Rev 22 on the Prime emulator
Jim Wilcoxson
Prime emulator. All emulators are accessed via telnet with different
ports:
$ telnet prirun.dyndns.org 8001 (rev 19)
$ telnet prirun.dyndns.org 8002 (rev 20)
$ telnet prirun.dyndns.org 8003 (rev 21)
$ telnet prirun.dyndns.org 8004 (rev 22)
Here's a sample login session:
$ telnet prirun.dyndns.org 8004
Trying xx.xx.xx.xx...
Connected to prirun.dyndns.org.
Escape character is '^]'.
Welcome to the Prime Computer 50-series emulator, running Primos rev
22.1.4!
After logging in, use the Prime HELP command for assistance.
You are welcome to create a directory under GUEST for your files.
To report bugs or contact the author, send email to pri...@gmail.com
Enjoy your time travels! -Jim Wilcoxson aka JIMMY
Login please.
login jimmy
Password?
JIMMY (user 2) logged in Sunday, 03 Feb 08 12:55:36.
Welcome to PRIMOS version 22.1.4c
Copyright (c) Prime Computer, Inc. 1988.
Last login Sunday, 03 Feb 08 12:47:00.
OK, ab .abbrev
OK, ab -nq -ac pwd [unquote 'type [dir [pathname *]]']
OK, ab -nq -ac up [unquote 'a [dir [dir [pathname *]]];type [dir
[pathname *]]']
OK, co -end
OK, lo
JIMMY (user 2) logged out Sunday, 03 Feb 08 12:55:48.
Time used: 00h 00m connect, 00m 00s CPU, 00m 00s I/O.
Prime session disconnected
Connection closed by foreign host.
$
An ex-Prime user contributed a nice set of rev 22 Magsav dumps,
including:
- SPL
- PLP
- Midasplus
- FTN, the Fortran 66 compiler
- F77, the Fortran 77 compiler
- Basic/VM, the Basic compiler
- CI, the I-mode C compiler
- DBG, the source level debuger
- EMACS
Prime introduced a number of new features with rev 22:
- a new disk format, with support for extents (CAM files)
- disk mirroring
- support for dual-ported disk drives
- robust partitions
There were many other changes and enhancements. For details, attach
to INFO22.1, INFO22.1.1, INFO22.1.2, and INFO22.1.3, and look at
the .RUNO update files. There's also an UPINFO22.1.4 file: Prime used
those for smaller updates, where all update information was listed in
a single file.
There are a few installation problems, because it was hard to restore
the Magsav data from the dd format I had. DSM still isn't working
(sorry Andrew!), and complains about the SIT data. I think this is
related to Prime's internationalization support, and I don't have it
set up right. If someone knows how to set this up, shoot me an email.
All serialization has been removed, but if you see anything that I
missed, please send me a private email so I can remove it.
I'm hoping to get rev 23 configured next week.
Enjoy!
Jim
prirun at gmail.com
--
Software first. Software lasts!
Jim Wilcoxson
pr1me (one in the middle).
Enjoy!
Jim
prirun at gmail.com
On Feb 3, 1:09 pm, Jim Wilcoxson <pri...@gmail.com> wrote:
> Well, I finally got around to configuring a rev 22 system for the
> Prime emulator. All emulators are accessed via telnet with different
> ports:
>
> $ telnet prirun.dyndns.org 8001 (rev 19)
> $ telnet prirun.dyndns.org 8002 (rev 20)
> $ telnet prirun.dyndns.org 8003 (rev 21)
> $ telnet prirun.dyndns.org 8004 (rev 22)
--
Software first. Software lasts!
ap...@student.open.ac.uk
> Well, I finally got around to configuring a rev 22 system for the
> Prime emulator.
> There are a few installation problems, because it was hard to restore
> the Magsav data from the dd format I had. DSM still isn't working
> (sorry Andrew!), and complains about the SIT data. I think this is
> related to Prime's internationalization support, and I don't have it
> set up right. If someone knows how to set this up, shoot me an email.
Yes, SIT stands for Software Internationalization Tools. It uses an
EPF database for the internationalized string lookups. I can't
remember where it looks for them but I will login to the emulator and
have a poke around. It may be under DSM* somewhere.
Regards,
Andrew Marlow
Big and Blue
> Yes, SIT stands for Software Internationalization Tools. It uses an
> EPF database for the internationalized string lookups. I can't
> remember where it looks for them
Probably SIT*, since that existed.
file:///media/cdrom/cd/_bin/sit*
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_csd.edimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_csd_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_ed.edimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_ed_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_lnd.edimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_lnd_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_ncd.edimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_ncd_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_td.edimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_td_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_text_dbs/sit_usa.tdimg
file:///media/cdrom/cd/_bin/sit*/sit_characters_db.csimg
file:///media/cdrom/cd/_bin/sit*/sit_language_name_db.lnimg
file:///media/cdrom/cd/_bin/sit*/sit_national_cons_db.ncimg
--
Just because I've written it doesn't mean that
either you or I have to believe it.
marlow...@googlemail.com
> Well, I finally got around to configuring a rev 22 system for the
> Prime emulator. All emulators are accessed via telnet with different
> ports:
>
> $ telnet prirun.dyndns.org 8001 (rev 19)
> $ telnet prirun.dyndns.org 8002 (rev 20)
> $ telnet prirun.dyndns.org 8003 (rev 21)
> $ telnet prirun.dyndns.org 8004 (rev 22)
I just tried rev 22 and DSM is not running. I know there are problems
with SIT. Is that why the DSM server is not running? How about the
SYSTEM_MANAGER? Thats not running either. The SIM commands
(LIST_PROCESSES etc) still give errors about the start and stop time.
