Alternative _legal_ operating systems for VAX ?

[Andy Burns]

I thought I would look for other operating systems which ran on VAX and try
them out. The only current one I have found is NetBSD

I looked a while ago, bit it doesn't seem to support DSSI disks, so no use for
my aforementioned VS3300.

[Arne Vajhøj]

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?

Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

NetBSD may be the only one with current support. But do you need to run
a current version??

You did not run a current VMS version on it.

And resource wise it may be a better fit for an older OS
anyway.

Alternatively you could take up the Linux VAX port. Some
work was done many years ago and I believe they got pretty
far.

Arne

[Scott Dorsey]

Post by Arne Vajhøj
Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.
NetBSD may be the only one with current support. But do you need to run
a current version??

What is the status on 4.1 and 4.2BSD? I know that V/32 is pretty much
all AT&T code.

There used to be a UCSD P-System kit available for the vax as well. Have
not thought about that in ages.

Also... what is the story with Ultrix-32? Are the licenses transferrable?
--scott

[Scott Dorsey]

Post by Scott Dorsey

Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu as
open source options for VAX.
NetBSD may be the only one with current support. But do you need to run
a current version??

What is the status on 4.1 and 4.2BSD? I know that V/32 is pretty much
all AT&T code.

I have 4.3 Quasijarus, but I have only run it on the simulator.

Yes. 4.3 is "not constrained by the AT&T license" whatever the lawyers really
mean by that.
--scott

[Johnny Billquist]

Post by Arne Vajhøj

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?

Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

Post by Arne Vajhøj
NetBSD may be the only one with current support. But do you need to run
a current version??
You did not run a current VMS version on it.
And resource wise it may be a better fit for an older OS
anyway.

That is definitely true. It's not that pleasant to run current NetBSD on
any VAX at the moment. But there are also some problems/issues that
we're waiting for a new version of gcc to come along, where they have
been fixed.

Post by Arne Vajhøj
Alternatively you could take up the Linux VAX port. Some
work was done many years ago and I believe they got pretty
far.

I think it booted to single user mode. Not sure it got much further. Not
sure anyone would even want to try and pick that up, as Linux is
constantly changing under the hood, making it a big task to resume
something abandoned years ago.

Johnny

[Scott Dorsey]

Post by Johnny Billquist

Post by Arne Vajhøj
Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

I think you are talking about two different things. Mt. Xinu made a
number of commercial products. XINU was a unixlike kernel without much
else; it was not a full "OS distribution" as we would call it today,
but it was open source. It was an interesting demo for students but
it was not actually useful for anything.
--scott

[Johnny Billquist]

Post by Scott Dorsey

Post by Johnny Billquist

Post by Arne Vajhøj
Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

I think you are talking about two different things. Mt. Xinu made a
number of commercial products. XINU was a unixlike kernel without much
else; it was not a full "OS distribution" as we would call it today,
but it was open source. It was an interesting demo for students but
it was not actually useful for anything.

Oh! My bad then. Hmm, was Xinu the Tannenbaum thing?

Johnny

[Bill Gunshannon]

Post by Johnny Billquist

Post by Johnny Billquist

Post by Arne Vajhøj
Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

I think you are talking about two different things.  Mt. Xinu made a
number of commercial products.  XINU was a unixlike kernel without much
else; it was not a full "OS distribution" as we would call it today,
but it was open source.  It was an interesting demo for students but
it was not actually useful for anything.

Oh! My bad then. Hmm, was Xinu the Tannenbaum thing?

No, the was Minix and I don;t believe it was ever ported to the VAX.

bill

[Scott Dorsey]

Post by Johnny Billquist

Post by Scott Dorsey

Post by Johnny Billquist

Post by Arne Vajhøj
Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

I think you are talking about two different things. Mt. Xinu made a
number of commercial products. XINU was a unixlike kernel without much
else; it was not a full "OS distribution" as we would call it today,
but it was open source. It was an interesting demo for students but
it was not actually useful for anything.

Oh! My bad then. Hmm, was Xinu the Tannenbaum thing?

