From dave at fgh.geac.com.au  Mon Feb  1 21:21:28 1999
From: dave at fgh.geac.com.au (Dave Horsfall)
Date: Mon, 1 Feb 1999 22:21:28 +1100 (EST)
Subject: Old UNIX file system formats
In-Reply-To: <199902010020.TAA29097@math.uwaterloo.ca>
Message-ID: <Pine.GSO.4.03.9902012214270.10349-100000@fgh>

On Sun, 31 Jan 1999, Ken Wellsch wrote:

> I didn't see mention of the flag "HUGE" WRT the V6 file format.
> Now I may be being mislead from my memory of Venix 1.x which is
> a derivative of V6 (while Venix 2.x is SysIII I think).  If the
> HUGE bit is set in the i-node, then and only then is the 8th
> index pointer treated as the indirection variety.  Thus 8 block
> or less files I think are directly indexed.  -- Ken

I think you're confusing LARGE with HUGE.  I don't have my Edition 5
manual handy, but in my Edition 6 manual if the LARGE bit is set then
the first seven inode pointers are indirect blocks (otherwise all
eight blocks are direct), and the eighth block is a double-indirect block.

-- 
Dave Horsfall VK2KFU  dave at geac.com.au  Ph: +61 2 9978-7493 Fx: +61 2 9978-7422
Geac Computers P/L (FGH Division) 2/57 Christie St, St Leonards 2065, Australia


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id EAA29200
	for pups-liszt; Tue, 2 Feb 1999 04:57:31 +1100 (EST)

From erin at coffee.corliss.net  Tue Feb  2 04:01:14 1999
From: erin at coffee.corliss.net (Erin W. Corliss)
Date: Mon, 1 Feb 1999 10:01:14 -0800 (PST)
Subject: Old UNIX file system formats
In-Reply-To: <199902010020.TAA29097@math.uwaterloo.ca>
Message-ID: <Pine.LNX.3.96.990201095832.20104A-100000@coffee.corliss.net>

> I didn't see mention of the flag "HUGE" WRT the V6 file format.
> Now I may be being mislead from my memory of Venix 1.x which is
> a derivative of V6 (while Venix 2.x is SysIII I think).  If the
> HUGE bit is set in the i-node, then and only then is the 8th
> index pointer treated as the indirection variety.  Thus 8 block
> or less files I think are directly indexed.  -- Ken

Hmm...  I wrote a disk image editor in Visual Basic without knowing the
specs for the filesystem -- I set it up so that if the 9th pointer is zero
and the filesize is greater than one block, then it assumed the block
pointed to by the 8th pointer was a list of blocks in the file.



From mxs46 at k2.scl.cwru.edu  Tue Feb  2 15:22:10 1999
From: mxs46 at k2.scl.cwru.edu (Michael Sokolov)
Date: Tue, 2 Feb 1999 00:22:10 -0500
Subject: Hardware free to a good home
Message-ID: <199902020522.AAA07686@skybridge.scl.cwru.edu>

Dear Ladies and Gentlemen,

I have got some hardware I have to get rid of by the end of February, and it's
free to any of you guys if you are willing to come and pick it up in Cleveland,
Ohio, USA.

Last November I received a load of equipment from one company here in
Cleveland. It was a complicated network of CPUs and peripherals of all makes
and models put together by Xerox and intended to be used as a dedicated
document processing system. The CPUs included one VAX, three unidentifiable
towers, and a bunch of PCs. I'm using the VAX and all disk and tape drives
myself for my own purposes, and I'm selling the PCs, but still I've got those
three unidentifiable towers and three very funky monitors that were attached to
them. There is also a very funky laser printer attached to one of them. Given
that the VAX and all disk and tape drives have been taken out of the equation,
it's unlikely that the rest of the stuff can still be used for that dedicated
document processing whatever thing, but the towers have some apparently generic
controller boards in them (VME or something like that) and other parts that can
be raided for. Who knows, maybe even the CPUs are standard (probably some 68K),
in which case someone who knows more about this than I do (NULL) may be able to
actually use these machines for something.

The only identification on this equipment are the Xerox model numbers. One of
the towers was called NS8090 File Server. It had an external SCSI hard disk and
an Exabyte tape drive, but I've reused these for my own purposes. The other two
towers were called 6085 workstations, and they were diskless from the beginning
(as far as I can tell they don't have any mass storage controllers). All three
have monitors with very funny connectors. Aside from the Xerox model numbers
which tell me absolutely nothing, there are no hints whatsoever as to what the
CPU architecture is and all that. All towers have AUI Ethernet ports.

The laser printer is called NS8000 Laser CP, and it was attached to the tower
that was called the NS8090 File Server. The connectors are 25-pin like the
serial and parallel ones, but they have slide locks like on AUI. These slide
locks and the fact that the printer was apparently never intended to be
connected to anything except an "NS8090 File Server" suggest that the printer's
interface is not parallel or serial, but something very funny.

It has been suggested to me that I take the boxes apart, ID as many boards as
possible, and try to sell/donate them to whoever finds them useful (and the
cabinets and such would probably have to be scrapped). However, the thing is,
I don't really have time for all this, and it's naive to think that any of this
stuff has any significant cash value.

Right now I'm in the process of moving to another (cheaper) apartment in
another part of Cleveland, and really don't feel like hauling that junk around
with me. I have got these three CPU towers, three monitors, and one laser
printer, all absolutely unidentifiable, that I have to get rid of. Given what a
great job I've done at identifying and describing this stuff, it would be naive
for me to expect to sell it. Therefore, I'm giving it away for free to anyone
who is willing to come and pick it up. I have to vacate this apartment by the
end of February, and if no one picks this stuff up, I will have no choice but
to throw it in the big dumpster, which would be a great pity if this stuff is
actually useful for something.

Once again, I'm in Cleveland, Ohio, USA.

Michael Sokolov
TUHS 4BSD Coordinator
4.3BSD-* Maintainer
Quasijarus Project Principal Architect & Developer
Phone: 440-449-0299 or 216-217-2579
ARPA Internet SMTP mail: mxs46 at k2.scl.cwru.edu
TUHS WWW page: http://minnie.cs.adfa.edu.au/TUHS/
Quasijarus WWW page: http://minnie.cs.adfa.edu.au/Quasijarus/

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA03045
	for pups-liszt; Wed, 3 Feb 1999 01:53:24 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Wed Feb  3 00:52:35 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Tue, 2 Feb 1999 09:52:35 -0500 (EST)
Subject: Old UNIX file system formats
In-Reply-To: <Pine.LNX.3.96.990201095832.20104A-100000@coffee.corliss.net> from "Erin W. Corliss" at Feb 1, 99 10:01:14 am
Message-ID: <199902021452.JAA29320@math.uwaterloo.ca>

I shouldn't have posted without doing the proper research.  I took a
gander at PUPS/Tools/Filesys/traverse.c.gz which I'm quite sure is one
of the tools I wrote when I was finally able to figure out the contents
of that V6 tape I had (also with no docs - it was such irony to look
at the setup document on the tape *after* figuring the format out that
clearly describes the block layout 8-).  I notice traverse.c.gz does
indeed use the LARG flag, not HUGE.  Since few care, I'll not bother
extracting enough of Venix 1.x to see whether that is where I met the
HUGE flag or it is just my faulty memory...  -- Ken

| From owner-pups at minnie.cs.adfa.edu.au  Mon Feb  1 13:06:04 1999
| 
| Hmm...  I wrote a disk image editor in Visual Basic without knowing the
| specs for the filesystem -- I set it up so that if the 9th pointer is zero
| and the filesize is greater than one block, then it assumed the block
| pointed to by the 8th pointer was a list of blocks in the file.


From bdc at world.std.com  Thu Feb  4 05:11:56 1999
From: bdc at world.std.com (Brian D Chase)
Date: Wed, 3 Feb 1999 11:11:56 -0800 (PDT)
Subject: MicroVAX I console port question.
Message-ID: <Pine.SGI.3.95.990203104445.22510A-100000@world.std.com>

Off-topic, but maybe somewhat related to MicroPDP-11's... I've got a
MicroVAX I in a BA23 enclosure.  I'm presently a bit thrown by the serial
console port.  I'm used to the 9pin MicroVAX II ports.  From what I've
been told, the DB25-M connector for the console requires a special serial
cable for connecting the MicroVAX I up to a terminal.  A null modem cable
is not adequate. 

First, I just wanted to verify that this is correct.  So far I haven't
been able to access a console prompt using either a null modem cable, or a
straight through cable.  So either the system isn't working correctly, or
I need to get the cable right. Secondly, if it does require a special
cable, then what are the pinouts for that cable?

I'm guessing that aside from the processor itself, the MicroVAX I is
probably a lot closer in design to its contemporary MicroPDP-11 systems.
So I'm hoping that the 25pin console port is simillar to what some of you
have worked with on your Q-bus PDP-11's. 

-brian.
---
Brian "JARAI" Chase | http://world.std.com/~bdc/ | VAXZilla LIVES!!!




