP in RSTS/E V10
NODVR, a software DDCMP driver was written for V9.6 as part of PRO
support (an unreleased and unsanctioned effort). In V10.0 it was made
part of the released product. These notes were reverse engineered
from the V10.1 code, though it is likely they are valid also for
V10.0.
NODVR is an "auxiliary driver", essentially a software coprocessor for
a standard device. In this case, it is associated with a terminal
line, and hooks to the terminal driver to become an alternate
character processor. Any physical terminal port other than the
console can be operated in DDCMP mode, but the PRO display,
pseudo-keyboards, and LAT terminals cannot.
Unlike other DDCMP devices, NODVR does not support use from
application programs. It can only be used as a DECnet data link.
Enabling NODVR support in RSTS.
While the software is built into the OS during SYSGEN, it is not
loaded unless enabled by a flag in the system configuration file
[0,1]CONFIG.SYS. To set this flag:
$ set system/async
The change will take effect after a system reboot. You can use "show
system" to confirm that the software DDCMP driver is loaded.
Starting DDCMP on a terminal line.
Running the DDCMP driver on a given terminal line is a two step
process. The first step enables DDCMP mode by binding an NODVR unit
to the terminal line. The second step starts the DDCMP state machine
for that unit.
1. Set DDCMP mode for a terminal
Before this is done, either the terminal line must be not owned, or it
must be owned (assigned) by the requesting job.
To enable DDCMP mode for a given terminal line, issue this DCL
command:
$ set term kb/proto=ddcmp
Alternatively in a program, open NO0: and issue a .SPEC request with the
following arguments:
XRB: 3 (Function code: set DDCMP mode)
XRBC: Terminal unit number to set to DDCMP mode
On successful completion, the terminal line is now associated with the
DDCMP driver. As part of this, an NO: DDB has been created, where
is the terminal line unit number specified. The terminal line is
now marked as owned by DECnet (owner job appears as "TRN" in the SHOW
DEVICES/ALLOC command).
2. Start DDCMP circuit
Issue a DECnet circuit start request, which is a .MESAG with the
following arguments:
FQFIL = -21 (Function code: circuit control)
FQSIZM = 3 (Subfunction: circuit on)
FQ$MFL = Routing circuit flags [11]
LF.VER ==> Verification required
LF.ANS ==> Answer mode
LF.RST ==> Restartable (512)
LF.TRA ==> Trace for this circuit
FQNAM1 = Originating queue limit [7]
FQNAM1+2 = Recall timer [9]
FQBUFL = Hello timer interval [15]
FQMODE = Operational mode (see below) [17]
FQFLAG = Circuit cost [19]
FQ$CTM = Counter timer [21]
FQDEV,
FQDEVN = Device to use ("TT" in FQDEV, in FQDEVN) [23..26]
FQCLUS = Buffer quota (byte) [27]
Numbers in [ ] indicate character positions in the Basic-PLUS syscall.
FQMODE bit 14 (8192) set indicates a request for sync mode. This is
only valid on a PRO, and then only for the terminal line corresponding
to the communication port. No other mode bits apply to the async DDMP
driver.
Make sure the XRB is zeroed since this is not a data transferring case
of .MESAG. Note that the device name is given as "TT" not "KB" or
"NO"; the DECnet circuit start code recognizes this name as a special
case.
Stopping DDCMP
This involves the reverse of the two steps given above.
1. Stop DDCMP circuit
Issue a DECnet circuit stop request, which is a .MESAG with the
following arguments:
FQFIL = -21 (Function code: circuit control)
FQSIZM = 4 (Subfunction: circuit off)
FQDEV,
FQDEVN = Device to use ("TT" in FQDEV, in FQDEVN)
Make sure the XRB is zeroed since this is not a data transferring case
of .MESAG. Note that the device name is given as "TT" not "KB" or
"NO"; the DECnet circuit start code recognizes this name as a special
case.
2. Disable DDCMP mode for the terminal line
To return the terminal line that was used to normal operation rather
than being bound to the Async DDCMP driver, use the DCL command:
$ set term kb/proto=none
Alternatively in a program, open NO0: and issue a .SPEC request with the
following arguments:
XRB: 4 (Function code: set normal mode)
XRBC: Terminal unit number to revert to normal mode
On successful completion, the
RE: Make behavior on TOPS20 Panda distribution
From: David Griffith Sent: Wednesday, March 11, 2020 2:41 AM > I've successfully managed to get a simulated TOPS20 environment going with > networking. Now there's a new strange hitch: the behavior of make. Would > someone please look at this and tell me what I'm doing wrong? > For a simple Makefile, it works fine if I just type "make", but if I type > "make ", I get the commands sent to the command line, but no > responses. The display then stalls at the eighth source file. Strangely, if > I scroll up in the xterm, it gets dragged back down as if more stuff is being > printed. "make" is not native to TOPS-20, and is unknown to the vast majority of TOPS-20 programmers. The best thing to do is to get the source for make and run it under DDT to see where it's failing to perform as you expect, and fix it yourself. It is unfortunate that Unix make was ever ported to TOPS-20, because there was already a MAKE command which invoked TECO with an empty buffer. This is the source of the joke response to "MAKE LOVE"; it has nothing to do with your issue with the Unix program. Instead of using a badly ported Unix program, you ought to learn to build TOPS-20 control files and do your compiles that way. Just my $0.02. Rich
JOSS-II: good news and bad news
Hello, A small group of volunteers are working on restoring JOSS-II, the PDP-6 version, to a working state. We are typing in a program listing provided by the RAND corporation. The listing is the supervisor, which is a large part of the system. So far good news. The bad news is that it's not a complete system. The Computer History Museum seems to have additional program listings here: https://www.computerhistory.org/collections/catalog/102734547 But they have not been scanned. They archvists are willing to let us see the collection, but no one of us lives near the museum. Is there anyone interested in this kind of thing that are willing to help out?