No, that was Minix. Minix was actually more complete and useful than XINU.
--scott

[Simon Clubley]

Post by Scott Dorsey
No, that was Minix. Minix was actually more complete and useful than XINU.

Well, Intel certainly thinks it was useful. :-)

Simon.

[Johnny Billquist]

Post by Johnny Billquist

Post by Arne Vajhøj

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?

Wikipedia lists BSD 4.3, NetBSD, OpenBSD and something called Xinu
as open source options for VAX.

As far as I know Mt. Xinu (read it backwards? ;-) ) wasn't free. I ran
it on an 8650 for a while. I think I still know where the manuals are,
but I'm unsure if I know where the tapes are...

https://www.cs.purdue.edu/homes/dec/xlicense.html

My bad. Xinu and Mt. Xinu are apparently two different things. Sorry for
the noise.

Post by Johnny Billquist

Post by Arne Vajhøj
NetBSD may be the only one with current support. But do you need to run
a current version??
You did not run a current VMS version on it.
And resource wise it may be a better fit for an older OS
anyway.

That is definitely true. It's not that pleasant to run current NetBSD
on any VAX at the moment. But there are also some problems/issues that
we're waiting for a new version of gcc to come along, where they have
been fixed.

Post by Arne Vajhøj
Alternatively you could take up the Linux VAX port. Some
work was done many years ago and I believe they got pretty
far.

I think it booted to single user mode. Not sure it got much further.
Not sure anyone would even want to try and pick that up, as Linux is
constantly changing under the hood, making it a big task to resume
something abandoned years ago.

Remember that Simon believes using C makes porting easy.
Great opportunity to demonstrate.
:-)

Well, this has close to nothing to do with C.
If the API is changed, it's a headache.
And porting operating systems requires a lot of code to deal with the
specifics of the hardware, which is inherently not that portable.

And thus - if you have code that deals with the hardware, but which is
using an API that no longer exists, you need to either rewrite all the
code that deals with the hardware, or rewrite all the code that makes
use of the API to the rest of the OS, which might also force changes in
code that deals with the hardware.

Makes no difference which language you are working in, and no matter how
easy a language would be for porting, the problem here don't change.

All that said, I actually do believe that C is easier to port than
Macro-32. :-)

Johnny

[Scott Dorsey]

Sadly, language has little to do with it. Too much of the
DEC stuff is still proprietary and trade secret locked. Last
I heard that was the reason why neither NetBSD or OpenBSD ever
had a functioning X11.

Ever had a functioning X11 _server_. The clients work just fine,
it's just that the DEC framebuffers are nontrivial to develop a
display server for given the available documentation.
--scott

[Bill Gunshannon]

Post by Scott Dorsey

Sadly, language has little to do with it. Too much of the
DEC stuff is still proprietary and trade secret locked. Last
I heard that was the reason why neither NetBSD or OpenBSD ever
had a functioning X11.

Ever had a functioning X11 _server_. The clients work just fine,
it's just that the DEC framebuffers are nontrivial to develop a
display server for given the available documentation.

That was my point exactly. Information on the DEC video hardware
is not readily available and thus no X-server for the VAX. I
have a bunch of VAXServer-3100's. Not much point in putting one
on my desk if I have to put something else on my desk to view
the output from the VS3100.

bill

[Simon Clubley]

Post by Scott Dorsey

Sadly, language has little to do with it. Too much of the
DEC stuff is still proprietary and trade secret locked. Last
I heard that was the reason why neither NetBSD or OpenBSD ever
had a functioning X11.

Ever had a functioning X11 _server_. The clients work just fine,
it's just that the DEC framebuffers are nontrivial to develop a
display server for given the available documentation.
--scott

Someone added basic framebuffer support to simh, and it does mostly
work ok with VAX/VMS. It's been a while but I do remember being able
to get a DECwindows display running under simh (although there might
have been some mouse issues (I can't remember for sure)).

Of course, as of this year, that's now obsolete... :-)

Simon.

[Scott Dorsey]

Post by Simon Clubley

Post by Scott Dorsey

Sadly, language has little to do with it. Too much of the
DEC stuff is still proprietary and trade secret locked. Last
I heard that was the reason why neither NetBSD or OpenBSD ever
had a functioning X11.