From cdl at mpl.ucsd.edu  Thu Feb  4 06:37:53 1999
From: cdl at mpl.ucsd.edu (Carl Lowenstein)
Date: Wed, 3 Feb 1999 12:37:53 -0800 (PST)
Subject: MicroVAX I console port question.
Message-ID: <199902032037.MAA03843@mpl.ucsd.edu>

> From owner-pups at minnie.cs.adfa.edu.au Wed Feb  3 11:35 PST 1999
> Date: Wed, 3 Feb 1999 11:11:56 -0800 (PDT)
> From: Brian D Chase <bdc at world.std.com>
> To: PUPS Mailing List <pups at minnie.cs.adfa.oz.au>
> Subject: MicroVAX I console port question.
> Mime-Version: 1.0
> 
> Off-topic, but maybe somewhat related to MicroPDP-11's... I've got a
> MicroVAX I in a BA23 enclosure.  I'm presently a bit thrown by the serial
> console port.  I'm used to the 9pin MicroVAX II ports.  From what I've
> been told, the DB25-M connector for the console requires a special serial
> cable for connecting the MicroVAX I up to a terminal.  A null modem cable
> is not adequate. 

My MicroVax (I and only) handbook vintage 1984 says the cable is a "BC22D-10".

VAX Systems and Options Catalog Oct 1984 describes the BC22D-10 as
"A fully shielded null modem cable".  Two DB25F connectors, 6 wires.

The pins in use are 1,2,3,6,7,20.  I would expect that "null modem"
means (from one end to the other) connect 2-3, 3-2, 7-7, 6-20, 20-6.

The implication is that the computer might need to see DTR asserted
before it talks to the terminal.

    carl

        carl lowenstein         marine physical lab     u.c. san diego
        {decvax|ucbvax} !ucsd!mpl!cdl                 cdl at mpl.ucsd.edu
                                                  clowenstein at ucsd.edu


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id IAA08853
	for pups-liszt; Thu, 4 Feb 1999 08:25:02 +1100 (EST)

From bdc at world.std.com  Thu Feb  4 07:24:36 1999
From: bdc at world.std.com (Brian D Chase)
Date: Wed, 3 Feb 1999 13:24:36 -0800 (PDT)
Subject: MicroVAX I console port question.
In-Reply-To: <199902032037.MAA03843@mpl.ucsd.edu>
Message-ID: <Pine.SGI.3.95.990203132057.7730C-100000@world.std.com>

On Wed, 3 Feb 1999, Carl Lowenstein wrote:

> My MicroVax (I and only) handbook vintage 1984 says the cable is a "BC22D-10".
> 
> VAX Systems and Options Catalog Oct 1984 describes the BC22D-10 as
> "A fully shielded null modem cable".  Two DB25F connectors, 6 wires.
> 
> The pins in use are 1,2,3,6,7,20.  I would expect that "null modem"
> means (from one end to the other) connect 2-3, 3-2, 7-7, 6-20, 20-6.
> 
> The implication is that the computer might need to see DTR asserted
> before it talks to the terminal.

Or given that everything seems to be fine on my end null modem cable-wise,
it's possible that something more serious is wrong with my MicroVAX I.
Does your handbook list what a flashing "1" LED error code means?  
I'll double-check my cabling as well.

-brian.
---
Brian "JARAI" Chase | http://world.std.com/~bdc/ | VAXZilla LIVES!!!


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id GAA13300
	for pups-liszt; Fri, 5 Feb 1999 06:08:49 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb  5 05:07:56 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Thu, 4 Feb 1999 20:07:56 +0100 (MET)
Subject: Memory Management
In-Reply-To: <Pine.LNX.3.96.990121163113.412A-100000@coffee.corliss.net>
Message-ID: <Pine.VUL.3.93.990204192532.4375B-100000@Zeke.Update.UU.SE>

On Thu, 21 Jan 1999, Erin W. Corliss wrote:

> The documentation that Warren gave me describes the memory management
> scheme.  It says that when the machine is first started, the memory
> management unit is disabled -- anyone know how to enable it, and where the
> segmentation registers are (I'm assuming they are in the 0160000-0177777
> range somewhere)?

I haven't seen anyone answering this, so here I go...

Reg.	Addr.

MMR0	777572
MMR1	777574
MMR2	777576
MMR3	772516
UIPAR	777640-777656
UDPAR	777660-777676
UIPDR	777600-777616
UDPDR	777620-777636
SIPAR	772240-772256
SDPAR	772260-772276
SIPDR	772200-772216
SDPDR	772220-772236
KIPAR	772340-772356
KDPAR	772360-772376
KIPDR	772300-772316
KDPDR	772320-772336

xy in xyP?R is:

x:	U - User
	S - Supervisor
	K - Kernel
y:	I - Instruction
	D - Data

PAR is Page Address Register
PDR is Page Description Register

Okay, so for the layout of the registers...

MMR0:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 ! ! ! !     ! ! ! \-/ ! \---/ +-- Enable relocation
 ! ! ! !     ! ! !  !  !   +------ Page number
 ! ! ! !     ! ! !  !  +---------- Page address space I/D
 ! ! ! !     ! ! !  +------------- Page mode
 ! ! ! !     ! ! +---------------- Instruction completed
 ! ! ! !     ! +------------------ Maintenance mode
 ! ! ! !     +-------------------- Enable memory management trap
 ! ! ! +-------------------------- Trap-Memory management
 ! ! +---------------------------- Abort-Read only access violation
 ! +------------------------------ Abort-Page length error
 +-------------------------------- Abort-Non resident

The page info is for when a trap/fault occurs, and tells in which page it
occured.The rest should be pretty obvious.

MMR1:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 \-------/ \---/ \-------/ \---/
     !       !       !       +---- Register number
     !       !       +------------ Amount changed (2 compl.)
     !       +-------------------- Register numbe
     +---------------------------- Amount changed (2 compl.)

Low byte is written first, and this register tells how much registers have
changed part way through an instruction, which needs to be undone to start
the intruction again.

MMR2:

Virtual address of instruction where fault occured.

MMR3:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
                     ! !   ! ! +--- Enable user D space
                     ! !   ! +----- Enable supervisor D space
                     ! !   +------- Enable kernel D space
                     ! +----------- Enable 22-bit mapping
                     +------------- Enable unibus map

If 22-bit mapping isn't enabled, the machine will be in 18-bit addressign
when MMU is enabled. Unibus-mapping is something I'll skip for now. You
need it for DMA on a 22-bit unibus machine only.

Note that at the end of a MMU trap/abort, MMR0 bit 15-12 must be cleared
for MMR1 and MMR2 to become active again.


>From a virtual address (VA), you get to the physical address (PA) like
this:

APF=VA[15:13]
DF=VA[12:0]

PA=PAR(APF)*64+DF


The PDR looks loke this:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   \-----------/ ! !     ! \---/
         !       ! !     !   +----- ACF
         !       ! !     +--------- ED
         !       ! +--------------- W
         !       +----------------- A
         +------------------------- PLF

ACF - Access Control Field
	000 - Non resident; abort on all accesses
	001 - Read only; abort on write attempt, memory mgmt trap on read
	010 - Read only; abort on write attempt
	011 - Unused; abort on all accesses - reserved for future use
	100 - Read/Write; memory mgmt trap upon completion of read or write
	101 - Read/Write; memory mgmt trap upon completion of write
	110 - Read/Write; no system trap/abort action
	111 - Unused; abort on all accesses - reserved for future use

A - Access to page has been made.
W - Page has been written to since PAR/PDR was loaded
ED - Expansion direction
PLF - Page length field

Now, have fun...
	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol



From mxs46 at k2.scl.cwru.edu  Mon Feb  8 06:55:35 1999
From: mxs46 at k2.scl.cwru.edu (Michael Sokolov)
Date: Sun, 7 Feb 1999 15:55:35 -0500
Subject: Special Agent Michael Sokolov has moved
Message-ID: <199902072055.PAA09064@skybridge.scl.cwru.edu>

My new address is:

13444 Euclid Ave. Apt. 215
East Cleveland, OH  44112
USA

My new phone # is 216-761-3656 (voice mail not set up yet, will be done in a
couple of days).

I'm still not quite done with all move-related work, so it will be a few more
days before I catch up with my E-mail.

My hardware is laid out a lot better at the new place than at the old one, so
when I'm done hooking everything up, I'll have much better work conditions for
my Project. Also the new place is physically closer to the building where all
Cleveland ISPs are located, reducing the cost of leased lines and increasing
the probability of me getting one some day.

With the hardware taking up most of the space, I originally thought that my
apartment would look like Agent Mulder's, but it actually ended up being more
like Agent Scully's. Oh well, her place is pretty nice too, and so is mine now.

Just a reminder to all Quasijarus Project folks living in the USA, be sure to
watch the X-Files this evening. They'll finally tell us what really happened to
Mulder's sister, who is the cigarette-smoking man, and all the other cool
stuff.