I thought the Y2K problems were fixed here.
Also GUEST has attach rights only to DSM*. Can it have at least read
access please?
Jim Wilcoxson
> Also GUEST has attach rights only to DSM*. Can it have at least read
> access please?
I changed the DSM* ACL so GUEST has LUR access. If you figure out
what's messed up in the DSM installation, lemme know. I'm pretty sure
there are still Y2K issues with rev 22, because when it says "1908" in
compiler listings instead of 2008.
Jim
ap...@student.open.ac.uk
> On Feb 5, 11:17 am, marlow.and...@googlemail.com wrote:
>
> > Also GUEST has attach rights only to DSM*. Can it have at least read
> > access please?
>
> I changed the DSM* ACL so GUEST has LUR access. If you figure out
> what's messed up in the DSM installation, lemme know.
I'll take a look.
> I'm pretty sure
> there are still Y2K issues with rev 22, because when it says "1908" in
> compiler listings instead of 2008.
>
> Jim
I wonder if things would work if you told the emulator that the year
is 1958. That has the same calendar as 2008.
Is there any chance you could get the Y2K fixes from Computronics?
Jim Wilcoxson
> I wonder if things would work if you told the emulator that the year
> is 1958. That has the same calendar as 2008.
>
> Is there any chance you could get the Y2K fixes from Computronics?
Randy kindly offered his Y2K toolkit to me a few months ago, at no
charge for the public emulators I have running here. He is also still
offering commercial licenses.
I fixed the Y2K problems in Primos rev 19 by correcting the source and
will look into doing it for rev 22. When I get rev 23 configured, DSM
will work: I've seen it working with 23.2. To me, it's not such a big
deal that some software doesn't work with earlier revs. The date bugs
were there in Primos and/or DSM, so from this viewpoint, it is what it
is (or was!)
It bothers me a lot more that for historical accuracy and interest,
there is so much Prime software and source code that is still nowhere
to be found. I'd love to get my hands on early versions of Primos,
like the early teens, but sadly, that will probably never happen. If
anyone has Prime software (anything is welcome) they'd like to
contribute to the project, please get in touch with me.
ap...@student.open.ac.uk
> On Feb 6, 3:28 am, ap...@student.open.ac.uk wrote:
> > Is there any chance you could get the Y2K fixes from Computronics?
>
> Randy kindly offered his Y2K toolkit to me a few months ago, at no
> charge for the public emulators I have running here. He is also still
> offering commercial licenses.
Yes, Randy emailed me about this. I think that you think there is some
sort of license issue here. Perhaps you could explain please?
>
> I fixed the Y2K problems in Primos rev 19 by correcting the source and
> will look into doing it for rev 22.
Impressive.
> When I get rev 23 configured, DSM
> will work: I've seen it working with 23.2.
That sounds exciting!
> To me, it's not such a big
> deal that some software doesn't work with earlier revs. The date bugs
> were there in Primos and/or DSM, so from this viewpoint, it is what it
> is (or was!)
Since DSM was *the* main thing I worked on this is what would make it
a trip to Nostalgia City for me!
>
> It bothers me a lot more that for historical accuracy and interest,
> there is so much Prime software and source code that is still nowhere
> to be found.
Yes, this is truly amazing. You would think that there would be tapes
of all the source for all the apps somewhere, not to mention the PETs
and PETIs. Sure, the source for PRIMOS seems to be about in various
places but there were loads of apps too. Quite a bit was developed in
the UK. We maintained the source (using Clive Backhams CMAN product)
but we had to ship tapes to the US so I presume they got copies too.
What happened to them all? Its a mystery.
I wonder if anyone knows what happened in the UK. The main research
centre at Willen Lake was taken over by a company called VMark
Software. I wonder if any ex-primates that stayed at VMark know what
happened?
> I'd love to get my hands on early versions of Primos,
> like the early teens, but sadly, that will probably never happen.
Well, steady on. My introduction to PRIMOS was on a P400 running rev
17. It seemed good at the time but now it would probably be regarded
by most as a pile of poo. It didnt even have user logins! No ACLs, no
CPL, no FORTRAN 77 (only FORTRAN IV).
Jim Wilcoxson
> Well, steady on. My introduction to PRIMOS was on a P400 running rev
> 17. It seemed good at the time but now it would probably be regarded
> by most as a pile of poo. It didnt even have user logins! No ACLs, no
> CPL, no FORTRAN 77 (only FORTRAN IV).