Ever had a functioning X11 _server_. The clients work just fine,
it's just that the DEC framebuffers are nontrivial to develop a
display server for given the available documentation.

Someone added basic framebuffer support to simh, and it does mostly
work ok with VAX/VMS. It's been a while but I do remember being able
to get a DECwindows display running under simh (although there might
have been some mouse issues (I can't remember for sure)).

Yes. But it's harder the other way around.
--scott

[Simon Clubley]

Post by Scott Dorsey

Post by Simon Clubley

Post by Scott Dorsey

Sadly, language has little to do with it. Too much of the
DEC stuff is still proprietary and trade secret locked. Last
I heard that was the reason why neither NetBSD or OpenBSD ever
had a functioning X11.

Ever had a functioning X11 _server_. The clients work just fine,
it's just that the DEC framebuffers are nontrivial to develop a
display server for given the available documentation.

Someone added basic framebuffer support to simh, and it does mostly
work ok with VAX/VMS. It's been a while but I do remember being able
to get a DECwindows display running under simh (although there might
have been some mouse issues (I can't remember for sure)).

Yes. But it's harder the other way around.

Are you sure ?

The people who emulated the framebuffer hardware in software in simh
had to understand the framebuffer hardware well enough for the VAX/VMS
framebuffer device driver to recognise and use the emulated framebuffer.

If the public framebuffer documentation is really that limited, that
sounds like an impressive achievement to be able to emulate the
VAXstation framebuffer hardware in software well enough for VMS to
be able to start using it.

Simon.

[Bob Eager]

Sadly, language has little to do with it. Too much of the DEC stuff is
still proprietary and trade secret locked. Last I heard that was the
reason why neither NetBSD or OpenBSD ever had a functioning X11.

I find that hard to believe. I am using X11 on FreeBSD right now, and the
projects overlap quite a bit.

[Bob Eager]

Post by Bob Eager

Sadly, language has little to do with it. Too much of the DEC stuff
is still proprietary and trade secret locked. Last I heard that was
the reason why neither NetBSD or OpenBSD ever had a functioning X11.

I find that hard to believe. I am using X11 on FreeBSD right now, and
the projects overlap quite a bit.

Ah, I presume you mean VAX!

Will take a look...

[Bill Gunshannon]

Post by Bob Eager

Sadly, language has little to do with it. Too much of the DEC stuff is
still proprietary and trade secret locked. Last I heard that was the
reason why neither NetBSD or OpenBSD ever had a functioning X11.

I find that hard to believe. I am using X11 on FreeBSD right now, and the
projects overlap quite a bit.

FreeeBSD does not run on a VAX so that really has nothing to do with
the discussion. It is the VAX video hardware that is not supported
under NetBSD and, especially at this point, probably never will be.

bill

[chris]

Post by Bob Eager

Sadly, language has little to do with it. Too much of the DEC stuff is
still proprietary and trade secret locked. Last I heard that was the
reason why neither NetBSD or OpenBSD ever had a functioning X11.

I find that hard to believe. I am using X11 on FreeBSD right now, and the
projects overlap quite a bit.

FreeeBSD does not run on a VAX so that really has nothing to do with
the discussion. It is the VAX video hardware that is not supported
under NetBSD and, especially at this point, probably never will be.
bill

If you look at some early X11 sources, there is code to drive dec
frame buffer hardware, so there is info out there and probably more
if you have access to VMS or Ultrix 32 sources. Not impossible to
reverse engineer from that.

If you have Ultrix32 and VMS sources and you work from them you
are not reverse-engineering and anything you use those sources
for would still be encumbered,

Not suggesting that you should copy the code, but rather as an aid to
understanding how the hardware works. Still encumbered perhaps, but
it's now so old I doubt if anyone would care if someone wrote new
drivers for the hardware. Istr, there's loads of info in the public
domain on VCB... series qbus frame buffers out there, at least for
the mono versions, if not the GPX series boards. All i'm saying is
that you need to think creatively if you want to keep obsolete
historical hardware working.

Some of the very earliest work on X11 was done on a VS100 (aka, Vax 725)
with E&S frame buffer...