Special Agent Michael Sokolov
TUHS 4BSD Coordinator
4.3BSD-* Maintainer
Quasijarus Project Principal Architect & Developer
Phone: 216-761-3656
ARPA Internet SMTP mail: mxs46 at k2.scl.cwru.edu
TUHS WWW page: http://minnie.cs.adfa.edu.au/TUHS/
Quasijarus WWW page: http://minnie.cs.adfa.edu.au/Quasijarus/


From entropy at zippy.bernstein.com  Wed Feb 17 14:23:11 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Tue, 16 Feb 1999 23:23:11 -0500 (EST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902170423.XAA24481@zippy.bernstein.com>

I'm trying to compile a 2.9BSD kernel using the distribution from the
pups archive.

"make unix" failed:
Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.

I looked in the usr.tar from the distribution, and I don't see mbuf.h
anywhere.

Does anyone know where I can find a copy of this file?

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id QAA11005
	for pups-liszt; Wed, 17 Feb 1999 16:17:41 +1100 (EST)

From sms at moe.2bsd.com  Wed Feb 17 15:15:02 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Tue, 16 Feb 1999 21:15:02 -0800 (PST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902170515.VAA23159@moe.2bsd.com>

Hi -

> I'm trying to compile a 2.9BSD kernel using the distribution from the
> pups archive.
> 
> "make unix" failed:
> Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> 
> I looked in the usr.tar from the distribution, and I don't see mbuf.h
> anywhere.
> 
> Does anyone know where I can find a copy of this file?

	That's not _all_ your missing ;-)

	Unless you have the 1985 Seismo (or Harvard - depends where you
	got the tape from) update tape to 2.9 the networking code won't
	compile much less run.  Been there, done that.  It was a fun couple
	weeks coming to the realization that the networking code hadn't
	been fully integrated and compiled in 2.9

	I believe the 2.9-Seismo update is in the PUPS archive (should be
	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
	painful upgrade process because it changes the a.out header format
	for overlaid processes (goes from 7 to 15 overlays).  If you're not
	real careful you'll have (as I did ;-)) a real mess:  can't finish
	the upgrade because the old kernel doesn't support the new overlaid
	processes but you can't build a new kernel because doing so needs
	those processes.  Something like that.  It was "interesting" ;)

	Steven Schultz

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id QAA11030
	for pups-liszt; Wed, 17 Feb 1999 16:24:28 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Wed Feb 17 15:26:09 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Wed, 17 Feb 1999 16:26:09 +1100 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <199902170515.VAA23159@moe.2bsd.com> from "Steven M. Schultz" at "Feb 16, 1999  9:15: 2 pm"
Message-ID: <199902170526.QAA14818@henry.cs.adfa.edu.au>

In article by Steven M. Schultz:
> Hi -
> 
> > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > pups archive.
> > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > Does anyone know where I can find a copy of this file?
> 
> 	That's not _all_ your missing ;-)
> 
> 	Unless you have the 1985 Seismo (or Harvard - depends where you
> 	got the tape from) update tape to 2.9 the networking code won't
> 	compile much less run.  Been there, done that.  It was a fun couple
> 	weeks coming to the realization that the networking code hadn't
> 	been fully integrated and compiled in 2.9
> 
> 	I believe the 2.9-Seismo update is in the PUPS archive (should be
> 	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> 	painful upgrade process because it changes the a.out header format
> 	for overlaid processes (goes from 7 to 15 overlays).  If you're not
> 	real careful you'll have (as I did ;-)) a real mess:  can't finish
> 	the upgrade because the old kernel doesn't support the new overlaid
> 	processes but you can't build a new kernel because doing so needs
> 	those processes.  Something like that.  It was "interesting" ;)
> 
> 	Steven Schultz

Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro. 
I'm sure he will keep us informed :-)

'Night!

	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id RAA11134
	for pups-liszt; Wed, 17 Feb 1999 17:01:58 +1100 (EST)

From djenner at halcyon.com  Wed Feb 17 16:00:45 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Tue, 16 Feb 1999 22:00:45 -0800
Subject: 2.9BSD:  mbuf.h
References: <199902170526.QAA14818@henry.cs.adfa.edu.au>
Message-ID: <36CA5B0D.8A2B2629@halcyon.com>

Speaking of the Pro, I have one and have been trying to get Venix
to run on it.  The rub is, there are two versions: one directly from
VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Venix/Pro is freely available on the Internet at ftp.update.uu.se,
but Pro/Venix seems to be a little harder to find.  Pro/Venix is
much to be preferred because you can reconfigure the kernel (in
binary) to include different drivers, etc.

I've been able to acquire all the documentation and all (almost) the
disks for Pro/Venix 2.0.  A couple of the disks are apparently
unusable or missing in the set I have.

It seems to me that Pro/Venix is a potential candidate for the PUPS
archive, the snag being DEC/Compaq residual interests in it.  PUPS
covers the AT&T part, VenturCom has "given away" their part, and
DEC/Compaq is all that's left.

So:
	1) Could this be a PUPS addition, if a good copy be found?
	2) If someone has a copy, but worries about the DEC/Compaq
	   aspects, can a good copy of the disks I have be acquired?
	   (Anyone in this category might want to respond directly
	    to me instead of posting to the mailing lists.)  After
	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
	   objections are probably to worry about the AT&T part,
	   which the Ancient Unix license covers...

Actually, I'm amazed I've gotten as far as I have with this, because
I've been pretty passive about finding it.  It's only taken 2 years
so far.

Dave

Warren Toomey wrote:
> 
> In article by Steven M. Schultz:
> > Hi -
> >
> > > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > > pups archive.
> > > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > > Does anyone know where I can find a copy of this file?
> >
> >       That's not _all_ your missing ;-)
> >
> >       Unless you have the 1985 Seismo (or Harvard - depends where you
> >       got the tape from) update tape to 2.9 the networking code won't
> >       compile much less run.  Been there, done that.  It was a fun couple
> >       weeks coming to the realization that the networking code hadn't
> >       been fully integrated and compiled in 2.9
> >
> >       I believe the 2.9-Seismo update is in the PUPS archive (should be
> >       on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> >       painful upgrade process because it changes the a.out header format
> >       for overlaid processes (goes from 7 to 15 overlays).  If you're not
> >       real careful you'll have (as I did ;-)) a real mess:  can't finish
> >       the upgrade because the old kernel doesn't support the new overlaid
> >       processes but you can't build a new kernel because doing so needs
> >       those processes.  Something like that.  It was "interesting" ;)
> >
> >       Steven Schultz
> 
> Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro.
> I'm sure he will keep us informed :-)
> 
> 'Night!
> 
>         Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id RAA11133
	for pups-liszt; Wed, 17 Feb 1999 17:01:50 +1100 (EST)

From djenner at halcyon.com  Wed Feb 17 16:00:45 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Tue, 16 Feb 1999 22:00:45 -0800
Subject: 2.9BSD:  mbuf.h
References: <199902170526.QAA14818@henry.cs.adfa.edu.au>
Message-ID: <36CA5B0D.8A2B2629@halcyon.com>

Speaking of the Pro, I have one and have been trying to get Venix
to run on it.  The rub is, there are two versions: one directly from
VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Venix/Pro is freely available on the Internet at ftp.update.uu.se,
but Pro/Venix seems to be a little harder to find.  Pro/Venix is
much to be preferred because you can reconfigure the kernel (in
binary) to include different drivers, etc.

I've been able to acquire all the documentation and all (almost) the
disks for Pro/Venix 2.0.  A couple of the disks are apparently
unusable or missing in the set I have.

It seems to me that Pro/Venix is a potential candidate for the PUPS
archive, the snag being DEC/Compaq residual interests in it.  PUPS
covers the AT&T part, VenturCom has "given away" their part, and
DEC/Compaq is all that's left.

So:
	1) Could this be a PUPS addition, if a good copy be found?
	2) If someone has a copy, but worries about the DEC/Compaq
	   aspects, can a good copy of the disks I have be acquired?
	   (Anyone in this category might want to respond directly
	    to me instead of posting to the mailing lists.)  After
	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
	   objections are probably to worry about the AT&T part,
	   which the Ancient Unix license covers...

Actually, I'm amazed I've gotten as far as I have with this, because
I've been pretty passive about finding it.  It's only taken 2 years
so far.

Dave

Warren Toomey wrote:
> 
> In article by Steven M. Schultz:
> > Hi -
> >
> > > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > > pups archive.
> > > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > > Does anyone know where I can find a copy of this file?
> >
> >       That's not _all_ your missing ;-)
> >
> >       Unless you have the 1985 Seismo (or Harvard - depends where you
> >       got the tape from) update tape to 2.9 the networking code won't
> >       compile much less run.  Been there, done that.  It was a fun couple
> >       weeks coming to the realization that the networking code hadn't
> >       been fully integrated and compiled in 2.9
> >
> >       I believe the 2.9-Seismo update is in the PUPS archive (should be
> >       on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> >       painful upgrade process because it changes the a.out header format
> >       for overlaid processes (goes from 7 to 15 overlays).  If you're not
> >       real careful you'll have (as I did ;-)) a real mess:  can't finish
> >       the upgrade because the old kernel doesn't support the new overlaid
> >       processes but you can't build a new kernel because doing so needs
> >       those processes.  Something like that.  It was "interesting" ;)
> >
> >       Steven Schultz
> 
> Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro.
> I'm sure he will keep us informed :-)
> 
> 'Night!
> 
>         Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11461
	for pups-liszt; Wed, 17 Feb 1999 19:23:34 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:22:50 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:22:50 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <36CA5B0D.8A2B2629@halcyon.com> (djenner@halcyon.com)
Message-ID: <199902170822.DAA24861@zippy.bernstein.com>

>Date: Tue, 16 Feb 1999 22:00:45 -0800
>From: "David C. Jenner" <djenner at halcyon.com>
>
>Speaking of the Pro, I have one and have been trying to get Venix
>to run on it.  The rub is, there are two versions: one directly from
>VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
for the Pro-380.  I never heard of a distinction between Venix/Pro
vs. Pro/Venix.  Then again, I got into this game fairly late...I
bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
installed (also with original install media and docs).