Wait - my intro to Primos was rev 12, with 6-character filenames (rev
13 introduced 32-character filenames), and we had user logins even
back then. There was no central user database like the SAD that came
into being at rev 19. The way it worked is that any top-level UFD
(User File Directory) was also a valid login. UFD's had an owner and
non-owner password. You had to know the owner password to login.
Unfortunately, all of the Prime system directories were shipped
without passwords, so anyone knowing this could login to SYSTEM,
CMDNC0, LIB, etc! Individual files had both owner and non-owner
protection numbers. So when you attached to a UFD as a non-owner, you
only got the non-owner access rights for all of the files. If you
didn't know the owner or non-owner password, you couldn't attach to
the UFD. There was also a "magic" password that worked on the MFD: I
think it was ^z^y^x^w^v^u, or something like that. It always made you
an owner of the MFD. With that, you could use GPASS subroutine calls
to get the owner and non-owner password of any directory underneath.
Funny, eh?
Prime corrected all this nonsense with ACL directories and user
profiles at rev 19.
Jim
ap...@student.open.ac.uk
> On Feb 8, 12:10 pm, ap...@student.open.ac.uk wrote:
>
> > Well, steady on. My introduction to PRIMOS was on a P400 running rev
> > 17. It seemed good at the time but now it would probably be regarded
> > by most as a pile of poo. It didnt even have user logins! No ACLs, no
> > CPL, no FORTRAN 77 (only FORTRAN IV).
>
> Wait - my intro to Primos was rev 12, with 6-character filenames (rev
> 13 introduced 32-character filenames), and we had user logins even
> back then.
Well that sounds like a large quivering pile of poo emitting great
gobs of green steam!
> There was no central user database like the SAD that came
> into being at rev 19. The way it worked is that any top-level UFD
> (User File Directory) was also a valid login.
It was still like that at rev 17, which was a gaping security hole.
That is why many universities hacked the source of PRIMOS to implement
proper logins. In those days the source of PRIMOS was given to unis.
> UFD's had an owner and
> non-owner password. You had to know the owner password to login.
> Unfortunately, all of the Prime system directories were shipped
> without passwords, so anyone knowing this could login to SYSTEM,
> CMDNC0, LIB, etc!
Indeed. What a blunder.
> Individual files had both owner and non-owner
> protection numbers. So when you attached to a UFD as a non-owner, you
> only got the non-owner access rights for all of the files. If you
> didn't know the owner or non-owner password, you couldn't attach to
> the UFD.
As I recall, hardly any use was made of this, even though it was
potentially useful. I did see if used once in the post rev18 era as a
poor man's simulation of an 'su'. As we know, in Unix you can made a
program SUID so that it can read/writes that the invoking user cannot
read directly. What this Prime program did was to achieve something
similar by putting the files in an old-style password directory. The
program knew the password but ordinary people did not. Unfortunately
this was not thought through properly and the program stored the
password in plaintext, making it very easy to grab.
> There was also a "magic" password that worked on the MFD: I
> think it was ^z^y^x^w^v^u, or something like that. It always made you
> an owner of the MFD. With that, you could use GPASS subroutine calls
> to get the owner and non-owner password of any directory underneath.
> Funny, eh?
I don't know whether to laugh or cry....
>
> Prime corrected all this nonsense with ACL directories and user
> profiles at rev 19.
Yes, but that came in at rev 18 didn't it? I really like the PRIMOS
ACL system, I think it is much better than Unix permissions and indeed
anything else in this area that I have seen in other OSs.
> Jim
Big and Blue
> What this Prime program did was to achieve something
> similar by putting the files in an old-style password directory. The
> program knew the password but ordinary people did not. Unfortunately
> this was not thought through properly and the program stored the
> password in plaintext, making it very easy to grab.
I came across a program that assembled the password at run time, which
was somewhat better, but didn't stop anyone who could use (V)PSD. Mind you,
in the end I just patched the ATTACH call so that the first arg had a value
of K$HOME rather than K$CURR (just had to point it a a 16-bit 1 rather than
a 16-bit 0, IIRC - easy enough to create a 16-bit 1 in the first 512 bytes
if necessary) and knew I'd end up attached as owner when the program
stopped, which was simpler...
ap...@student.open.ac.uk
> Well, I finally got around to configuring a rev 22 system for the
> Prime emulator. All emulators are accessed via telnet with different
> ports:
>
> $ telnet prirun.dyndns.org 8001 (rev 19)
> $ telnet prirun.dyndns.org 8002 (rev 20)
> $ telnet prirun.dyndns.org 8003 (rev 21)
> $ telnet prirun.dyndns.org 8004 (rev 22)
I cannot connect to any of these at the moment; I get connection
refused. Are they down?
> I'm hoping to get rev 23 configured next week.
How's rev23 coming along please?
-Andrew Marlow
Jim Wilcoxson
On Feb 21, 2:57 pm, ap...@student.open.ac.uk wrote:
> I cannot connect to any of these at the moment; I get connection
> refused. Are they down?
No, they're not down. I'm using siteuptime.com to monitor the
emulators every 5 minutes, and it hasn't reported any outages. I'm
monitoring it from a server in London, figuring that if it works from
there, it'll work from in the US.
> How's rev23 coming along please?
I'll try to get to it soon...