It has nothing to do with the availability of X-11 sources. It is the
information regarding the DEC hardware that is not publicly available.

Well it does in fact, since some of the X11 sources do include frame
buffer drivers for DEC hardware. X11 is open source so it would be
fine to use any info in that to modify or write fresh code.

I am sure some people have fudged around and made some of it work. But
until it is free (which will never happen) you will never see a any
serious development of X-11 for the remaining legal free VAX OSes.
Too bad they never did a port of Plan9 for the VAX. :-)

Would have been interesting to have a play with that. Worked for AT&T
wireless many years ago and it was big news at the time...

bill

[bill]

Too bad they never did a port of Plan9 for the VAX.  :-)

Old subject, but I was doing some research and found some
interesting stuff.

Seems there actually was a port of Plan9 for the the VAX. :-)
Sadly, all references today say the sources have all been lost.

But, continuing this thread, there also was a port of Amoeba
for the VAX. I haven't looked yet but although they would be
very old I think sources may still exist.

How's this for a concept. Revive VAX Amoeba. Then build an
Amoeba Distributed System with all the remaining Hobbyist
VAXen. :-)

bill

[Dennis Boone]

Post by bill
How's this for a concept. Revive VAX Amoeba. Then build an
Amoeba Distributed System with all the remaining Hobbyist
VAXen. :-)

Bonus points for using a VAXstation 100 to access the Amoeba network. :)

De

[Arne Vajhøj]

Post by Dennis Boone

Post by bill
How's this for a concept. Revive VAX Amoeba. Then build an
Amoeba Distributed System with all the remaining Hobbyist
VAXen. :-)

Bonus points for using a VAXstation 100 to access the Amoeba network. :)

I had never heard about VAXstation 100 before.

Per Wikipedia it is an X terminal using a M68K processor
not a VAX processor that connects to a VAX using Unibus.

Arne

[Dennis Boone]

Post by Arne Vajhøj
Per Wikipedia it is an X terminal using a M68K processor
not a VAX processor that connects to a VAX using Unibus.

Exactly. And Lars has been working on an emulator.

De

[Dan Cross]

Post by bill

Too bad they never did a port of Plan9 for the VAX.  :-)

Old subject, but I was doing some research and found some
interesting stuff.
Seems there actually was a port of Plan9 for the the VAX. :-)
Sadly, all references today say the sources have all been lost.

Yes. VAX was the first machine that plan9 ran on.
However, it was a short-lived portf or a very early
version of the system, and they did not stick with
it: even then, they realized that the future lay
elsewhere.

Post by bill
But, continuing this thread, there also was a port of Amoeba
for the VAX. I haven't looked yet but although they would be
very old I think sources may still exist.
How's this for a concept. Revive VAX Amoeba. Then build an
Amoeba Distributed System with all the remaining Hobbyist
VAXen. :-)

Phffft....

- Dan C.

[Simon Clubley]

Post by Johnny Billquist
That is definitely true. It's not that pleasant to run current NetBSD on
any VAX at the moment. But there are also some problems/issues that
we're waiting for a new version of gcc to come along, where they have
been fixed.

Generating the ssh server keys in NetBSD running under simh was
certainly "interesting". :-) I left it running overnight and it
was done in the morning so I don't know how long that actually took. :-)

Regarding gcc, I notice there were discussions about removing VAX
support from gcc and the subsequent bounty to fix the immediate
issues to stop that from happening.

Are those the issues you were talking about ?

Simon.

[Simon Clubley]

Post by Johnny Billquist
I think it booted to single user mode. Not sure it got much further. Not
sure anyone would even want to try and pick that up, as Linux is
constantly changing under the hood, making it a big task to resume
something abandoned years ago.

Remember that Simon believes using C makes porting easy.
Great opportunity to demonstrate.
:-)

Hey!!! :-)

Simon said "easier" not "easy" and how much easier depends on how
well the OS has been designed internally for portability.

Simon.