>Venix/Pro is freely available on the Internet at ftp.update.uu.se,
>but Pro/Venix seems to be a little harder to find.  Pro/Venix is
>much to be preferred because you can reconfigure the kernel (in
>binary) to include different drivers, etc.
>
>I've been able to acquire all the documentation and all (almost) the
>disks for Pro/Venix 2.0.  A couple of the disks are apparently
>unusable or missing in the set I have.

I have the following archives of Venix-related stuff that I snagged
from the net a few years back.  If you think any of them might contain
what you're looking for, let me know and I'll give more detail about
their contents.

-rw-r--r--   1 entropy  user         3833 Oct 17  1997 README
-rw-r--r--   1 entropy  user          532 Oct 17  1997 README.VAX
-rw-r--r--   1 entropy  user        30819 Oct 17  1997 RX50.notes
-rw-r--r--   1 entropy  user      2530759 Oct 17  1997 Venix1.tar.Z
-rw-r--r--   1 entropy  user      2503931 Oct 17  1997 Venix2.tar.Z
-rw-r--r--   1 entropy  user        15817 Oct 17  1997 cathang.txt
-rw-r--r--   1 entropy  user       332543 Oct 17  1997 mopimage.tar.Z
-rw-r--r--   1 entropy  user          443 Oct 17  1997 nbsdrx50.readme
-rw-r--r--   1 entropy  user       897510 Oct 17  1997 nbsdrx50.zip
-rw-r--r--   1 entropy  user       155648 Oct 17  1997 pppd
-rw-r--r--   1 entropy  user       193536 Oct 17  1997 pr0801eng.sys
-rw-r--r--   1 entropy  user        14153 Oct 17  1997 raind112.zip
-rw-r--r--   1 entropy  user         6621 Oct 17  1997 rx50.zip
-rw-r--r--   1 entropy  user        81152 Oct 17  1997 teledisk.zip
-rw-r--r--   1 entropy  user        61440 Oct 17  1997 venix.tar
-rw-r--r--   1 entropy  user          116 Oct 17  1997 venix1.readme
-rw-r--r--   1 entropy  user      1119490 Oct 17  1997 venix1s.zip
-rw-r--r--   1 entropy  user      1095824 Oct 17  1997 venix1u.zip
-rw-r--r--   1 entropy  user          424 Oct 17  1997 venix2.readme
-rw-r--r--   1 entropy  user      1058970 Oct 17  1997 venix2s.zip
-rw-r--r--   1 entropy  user      1145720 Oct 17  1997 venix2u.zip
-rw-r--r--   1 entropy  user       332362 Oct 17  1997 vnx2u2u5.zip

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11492
	for pups-liszt; Wed, 17 Feb 1999 19:33:00 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:32:05 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:32:05 -0500 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <199902170515.VAA23159@moe.2bsd.com> (sms@moe.2bsd.com)
Message-ID: <199902170832.DAA24878@zippy.bernstein.com>

>Date: Tue, 16 Feb 1999 21:15:02 -0800 (PST)
>From: "Steven M. Schultz" <sms at moe.2bsd.com>
>
>	I believe the 2.9-Seismo update is in the PUPS archive (should be
>	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
>	painful upgrade process because it changes the a.out header format
>	for overlaid processes (goes from 7 to 15 overlays).  If you're not
>	real careful you'll have (as I did ;-)) a real mess:  can't finish
>	the upgrade because the old kernel doesn't support the new overlaid
>	processes but you can't build a new kernel because doing so needs
>	those processes.  Something like that.  It was "interesting" ;)

Sounds like fun.  Any hints on the correct upgrade path to avoid this
lossage?

Better yet, would you be willing and able to upload a disk image or
tar file of an upgraded system to the PUPS archive (or directly to me
if it's not of general interest), so I could use that as a starting
point?

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11533
	for pups-liszt; Wed, 17 Feb 1999 19:43:19 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:42:40 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:42:40 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902170842.DAA24887@zippy.bernstein.com>


As already mentioned in previous messages, I'm working on getting
2.9BSD onto a Pro 350.  I'm using 2.9BSD as a starting point because
it claims to support machines without split i/d.  The 350 uses the
F-11 chipset, which I have read does not support split i/d.

I would prefer to use 2.11BSD because I understand it's still actively
used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
says that it needs split i/d to run.  Can anyone give me more detail
about this?  Was support for machines without split i/d removed from
the kernel, or is it just that some of the programs are too big to fit
in a single 64k segment?

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA12587
	for pups-liszt; Thu, 18 Feb 1999 01:36:09 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 00:35:30 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902170822.DAA24861@zippy.bernstein.com> from "maximum entropy" at Feb 17, 99 03:22:50 am
Message-ID: <199902171435.JAA12462@math.uwaterloo.ca>

| Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
| they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
| for the Pro-380.  I never heard of a distinction between Venix/Pro
| vs. Pro/Venix.  Then again, I got into this game fairly late...I
| bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
| installed (also with original install media and docs).

My time playing with Pro's faded out before Venix 2 was available (free)
for me to try.  I've played a fair bit with Venix 1.1 on both Pro 350's
and Pro 380's.  The Venix 1 series I feel is basically V6 derived while
I understood the Venix 2 series was derived from Sys III.

About a year ago Rick Macklem that did a port to the Pro series mailed
me his "Pro stuff" which included a tape and floppies.  I've forgotten
what all is in that stash, but taking a peek at some old mail he mentions:

> The stuff I did went out on a Usenix distribution tape in about 1983/84
> and had to be merged into a 2.9BSD distribution. I did generate floppy
> sets for a few people, because that was the only easy way to get it
> installed. (The first install here was actually done by downloading the
> kernel over the serial port talking to the PDP 11 prom (ODS?).)

I had thought his set of patches were in the PUPS archive.  In fact I
see the patches under PUPS/Distributions/ucb/2.9-pro350.

-- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA12654
	for pups-liszt; Thu, 18 Feb 1999 01:42:25 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 00:42:05 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 09:42:05 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
In-Reply-To: <199902170842.DAA24887@zippy.bernstein.com> from "maximum entropy" at Feb 17, 99 03:42:40 am
Message-ID: <199902171442.JAA15916@math.uwaterloo.ca>

| I would prefer to use 2.11BSD because I understand it's still actively
| used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
| says that it needs split i/d to run.  Can anyone give me more detail
| about this?  Was support for machines without split i/d removed from
| the kernel, or is it just that some of the programs are too big to fit
| in a single 64k segment?

Have you been able to acquire the documentation for the DECNA card?  I
think that is roughly what it is called.  The Pro Ethernet card.  A few
old timers like myself and Dan Lanciani talked years ago about running
things on a Pro and no-one seems to know much about this relatively
critical bit of documentation.  Again referring to Rick Macklem's
correspondence (I believe I was asking him, again, about these docs):

> Well, the short answer is "I'm not sure what the answers are". At one
> point someone mentioned they were putting the Pro stuff into 2BSD, but
> I'm not sure if they actually did it. (The guys that used it the most
> had it running on a lab of Pro380s at Columbia U. (I think. It's the
> one right in NY city.)) His name was Charlie Kim (again, I think?) and
> did some stuff to it so that it worked reasonably well on a Pro380, but
> I have no idea how you might find him now. (It was a real dog on a Pro350
> because it didn't have separate I and D space.)

The rumors we were able to find all pointed to this place and person
WRT documentation for the ethernet card.

-- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id CAA12725
	for pups-liszt; Thu, 18 Feb 1999 02:12:16 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 01:11:36 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 10:11:36 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902171435.JAA12462@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:35:30 -0500 (EST))
Message-ID: <199902171511.KAA25176@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
>
>About a year ago Rick Macklem that did a port to the Pro series mailed
>me his "Pro stuff" which included a tape and floppies.  I've forgotten
>what all is in that stash, but taking a peek at some old mail he mentions:

Would you be able to send images (rx50 teledisk, or plain dd dumps) of
these disks to me or to the archive?

>I had thought his set of patches were in the PUPS archive.  In fact I
>see the patches under PUPS/Distributions/ucb/2.9-pro350.

Those files aren't 100% complete.  Excerpt of a mail I sent last night
to Warren Toomey:

#The instructions in boot.doc are mangled.
#The patches included are reversed, and didn't apply cleanly to one of
#the files (/usr/src/net/sys/sys/machdep.c).  Also, it looks like the
#guy that produced that set of changes forgot to include his
#modifications to /usr/src/sys/conf/config, but I managed to hack
#together something that might work.

Then there's the fact that the 2.9 distribution won't even compile,
and the 2.9 upgrade patches are a nightmare...

Maybe I'll just stick to venix :-)

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id CAA12848
	for pups-liszt; Thu, 18 Feb 1999 02:44:50 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 01:44:04 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 10:44:04 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902171435.JAA12462@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:35:30 -0500 (EST))
Message-ID: <199902171544.KAA25215@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
>
>About a year ago Rick Macklem that did a port to the Pro series mailed
>me his "Pro stuff" which included a tape and floppies.  I've forgotten
>what all is in that stash, but taking a peek at some old mail he mentions:

Would you be able to send images (rx50 teledisk, or plain dd dumps) of
these disks to me or to the archive?

>I had thought his set of patches were in the PUPS archive.  In fact I
>see the patches under PUPS/Distributions/ucb/2.9-pro350.

Those files aren't 100% complete.  Excerpt of a mail I sent last night
to Warren Toomey:

#The instructions in boot.doc are mangled.d
#The patches included are reversed, and didn't apply cleanly to one of
#the files (/usr/src/net/sys/sys/machdep.c).  Also, it looks like the
#guy that produced that set of changes forgot to include his
#modifications to /usr/src/sys/conf/config, but I managed to hack
#together something that might work.

Then there's the fact that the 2.9 distribution won't even compile,
and the 2.9 upgrade patches are a nightmare...

Maybe I'll just stick to venix :-)

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12922
	for pups-liszt; Thu, 18 Feb 1999 03:12:05 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 02:11:24 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 11:11:24 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
In-Reply-To: <199902171442.JAA15916@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:42:05 -0500 (EST))
Message-ID: <199902171611.LAA25258@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:42:05 -0500 (EST)
>
>Have you been able to acquire the documentation for the DECNA card?  I

I haven't looked for it.  The DECNA is optional, and my Pro doesn't
have it.  All Pro's came with an AUI port, but without the card it
doesn't do anything.

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12934
	for pups-liszt; Thu, 18 Feb 1999 03:12:18 +1100 (EST)

From djenner at halcyon.com  Thu Feb 18 02:11:12 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Wed, 17 Feb 1999 08:11:12 -0800
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
References: <199902171435.JAA12462@math.uwaterloo.ca>
Message-ID: <36CAEA1F.D5D7C838@halcyon.com>

I haven't tried the 2.9 stuff at all on a Pro.  I have had it
running on an 11/23+ (w/binary license) for 10 years.  The
problem is the networking, as you have found.

Venix/Pro 1.1 and 2.0 run just fine on the Pro 380, and it's
pretty painless to install.  I have distribution disks for
Pro/Venix 1.1, but the install disk has apparently been
overwritten with the 2.0 installation disk.  And my distribution
for 2.0 is missing a couple of original disks; I have copies of
those disks, but they get read errors.

I guess the 2.9 stuff would be interesting if you got it to
work on the Pro, especially if you got networking to work.
I don't have any docs on the DECNA, but they must exist.  It's
probably pretty close to the DEQNA.

Dave

Ken Wellsch wrote:
> 
> | Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
> | they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
> | for the Pro-380.  I never heard of a distinction between Venix/Pro
> | vs. Pro/Venix.  Then again, I got into this game fairly late...I
> | bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
> | installed (also with original install media and docs).
> 
> My time playing with Pro's faded out before Venix 2 was available (free)
> for me to try.  I've played a fair bit with Venix 1.1 on both Pro 350's
> and Pro 380's.  The Venix 1 series I feel is basically V6 derived while
> I understood the Venix 2 series was derived from Sys III.
> 
> About a year ago Rick Macklem that did a port to the Pro series mailed
> me his "Pro stuff" which included a tape and floppies.  I've forgotten
> what all is in that stash, but taking a peek at some old mail he mentions:
> 
> > The stuff I did went out on a Usenix distribution tape in about 1983/84
> > and had to be merged into a 2.9BSD distribution. I did generate floppy
> > sets for a few people, because that was the only easy way to get it
> > installed. (The first install here was actually done by downloading the
> > kernel over the serial port talking to the PDP 11 prom (ODS?).)
> 
> I had thought his set of patches were in the PUPS archive.  In fact I
> see the patches under PUPS/Distributions/ucb/2.9-pro350.
> 
> -- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12957
	for pups-liszt; Thu, 18 Feb 1999 03:17:55 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 02:15:01 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 08:15:01 -0800 (PST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902171615.IAA02324@moe.2bsd.com>

Hi

> Sounds like fun.  Any hints on the correct upgrade path to avoid this
> lossage?

	Oh, it's not _completely_ irrecoverable and is "fun" in a perverse
	way.

	First go thru all of the executable directories (/bin, /usr/bin,...)
	and identify all of the overlaid executables and save copies of them.
	Shouldn't be too many but the important one is 'ex'/'vi'.  A number
	of programs rely on using 'ex' scripts to edit generated files (the
	kernel makefiles are _good_ examples;)), and so on.  Having an older
	copy of 'ex'/'vi' is the main thing I remember as saving the day.

> Better yet, would you be willing and able to upload a disk image or
> tar file of an upgraded system to the PUPS archive (or directly to me

	Oh, I have no 2.9 systems - this was all done 10 years ago.  The
	systems I run now use MSCP/TMSCP devices and 2.9 lacks support
	for those.

	Steve

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA13020
	for pups-liszt; Thu, 18 Feb 1999 03:32:46 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 02:32:00 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 08:32:00 -0800 (PST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902171632.IAA02404@moe.2bsd.com>

Hi -

> From: maximum entropy <entropy at zippy.bernstein.com>
> I would prefer to use 2.11BSD because I understand it's still actively
> used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
> says that it needs split i/d to run.  Can anyone give me more detail
> about this?  Was support for machines without split i/d removed from
> the kernel, or is it just that some of the programs are too big to fit
> in a single 64k segment?

	Oh, support was NOT removed.  Non-split executables (magic number
	0407 and 0410) will still run.  

	The kernel will not fit - without split I/D it is impossible to
	create a /unix image that fits within a single 64kb (actually 48kb
	since the kernel stack takes 1 segment and the 'u' area takes
	another) address space.

	I actually went thru the exercise once (2.10 era) of creating a bare 
	bones kernel that would fit in - at least the linker said it would.
	That was only done by ripping out lots of stuff - no networking, no
	statistics gathering, almost no drivers, etc.  Never 'ran' it though
	since there seemed to be little point in such a stripped down system.

	Even V7 was hard pressed to run on a non-split machine!  In fact there
	was a paper written about shoehorning V7 onto an 11/40 and the hoops
	that needed to be jumped thru.  Not sure but that document might be
	in the /usr/doc tree of one of the PUPS Distributions hierarchy.

	Steven


From wkt at henry.cs.adfa.edu.au  Thu Feb 18 08:25:47 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Thu, 18 Feb 1999 09:25:47 +1100 (EST)
Subject: Pro/Venix
In-Reply-To: <36CA5B0D.8A2B2629@halcyon.com> from "David C. Jenner" at "Feb 16, 1999 10: 0:45 pm"
Message-ID: <199902172225.JAA16961@henry.cs.adfa.edu.au>

In article by David C. Jenner:
> It seems to me that Pro/Venix is a potential candidate for the PUPS
> archive, the snag being DEC/Compaq residual interests in it.  PUPS
> covers the AT&T part, VenturCom has "given away" their part, and
> DEC/Compaq is all that's left.
> 
> So:
> 	1) Could this be a PUPS addition, if a good copy be found?
> 	2) If someone has a copy, but worries about the DEC/Compaq
> 	   aspects, can a good copy of the disks I have be acquired?
> 	   (Anyone in this category might want to respond directly
> 	    to me instead of posting to the mailing lists.)  After
> 	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
> 	   objections are probably to worry about the AT&T part,
> 	   which the Ancient Unix license covers...
> 
> Dave

If we could get DEC/Compaq to allow access to Pro/Venix by UNIX source
license holders, then yes I would certainly add it to the Archive. If
there's no source code, and SCO are happy, then it could go up for anon ftp.

Cheers,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id KAA14692
	for pups-liszt; Thu, 18 Feb 1999 10:21:58 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Thu Feb 18 09:18:40 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Thu, 18 Feb 1999 10:18:40 +1100 (EST)
Subject: Help with regs on Pro serial ports
Message-ID: <199902172318.KAA18083@henry.cs.adfa.edu.au>

I'm trying to help get the kernel for the version of 2.9BSD ported to the
Pro-350. The patches supplied by Rick Macklem are slightly incomplete, e.g
there is no config shell script which knows about the new device drivers etc.

Anyway, one vital missing file is pcreg.h, which holds the structure
describing the registers of the serial ports on the Pro-350. By perusing
the file dev/pc.c, I've worked out that the struct looks something like:


struct pcdevice {
	??? baud;
	??? cdb;
	??? csa;
	??? csb;
	??? csr;
	??? dbuf;
	??? mc0;
	??? mc1;
	??? mode;
	??? stat;
}

where the fields are not in the correct order, and I have no idea what
C type each is. If anybody can help recreate this file, could they
email me?!