[Simon Clubley]

Xinu was an educational OS written by Doug Comer at Purdue to teach
os classes. For the original system, it was assumed that a Vax
running BSD 4.3 or similar, typically a 750 class machine, did the
development, and a pdp11 target machine to run the developed code.
It was quite complete for it's time and was interested enough to
buy the book and later x86 and 68000 versions that ran on pc or sun 3
hardware. Not really a mainstream OS, whatever Wiki says...

There are a number of academic created operating systems that are
interesting in their own right. None run on VAX however. :-)

Here's one I look at every so often:

http://www.helenos.org/

Simon.

[chris]

Post by Simon Clubley

Xinu was an educational OS written by Doug Comer at Purdue to teach
os classes. For the original system, it was assumed that a Vax
running BSD 4.3 or similar, typically a 750 class machine, did the
development, and a pdp11 target machine to run the developed code.
It was quite complete for it's time and was interested enough to
buy the book and later x86 and 68000 versions that ran on pc or sun 3
hardware. Not really a mainstream OS, whatever Wiki says...

There are a number of academic created operating systems that are
interesting in their own right. None run on VAX however. :-)
http://www.helenos.org/
Simon.

A bit late getting back to this, but that looks quite interesting.
Message passing design has been my favourite ipc method for some
time now. Have written a simple round robin, run to completion
task scheduler where each task is a state machine that waits for
the next message, doing nothing otherwise. Specifically for low
resource embedded work, completely in C, with no asm low level
context switching, so completely portable. A sort of proof of
concept idea and have used it for one of my own projects.

I try to keep up with the mainstream OS development, releases
and typically evaluate several over thw course of each year. Recent
examples include Open Solaris Hipster, (very well sorted) and
Devuan Linux, sans systemd. There is also SmartOs from Joyent,
a hypervisor that combines features of Open Solaris with Linux
KVM to provide a virtualisation environment, with templates for
each OS hosted. Did a quick install recently, took less than half
an hour and will revisit that later in the year. FreeBSD is still
the goto OS for most work here at present though...

Chris

[Don North]

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

I have NetBSD still running OK on my other VAX VLC box. It is an old version,
but works just fine. My OpenVMS 7.4 VAX VLC is now brain dead.

NetBSD 5.1 (GENERIC) #0: Sat Nov 6 19:48:36 UTC 2010

Welcome to NetBSD!

Sun Jan 2 10:47:37 PST 2022
donorth(1) uname -a
NetBSD netbsdvax.home.lan 5.1 NetBSD 5.1 (GENERIC) #0: Sat Nov 6 19:48:36 UTC
2010
***@b8.netbsd.org:/home/builds/ab/netbsd-5-1-RELEASE/vax/201011061943Z-obj/home/builds/ab/netbsd-5-1-RELEASE/src/sys/arch/vax/compile/GENERIC
vax
donorth(2)

[Arne Vajhøj]

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

Some old unix distributions, Ultrix32,

Ultrix32 is not legal for use without a license which you can not get.

Yes. Commercial licensed product.

                                       for example. 4.3 BSD works
on Vaxen up to MicroVAX II and possibly later. Having installed
4.3 bsd on both a Vax 730 and 725 many years ago, can confirm
that it is stable, works with most interfaces including mscp and
tmscp disk and tape controllers and good for serious work.

Why run 4.3 BSD when NetBSD is the continued development from there.

Sometimes an OS from time T runs better on HW or simulated HW
from time T than OS from time T + 20 years.

Often newer OS versions require more resources (CPU, memory, disk).

I have no idea how slim NetBSD can be.

                                                           Iirc,
I installed from 9 track magtape, but can be done via TK50 if you
can find a way to write the tape. I used an Emulex QD33 and UD33
smd disk controllers and disks for the work here.
All that stuff is available online for free now and probably far
more robust than NetBSD Vax, for example...

I really don't see how 4.3 can be more robust than the versions
developed from it.  :-)

Newer versions are not always more stable than the older versions.
Because new features can have problems.

I doubt it is the case here. The 3 modern *BSD's have a pretty
good reputation for stability.

Arne

[chris]

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

Some old unix distributions, Ultrix32,

Ultrix32 is not legal for use without a license which you can not get.