I've included below the C comments at the top of dev/pc.c.

If anybody has Rick Macklem's email address, could they pass that on too?
I will email him and see if he's got a more complete set of patches somewhere.

Many thanks in advance,

	Warren

/*
 * This driver handles the two serial ports on the back of the
 * pro3xx system unit. Although not software compatible, they
 * are handled as minor device 0 & 1 respectively, for the printer
 * and communication port. Modem control is included but no sync
 * serial support for the com. port.
 * NOTE: The DSR line in the printer port is used for carrier
 * detect so terminals or modems should be cabled accordingly.
 * Local terminal cables should jumper DTR-CDT so that the carrier
 * will appear to be up or PC_SOFTCAR defined and devs or'd with 0200.
 * NOTE2: The interrupt service routines are as follows:
 *      plrint - printer port receive
 *      plxint - printer port transmit
 *      cmintr - communication port com. interrupt
 * Modem transition interrupts are NOT used.
 */

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA15025
	for pups-liszt; Thu, 18 Feb 1999 11:31:53 +1100 (EST)

From allisonp at world.std.com  Thu Feb 18 10:31:29 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Wed, 17 Feb 1999 19:31:29 -0500
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902180031.AA13175@world.std.com>


<Venix/Pro is freely available on the Internet at ftp.update.uu.se,
<but Pro/Venix seems to be a little harder to find.  Pro/Venix is
<much to be preferred because you can reconfigure the kernel (in
<binary) to include different drivers, etc.

The UU.SE and gatway.dec.com version of it I ahve running on my PRO-350
for the last year or more.

I'd like to have SLIP/PPP running on it or even be able to tweek it.

<	1) Could this be a PUPS addition, if a good copy be found?

Oh hand I'd say yes.  

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA15035
	for pups-liszt; Thu, 18 Feb 1999 11:32:11 +1100 (EST)

From allisonp at world.std.com  Thu Feb 18 10:31:41 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Wed, 17 Feb 1999 19:31:41 -0500
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902180031.AA13305@world.std.com>

<As already mentioned in previous messages, I'm working on getting
<2.9BSD onto a Pro 350.  I'm using 2.9BSD as a starting point because
<it claims to support machines without split i/d.  The 350 uses the
<F-11 chipset, which I have read does not support split i/d.

The F11 does not do I&D split but does have user/system.  

<I would prefer to use 2.11BSD because I understand it's still actively
<used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
<says that it needs split i/d to run.  Can anyone give me more detail
<about this?  Was support for machines without split i/d removed from
<the kernel, or is it just that some of the programs are too big to fit
<in a single 64k segment?

It's my understanding that 2.11 will run on F11 systems (pro350 and 11/23)
if properly configured but the only binaries loose are for split I&D.
So if properly configured you can get 2.11 to utilize the user/system
spaces.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA15404
	for pups-liszt; Thu, 18 Feb 1999 12:32:52 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 11:25:53 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 17:25:53 -0800 (PST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902180125.RAA05895@moe.2bsd.com>

Hi -

> From: allisonp at world.std.com (Allison J Parent)
> The F11 does not do I&D split but does have user/system.  

	Correct.  Some systems also have an 18bit only MMU which restricts
	memory to 248kb max (others have a 22bit MMU and can physically
	have more memory).

> It's my understanding that 2.11 will run on F11 systems (pro350 and 11/23)
> if properly configured but the only binaries loose are for split I&D.

	Not likely.  The kernel won't fit in 48kb that I know of.  And there
	will be no networking support since that requires supervisor mode
	which non-split I/D systems don't have.

> So if properly configured you can get 2.11 to utilize the user/system spaces.

	The skeleton of a Makefile for non-split a kernel exists but it 
	will take much work (it is essentially just a list of file that may
	or may not be 100% current) to kick into shape.  Also, remember that
	programs like 'csh', 'vi' and so on are not only split I/D but 
	overlaid - they will not run on a non-split machine. 

	Steven Schultz

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA15443
	for pups-liszt; Thu, 18 Feb 1999 12:44:04 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 11:43:27 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 20:43:27 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <36CAEA1F.D5D7C838@halcyon.com> from "David C. Jenner" at Feb 17, 99 08:11:12 am
Message-ID: <199902180143.UAA01509@math.uwaterloo.ca>

| I don't have any docs on the DECNA, but they must exist.  It's
| probably pretty close to the DEQNA.

The DECNA uses one of the earlier Intel network chips.  It lives
on the CTI bus, a bus like no other.  I believe the DEQNA is T-11
based and lives on the vastly better known Q-bus...  -- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id FAA18371
	for pups-liszt; Fri, 19 Feb 1999 05:47:10 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Fri Feb 19 04:46:35 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Thu, 18 Feb 1999 13:46:35 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <01BE5B64.56247680@SONAR> from "James Lothian" at Feb 18, 99 01:14:51 pm
Message-ID: <199902181846.NAA05766@math.uwaterloo.ca>

I'm going to give up as I seem to remember nothing anymore... sigh.
Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
over three years and certainly miss being able to pick up the part in
question before foaming at the mouth spouting nonsense. Many apologies
for suggesting such major inaccuracies.  -- Ken

P.S.  Allison describe the DEQNA as a state-driven device with PALs
      (I think) and that "big F" may the the gate array also mentioned.

| From simul8 at simul8.demon.co.uk  Thu Feb 18 12:27:23 1999
| 
| Just for the sake of being picky... the DEQNA is based on an Intel 
| microcontroller chip (something 8085-ish, I think). The ethernet chipset
| seems to be Fairchild (it's certainly got a big F on it.)
| 
| James


From erin at coffee.corliss.net  Fri Feb 19 06:39:35 1999
From: erin at coffee.corliss.net (Erin W. Corliss)
Date: Thu, 18 Feb 1999 12:39:35 -0800 (PST)
Subject: VaxMate
Message-ID: <Pine.LNX.3.96.990218123109.5732A-100000@coffee.corliss.net>


My local computer junk store has a VaxMate for sale.  I'm not sure of the
model -- It has a DB-25 serial port, 10-base-2 ethernet, and a phone-jack
like printer port on the back, as well as an internal ST-225 hard drive
and a 5.25 inch floppy drive.  

Anyway, when I turn it on it tries to boot up -- the graphical slider
thing on the screen gets about 90% of the way across and it displays the
number 83, which I assume is an eeror code since the number changes if you
boot it up with no keyboard.  Anyone know what the 83 means or where I can
get a list of VaxMate error codes?  Also, how intelligent is this machine
compared to a terminal?  Will it actually run a Vax operating system or
does it need a server?

--------------------------------------------------------
"...color flashing thunder crashing dynamite machine..."


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id KAA19267
	for pups-liszt; Fri, 19 Feb 1999 10:25:25 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 09:24:43 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 00:24:43 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902181846.NAA05766@math.uwaterloo.ca>
Message-ID: <Pine.VUL.3.93.990219002351.1502D-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Ken Wellsch wrote:

> I'm going to give up as I seem to remember nothing anymore... sigh.
> Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
> I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
> over three years and certainly miss being able to pick up the part in
> question before foaming at the mouth spouting nonsense. Many apologies
> for suggesting such major inaccuracies.  -- Ken
> 
> P.S.  Allison describe the DEQNA as a state-driven device with PALs
>       (I think) and that "big F" may the the gate array also mentioned.

You might not be totally out. I also thought the DEQNA was T-11 based,
since the DEUNA is. :-)

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19430
	for pups-liszt; Fri, 19 Feb 1999 11:22:47 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 10:22:33 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 19:22:33 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190022.AA25325@world.std.com>

<> I'm going to give up as I seem to remember nothing anymore... sigh.
<> Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
<> I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
<> over three years and certainly miss being able to pick up the part in
<> question before foaming at the mouth spouting nonsense. Many apologies
<> for suggesting such major inaccuracies.  -- Ken
<> 
<> P.S.  Allison describe the DEQNA as a state-driven device with PALs
<>       (I think) and that "big F" may the the gate array also mentioned.
<
<You might not be totally out. I also thought the DEQNA was T-11 based,
<since the DEUNA is. :-)

I have a DEQNA in front of me.  There is a micro and that is a 8751 8bitter.
The big chip is a LSI ASIC that is a linked list DMA controller.  No t-11.
The RQDXn(n={1,2,3} uses a t-11.  The DELQA also does not use a T-11.
Both use lots of logic in PALs and ASICs to perform several state machines
needed for eithenet.  At the time of development there were few complete
and fast enough chipsets for eithernet.  The DEQNA is mid 80s design and
quite old.