Well, it's available for free download at the unix historical society,
for years, so perhaps they organised some agreement to cover that ?.
Never run it here, but have run Ultrix 11, for evaluation. Does work
(on an 11/53), but not very stable.

for example. 4.3 BSD works
on Vaxen up to MicroVAX II and possibly later. Having installed
4.3 bsd on both a Vax 730 and 725 many years ago, can confirm
that it is stable, works with most interfaces including mscp and
tmscp disk and tape controllers and good for serious work.

Why run 4.3 BSD when NetBSD is the continued development from there.

Iirc,
I installed from 9 track magtape, but can be done via TK50 if you
can find a way to write the tape. I used an Emulex QD33 and UD33
smd disk controllers and disks for the work here.
All that stuff is available online for free now and probably far
more robust than NetBSD Vax, for example...

I really don't see how 4.3 can be more robust than the versions
developed from it. :-)

Perhaps it has improved, but last version run here was 1.01 or
similar and at that stage neither of the qbus network
interfaces were stable for long before falling over. Sort of
worked, but not well. Firmware bug on the network cards, from
memory. Perhaps I should give it another try...

Chris

bill

[Bill Gunshannon]

Post by chris

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

Some old unix distributions, Ultrix32,

Ultrix32 is not legal for use without a license which you can not get.

Well, it's available for free download at the unix historical society,
for years, so perhaps they organised some agreement to cover that ?.

Lots of commercial software is available for free on the Internet.
But having the software alone (I still have copies from when I could
actually run it legally) doesn't make it work. Things like users
require a license. :-)

Post by chris
Never run it here, but have run Ultrix 11, for evaluation. Does work
(on an 11/53), but not very stable.

Nothing comp[arable about them other than the name. Ultrix-11 is
V7 Unix and Ultrix-32 much later and never released for um-licened
use by AT&T.

Post by chris

for example. 4.3 BSD works
on Vaxen up to MicroVAX II and possibly later. Having installed
4.3 bsd on both a Vax 730 and 725 many years ago, can confirm
that it is stable, works with most interfaces including mscp and
tmscp disk and tape controllers and good for serious work.

Why run 4.3 BSD when NetBSD is the continued development from there.

Iirc,
I installed from 9 track magtape, but can be done via TK50 if you
can find a way to write the tape. I used an Emulex QD33 and UD33
smd disk controllers and disks for the work here.
All that stuff is available online for free now and probably far
more robust than NetBSD Vax, for example...

I really don't see how 4.3 can be more robust than the versions
developed from it. :-)

Perhaps it has improved, but last version run here was 1.01 or
similar and at that stage neither of the qbus network
interfaces were stable for long before falling over. Sort of
worked, but not well. Firmware bug on the network cards, from
memory. Perhaps I should give it another try...

Or compare the sources and see who broke what. :-)

bill

[Johnny Billquist]

Post by chris

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

Some old unix distributions, Ultrix32,

Ultrix32 is not legal for use without a license which you can not get.

Well, it's available for free download at the unix historical society,
for years, so perhaps they organised some agreement to cover that ?.
Never run it here, but have run Ultrix 11, for evaluation. Does work
(on an 11/53), but not very stable.

For a PDP-11 I would definitely point at 2.11BSD any day of the week.
It's really nice and good, and actually pretty useful even for "real" work.

Ultrix-11 or some older BSD would really only be for curiosity.

Post by chris

Iirc,
I installed from 9 track magtape, but can be done via TK50 if you
can find a way to write the tape. I used an Emulex QD33 and UD33
smd disk controllers and disks for the work here.
All that stuff is available online for free now and probably far
more robust than NetBSD Vax, for example...

I really don't see how 4.3 can be more robust than the versions
developed from it. :-)

Perhaps it has improved, but last version run here was 1.01 or
similar and at that stage neither of the qbus network
interfaces were stable for long before falling over. Sort of
worked, but not well. Firmware bug on the network cards, from
memory. Perhaps I should give it another try...

Around NetBSD 5 or so, it was maybe at its peak. After that it have
become really problematic.

(NetBSD never had a 1.01 version number, but back in the days when it
was still at 1.something, Qbus might have not been working that well.
For now, if someone wants NetBSD on a VAX, I would probably look at
version 4 or 5. But I do have some kind of hope that things will soon
improve.)