The DEUNA is quite different.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19476
	for pups-liszt; Fri, 19 Feb 1999 11:40:24 +1100 (EST)

From johnh at psych.usyd.edu.au  Fri Feb 19 10:40:17 1999
From: johnh at psych.usyd.edu.au (johnh at psych.usyd.edu.au)
Date: Fri, 19 Feb 1999 11:40:17 +1100 (EST)
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190040.LAA27443@psychwarp.psych.usyd.edu.au>


| Just for the sake of being picky... the DEQNA is based on an Intel
| microcontroller chip (something 8085-ish, I think). The ethernet chipset
| seems to be Fairchild (it's certainly got a big F on it.)
|

The DEQNA uses a Intel 8751 (an EPROM version of 8051 family). I suspect that
it may deal with the programming protocol and the ring buffers. The
chip with the F (with bars top and bottom of the letter) is probably
Fujitsu.

These boards had a fairly bad reputation for lockups and dropped packets.
There was a 20+ wire ECO along with a PAL chip (with 8 of the pins cut off)
soldered on top of another chip.

The replacement ethernet controller was the DELQA, which was a complete
redesign and used a 68000 processor.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19489
	for pups-liszt; Fri, 19 Feb 1999 11:41:28 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 10:41:03 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 01:41:03 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190022.AA25325@world.std.com>
Message-ID: <Pine.VUL.3.93.990219013735.1502E-100000@Zeke.Update.UU.SE>

Hi, Alison.

> <You might not be totally out. I also thought the DEQNA was T-11 based,
> <since the DEUNA is. :-)
> 
> I have a DEQNA in front of me.  There is a micro and that is a 8751 8bitter.
> The big chip is a LSI ASIC that is a linked list DMA controller.  No t-11.

I'm sorry. I didn't mean to imply that you were wrong, just that I was.

> The RQDXn(n={1,2,3} uses a t-11.  The DELQA also does not use a T-11.

Never looked carefully at RQDX?, but the DELQA uses an M68K, that much I
*do* know. (As do the DELUA)

> Both use lots of logic in PALs and ASICs to perform several state machines
> needed for eithenet.  At the time of development there were few complete
> and fast enough chipsets for eithernet.  The DEQNA is mid 80s design and
> quite old.

You obviously knows more about this than I do. :-)
However, as I said, atleast the DELQA have an M68K...
And the DEQNA is old, yes...

> The DEUNA is quite different.

Obviously. But it is also pretty old. Not as buggy though, which should
have been a clue. :-)

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA19674
	for pups-liszt; Fri, 19 Feb 1999 12:57:56 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 11:57:38 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 20:57:38 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190157.AA29020@world.std.com>

<I'm sorry. I didn't mean to imply that you were wrong, just that I was.

Not an argument, just posting to the group what went private by error.

<Never looked carefully at RQDX?, but the DELQA uses an M68K, that much I
<*do* know. (As do the DELUA)

Having two Qbus VAXen and several Qbus PDP-11s it's old turf.  Also I worked 
for DEC Engineering.  that and I've done a lot of hardware level work on my 
systems (repaired dead boards) so the designs are more familair.

<You obviously knows more about this than I do. :-)
<However, as I said, atleast the DELQA have an M68K...
<And the DEQNA is old, yes...

DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
(less buggy too) of the DEQNA and is largely logically the same as the 
DEQNA.

<> The DEUNA is quite different.
<
<Obviously. But it is also pretty old. Not as buggy though, which should
<have been a clue. :-)

Also The DELUA.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA19690
	for pups-liszt; Fri, 19 Feb 1999 12:59:39 +1100 (EST)

From eekg at ix.netcom.com  Fri Feb 19 11:48:59 1999
From: eekg at ix.netcom.com (Eric Edwards)
Date: Thu, 18 Feb 1999 20:48:59 -0500
Subject: 2.9BSD:  mbuf.h
Message-ID: <006401be5baa$06ce3da0$33d1b7c7@eric-edwards>

I'm not sure if anyone mentioned this, but you can build a working 2.9
kernel (sans network) from the sources by just commenting out the references
to the networking include files.  I think there is an offending reference in
syslocal.c also.

Eric Edwards
eekg at ix.netcom.com
mag at csh.rit.edu

-----Original Message-----
From: maximum entropy <entropy at zippy.bernstein.com>
To: pups at minnie.cs.adfa.edu.au <pups at minnie.cs.adfa.edu.au>
Date: Tuesday, February 16, 1999 11:36 PM
Subject: 2.9BSD: mbuf.h


>"make unix" failed:
>Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.



Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19762
	for pups-liszt; Fri, 19 Feb 1999 13:14:46 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 12:14:32 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 21:14:32 -0500
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190214.AA14211@world.std.com>


<The DEQNA uses a Intel 8751 (an EPROM version of 8051 family). I suspect th
<it may deal with the programming protocol and the ring buffers. The
<chip with the F (with bars top and bottom of the letter) is probably
<Fujitsu.

Correct on both cases.  

<These boards had a fairly bad reputation for lockups and dropped packets.
<There was a 20+ wire ECO along with a PAL chip (with 8 of the pins cut off
<soldered on top of another chip.

Actually there were revs A->n and each rev had a step.  The last one was 
N-11... it was marginal.  Good one tended to be good and the bad were PITA.
Also they tended to fail far often than MTBF predictions.

<The replacement ethernet controller was the DELQA, which was a complete
<redesign and used a 68000 processor.

The DELQA was not 68000. The board was far to small for that and had to be 
Qbus dual width and compatable with DEQNA. I have a few of them in my vaxen 
too.  The Unibus versions DEUNA and the later DELUA were 68k and very good. 
They were partly the reason why 730s and 750s were used for routers long 
after they were replaced for other tasks.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19811
	for pups-liszt; Fri, 19 Feb 1999 13:23:12 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 12:22:56 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 21:22:56 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190222.AA21412@world.std.com>

<DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
<(less buggy too) of the DEQNA and is largely logically the same as the 
<DEQNA.

Memory parity exception...  Eat foot time.

DELQA M7516 is 68k and lance chip... had to pull mine to check.  The M7504
however I am correct as I pulled one down from the shelf before dining on 
foot.  Oh and the reson I forgot it's 68k, was the DELQA is far more 
reliable!  that and I only open the BA123 one a year to check the fans and 
clean dust.  It just don't break. ;)

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19969
	for pups-liszt; Fri, 19 Feb 1999 13:36:31 +1100 (EST)

From SHOPPA at trailing-edge.com  Fri Feb 19 12:36:11 1999
From: SHOPPA at trailing-edge.com (Tim Shoppa)
Date: Thu, 18 Feb 1999 21:36:11 -0500
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <990218213611.202000b8@trailing-edge.com>

><The replacement ethernet controller was the DELQA, which was a complete
><redesign and used a 68000 processor.
>The DELQA was not 68000.

Hate to turn this into a "no it isn't, yet it is" sequence, but all
my DELQA's have prominent 68000's on 'em.

> The board was far to small for that

No, it isn't.  The 68000 is the quad pack, and is smaller than either
of the two custom gate arrays that does the Q-bus handshaking.

Tim.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id OAA20047
	for pups-liszt; Fri, 19 Feb 1999 14:07:20 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Fri Feb 19 13:06:14 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Fri, 19 Feb 1999 14:06:14 +1100 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <006401be5baa$06ce3da0$33d1b7c7@eric-edwards> from Eric Edwards at "Feb 18, 1999  8:48:59 pm"
Message-ID: <199902190306.OAA02230@henry.cs.adfa.edu.au>

In article by Eric Edwards:
> I'm not sure if anyone mentioned this, but you can build a working 2.9
> kernel (sans network) from the sources by just commenting out the references
> to the networking include files.  I think there is an offending reference in
> syslocal.c also.
> 
> Eric Edwards
> eekg at ix.netcom.com
> mag at csh.rit.edu

What is happening is that `make depend' invokes a script which finds
#includes in the source code, and builds a make dependency. However,
it's not very intelligent, and doesn't ignore:

#ifdef INET
#include <stuff>

when INET isn't defined. :-) This bites on several C files.
You just have to hand-prune the Makefile after make depend :-)
This is 2.9BSD, BTW, ignore if you're not using it.

Ciao!
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id VAA21635
	for pups-liszt; Fri, 19 Feb 1999 21:20:30 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 20:19:31 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 11:19:31 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190157.AA29020@world.std.com>
Message-ID: <Pine.VUL.3.93.990219111730.26158A-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Allison J Parent wrote:

> <You obviously knows more about this than I do. :-)
> <However, as I said, atleast the DELQA have an M68K...
> <And the DEQNA is old, yes...
> 
> DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
> (less buggy too) of the DEQNA and is largely logically the same as the 
> DEQNA.

Really? I have a DELQA sitting right in front of me, and when I look at
it, the large chip definitely says M68000. What could that be then?

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id VAA21652
	for pups-liszt; Fri, 19 Feb 1999 21:23:04 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 20:22:35 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 11:22:35 +0100 (MET)
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190214.AA14211@world.std.com>
Message-ID: <Pine.VUL.3.93.990219112030.26158B-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Allison J Parent wrote:

> <The replacement ethernet controller was the DELQA, which was a complete
> <redesign and used a 68000 processor.
> 
> The DELQA was not 68000. The board was far to small for that and had to be 
> Qbus dual width and compatable with DEQNA. I have a few of them in my vaxen 
> too.  The Unibus versions DEUNA and the later DELUA were 68k and very good. 