Johnny

[Bob Eager]

Post by Johnny Billquist
For a PDP-11 I would definitely point at 2.11BSD any day of the week.
It's really nice and good, and actually pretty useful even for "real" work.

If the machine is use doesn't have separate I and D space, it's a bit of
a tight fit. I remember the struggles we had.

In that case, there's always Sixth Edition.

And if it's a PDP-11 with no memory management, Mini-UNIX (a little known
thing, but a genuine Bell Labs one):

http://www.tavi.co.uk/unixhistory/mini-unix.html

[David Wade]

Post by Bob Eager

Post by Johnny Billquist
For a PDP-11 I would definitely point at 2.11BSD any day of the week.
It's really nice and good, and actually pretty useful even for "real" work.

If the machine is use doesn't have separate I and D space, it's a bit of
a tight fit. I remember the struggles we had.

Today I'm not even sure you can get a non-I/D space machine running 2.11.

Post by Bob Eager
In that case, there's always Sixth Edition.

Or Ultrix 11, as Bill suggested.
But then we're back to much less useful systems again.

Post by Bob Eager
And if it's a PDP-11 with no memory management, Mini-UNIX (a little known
  http://www.tavi.co.uk/unixhistory/mini-unix.html

Cool. Thanks for the link.
  Johnny

Whilst all these suggestions may work, I can't really see any point in
running anything other than VMS on my VAXs. That is what they were
bought for, that is what I have always done.

Given the lack of a legal OS I am quite inclined to send them to the
re-cyclers to be crushed and scrapped.

This would appear to be my only legal way to dispose of them

Thank you HP for "letting me go" and then destroying one of my few
remaining pleasures...

Dave
G4UGM

[Phillip Helbig (undress to reply)]

Post by David Wade
Whilst all these suggestions may work, I can't really see any point in
running anything other than VMS on my VAXs. That is what they were
bought for, that is what I have always done.
Given the lack of a legal OS I am quite inclined to send them to the
re-cyclers to be crushed and scrapped.
This would appear to be my only legal way to dispose of them

Why?

[Bill Gunshannon]

Post by David Wade

Post by Bob Eager

Post by Johnny Billquist
For a PDP-11 I would definitely point at 2.11BSD any day of the week.
It's really nice and good, and actually pretty useful even for "real" work.

If the machine is use doesn't have separate I and D space, it's a bit of
a tight fit. I remember the struggles we had.

Today I'm not even sure you can get a non-I/D space machine running 2.11.

Post by Bob Eager
In that case, there's always Sixth Edition.

Or Ultrix 11, as Bill suggested.
But then we're back to much less useful systems again.

Post by Bob Eager
And if it's a PDP-11 with no memory management, Mini-UNIX (a little known
  http://www.tavi.co.uk/unixhistory/mini-unix.html

Cool. Thanks for the link.
   Johnny

Whilst all these suggestions may work, I can't really see any point in
running anything other than VMS on my VAXs. That is what they were
bought for, that is what I have always done.
Given the lack of a legal OS I am quite inclined to send them to the
re-cyclers to be crushed and scrapped.
This would appear to be my only legal way to dispose of them

Why would you say that? As has been pointed out here there is
at least one legitimate option. If you don't want to run anything
but VMS find someone who will take them to run Unix. No reason
to destroy them other than spite.

Post by David Wade
Thank you HP for "letting me go" and then destroying one of my few
remaining pleasures...

HP did nothing that hasn't been the industry standard for as long
as I have been in this business.

bill

[Simon Clubley]

Post by David Wade
Whilst all these suggestions may work, I can't really see any point in
running anything other than VMS on my VAXs. That is what they were
bought for, that is what I have always done.

I agree. I'm on the verge of just deleting my simh instances and moving on.
I was hoping there was some unique OS (just like the HelenOS example I
posted about earlier) that actually ran on VAX but nothing has emerged.

Running older generic versions of Unix isn't something that appeals to
me because the OS is just a tool to me unless it's got something unique
about it (like VMS) that deserves to be explored in its own right.

Simon.

[Grant Taylor]

Post by David Wade
Given the lack of a legal OS I am quite inclined to send them to the
re-cyclers to be crushed and scrapped.

Why do that?

I guarantee that there are people who would be interested in acquiring
the hardware.

There are probably people that have perpetual license PAKs that would
happily run on them.

There are probably people that will run *BSD on them.

[Johnny Billquist]

Post by Johnny Billquist

Post by chris

Before deleting my simh instances, I thought I would look for
other operating systems which ran on VAX and try them out.
The only current one I have found is NetBSD (which certainly has
a _very_ retro 1980s Unix feel to it. :-)).
OpenBSD dropped VAX as an architecture a number of years ago.
Does anyone know of any other operating systems (including any
experimental ones) for VAX that are still legal to run under simh ?
If so, do you have any links to them ?
Thanks,
Simon.

Some old unix distributions, Ultrix32,

Ultrix32 is not legal for use without a license which you can not get.

Well, it's available for free download at the unix historical society,
for years, so perhaps they organised some agreement to cover that ?.
Never run it here, but have run Ultrix 11, for evaluation. Does work
(on an 11/53), but not very stable.

For a PDP-11 I would definitely point at 2.11BSD any day of the week.
It's really nice and good, and actually pretty useful even for "real" work.
Ultrix-11 or some older BSD would really only be for curiosity.

Depends on your wishes. Ultrix-11 runs on machines that can not
run 2.22. Try putting 2.11 on an RL02. :-)

True. If you can't run 2.11, then obviously you'll have to try alternative.

Post by Johnny Billquist

Post by chris

Iirc,
I installed from 9 track magtape, but can be done via TK50 if you
can find a way to write the tape. I used an Emulex QD33 and UD33
smd disk controllers and disks for the work here.
All that stuff is available online for free now and probably far
more robust than NetBSD Vax, for example...

I really don't see how 4.3 can be more robust than the versions
developed from it. :-)

Perhaps it has improved, but last version run here was 1.01 or
similar and at that stage neither of the qbus network
interfaces were stable for long before falling over. Sort of
worked, but not well. Firmware bug on the network cards, from
memory. Perhaps I should give it another try...

Around NetBSD 5 or so, it was maybe at its peak. After that it have
become really problematic.
(NetBSD never had a 1.01 version number, but back in the days when it
was still at 1.something, Qbus might have not been working that well.
For now, if someone wants NetBSD on a VAX, I would probably look at
version 4 or 5. But I do have some kind of hope that things will soon
improve.)

For the VAX, I doubt it. we will be stuck with what we have. Unless
someone wants to try porting RSX-180. :-)

Like I said - a much improved version of gcc for VAX is in the queue,
which will seriously improve on the current situation.
Exactly how much remains to be seen, though. I'm hoping that it will
once more be possible to do native builds. At which point I might also
look at some other things I've been putting off for a long time.

Johnny

[Simon Clubley]

NetBSD/vax have been suffering from a gcc that generates bad code for
some situations for quite a while, making the who system somewhat
unstable and totally unable to rebuild itself natively for about 10
years at least.

Do you happen to know what the OpenBSD people used for a compiler
when OpenBSD still supported the VAX architecture ?

Did they have some custom gcc patches or was it an earlier version
of gcc they used to build the OpenBSD VAX installation kit ?

Simon.

[Johnny Billquist]

Post by Simon Clubley

NetBSD/vax have been suffering from a gcc that generates bad code for
some situations for quite a while, making the who system somewhat
unstable and totally unable to rebuild itself natively for about 10
years at least.

Do you happen to know what the OpenBSD people used for a compiler
when OpenBSD still supported the VAX architecture ?
Did they have some custom gcc patches or was it an earlier version
of gcc they used to build the OpenBSD VAX installation kit ?

Earlier version. NetBSD used to be better as well, but later versions of
gcc introduced new problems on the VAX side. Which haven't been
addressed until recently.

Also, not sure if OpenBSD ever got around to dynamic linking and other
stuff on VAX. NetBSD did get there, but I'm not sure if anyone else
picked that up.

Johnny