Hate to disagree with you, Alison. The the DELQA really is 68000, take a
peek inside yourself. It is a dual-width too...

And the DEUNA is T-11, while the DELUA is 68000.
I have never bothered plugging in any DEUNAs myself, since DELUAs are
pretty common, and they atleast are pretty good. Never had any problems
with any of them.

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol



From g4klx at pop.agri.ch  Sun Feb 21 04:10:45 1999
From: g4klx at pop.agri.ch (Jonathan Naylor)
Date: Sat, 20 Feb 1999 19:10:45 +0100 (CET)
Subject: V7 filesystem work
Message-ID: <Pine.LNX.4.05.9902201906360.1385-100000@g4klx.agri.ch>

Hello All

A couple of weeks ago I hacked the program v7 from the bostic_tools to
work under all sorts of different Unix versions. It worked great and
allowed me to snoop around the V7 file system images from native Linux.
Anyone who wants a copy can send me an e-mail.

Anyway I had a few hours spare today, and decided to try adding the V7
filesystem to the Linux kernel. Results so far are encouraging:


g4klx:/usr/src/linux# ls -l /mnt
total 333
drwxrwxrwx   7 root     root          224 Sep 22  1988 .
drwxr-xr-x  19 root     root         1024 Feb 14 11:55 ..
drwxrwxr-x   2 3        3            2512 Sep 22  1988 bin
-rwxr-xr-x   1 3        3            8986 Jun  8  1979 boot
drwxrwxr-x   2 3        3             160 Sep 22  1988 dev
drwxrwxr-x   2 3        3             336 Sep 22  1988 etc
-rwxr-xr-x   1 daemon   daemon      53302 Jun  8  1979 hphtunix
-rwxr-xr-x   1 daemon   daemon      52850 Jun  8  1979 hptmunix
drwxrwxr-x   2 3        3             192 Sep 22  1988 lib
drwxrwxr-x   2 root     lp             96 Sep 22  1988 mdec
-rwxr-xr-x   1 root     daemon      50990 Jun  8  1979 rkunix
-rwxr-xr-x   1 root     daemon      51982 Jun  8  1979 rl2unix
-rwxr-xr-x   1 daemon   daemon      51790 Jun  8  1979 rphtunix
-rwxr-xr-x   1 daemon   daemon      51274 Jun  8  1979 rptmunix
g4klx:/usr/src/linux# df
Filesystem         1024-blocks  Used Available Capacity Mounted on
/dev/hda1            3031184 1920771   953665     67%   /
/dev/loop0              1919    1877       42     98%   /mnt
g4klx:/usr/src/linux#


I am using the loop block device to allow me to mount a file as a block
device, this saves me having to add a new partition to my disc. There
should be no reason why it won't work with a true disc partition. The V7
filesystem under Linux is read/write.

Anyone interested ?

Jonathan




From mallison at konnections.com  Sun Feb 21 12:13:38 1999
From: mallison at konnections.com (Mike Allison)
Date: Sat, 20 Feb 1999 19:13:38 -0700
Subject: V7 filesystem work
Message-ID: <199902210203.TAA18682@mail.konnections.com>

Yeah, I'm interested.  Can you write up what changes the linux port entailed???

-Mike


At 07:10 PM 2/20/99 +0100, g4klx at g4klx.demon.co.uk wrote:
>Hello All
>
>A couple of weeks ago I hacked the program v7 from the bostic_tools to
>work under all sorts of different Unix versions. It worked great and
>allowed me to snoop around the V7 file system images from native Linux.
>Anyone who wants a copy can send me an e-mail.
>
>Anyway I had a few hours spare today, and decided to try adding the V7
>filesystem to the Linux kernel. Results so far are encouraging:
>
>
>g4klx:/usr/src/linux# ls -l /mnt
>total 333
>drwxrwxrwx   7 root     root          224 Sep 22  1988 .
>drwxr-xr-x  19 root     root         1024 Feb 14 11:55 ..
>drwxrwxr-x   2 3        3            2512 Sep 22  1988 bin
>-rwxr-xr-x   1 3        3            8986 Jun  8  1979 boot
>drwxrwxr-x   2 3        3             160 Sep 22  1988 dev
>drwxrwxr-x   2 3        3             336 Sep 22  1988 etc
>-rwxr-xr-x   1 daemon   daemon      53302 Jun  8  1979 hphtunix
>-rwxr-xr-x   1 daemon   daemon      52850 Jun  8  1979 hptmunix
>drwxrwxr-x   2 3        3             192 Sep 22  1988 lib
>drwxrwxr-x   2 root     lp             96 Sep 22  1988 mdec
>-rwxr-xr-x   1 root     daemon      50990 Jun  8  1979 rkunix
>-rwxr-xr-x   1 root     daemon      51982 Jun  8  1979 rl2unix
>-rwxr-xr-x   1 daemon   daemon      51790 Jun  8  1979 rphtunix
>-rwxr-xr-x   1 daemon   daemon      51274 Jun  8  1979 rptmunix
>g4klx:/usr/src/linux# df
>Filesystem         1024-blocks  Used Available Capacity Mounted on
>/dev/hda1            3031184 1920771   953665     67%   /
>/dev/loop0              1919    1877       42     98%   /mnt
>g4klx:/usr/src/linux#
>
>
>I am using the loop block device to allow me to mount a file as a block
>device, this saves me having to add a new partition to my disc. There
>should be no reason why it won't work with a true disc partition. The V7
>filesystem under Linux is read/write.
>
>Anyone interested ?
>
>Jonathan
>
>
>
>


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA02555
	for pups-liszt; Sun, 21 Feb 1999 13:29:44 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Sun Feb 21 12:29:29 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Sun, 21 Feb 1999 13:29:29 +1100 (EST)
Subject: V7 filesystem work
In-Reply-To: <Pine.LNX.4.05.9902201906360.1385-100000@g4klx.agri.ch> from Jonathan Naylor at "Feb 20, 1999  7:10:45 pm"
Message-ID: <199902210229.NAA08687@henry.cs.adfa.edu.au>

In article by Jonathan Naylor:
> Hello All
> 
> A couple of weeks ago I hacked the program v7 from the bostic_tools to
> work under all sorts of different Unix versions. It worked great and
> allowed me to snoop around the V7 file system images from native Linux.
> Anyone who wants a copy can send me an e-mail.
> 
> Anyway I had a few hours spare today, and decided to try adding the V7
> filesystem to the Linux kernel. Results so far are encouraging:

> I am using the loop block device to allow me to mount a file as a block
> device, this saves me having to add a new partition to my disc. There
> should be no reason why it won't work with a true disc partition. The V7
> filesystem under Linux is read/write.
> 
> Anyone interested ?

I'd be happy to add any changes etc. into the Tools directory in the PUPS
Archive.

It's about time Unix could read the Unix filesystem again :-)

Ciao,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id UAA05201
	for pups-liszt; Sun, 21 Feb 1999 20:23:41 +1100 (EST)

From g4klx at pop.agri.ch  Sun Feb 21 19:07:50 1999
From: g4klx at pop.agri.ch (Jonathan Naylor)
Date: Sun, 21 Feb 1999 10:07:50 +0100 (CET)
Subject: V7 filesystem work
In-Reply-To: <199902210203.TAA18682@mail.konnections.com>
Message-ID: <Pine.LNX.4.05.9902210957580.303-100000@g4klx.agri.ch>

Hello Mike and the list

On Sat, 20 Feb 1999, Mike Allison wrote:
> Yeah, I'm interested.  Can you write up what changes the linux port entailed???
> 
> -Mike

I assume you mean the standalone V7 FS program rather than the kernel V7
FS support ?

The code was written in old C, and from a modern C programmers point of
view, rather sloppily. The warnings from the compiler were terrible, so I
added function prototypes, and made the code more ANSI C like. Then I got
rid of a few bugs, in one place I remember a character pointer being
assigned to a character.

I then typedef'd the data types so I could use int8, int16 and int32 in
the code to make it more portable. I stopped using structure overlays onto
the raw data as that is messy and is not good for (a) byte ordering and
(b) structure packing. It also allowed me to stop using the original V7
file headers which would have made a public release of the code
problematic.

The data is extracted from the raw block data by using special
architecturally neutral functions into locally held structures. That is a
particular win with the block number in three bytes trick that is used in
the inode.

It has been tested on i386/Linux with both glibc 1.0 and glibc 2.0 and
Alpha/Linux, no changes were needed.

Then I added a few new commands to let me look at the superblock and
bootblocks and a few other bits.

Then I released it.

I have just sent a copy of the program to Warren for inclusion in the PUPS
tools section. Its not very big.

Work is progressing on the V7 filesystem in the Linux kernel. Anyone who
wants the patches for that should send me an e-mail. I hope to get it into
the mainstream kernel in the Linux 2.3 series.

Jonathan



From wkt at henry.cs.adfa.edu.au  Tue Feb 23 15:25:51 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Tue, 23 Feb 1999 16:25:51 +1100 (EST)
Subject: 2.9BSD & Pro: another version
Message-ID: <199902230525.QAA00483@henry.cs.adfa.edu.au>

Ken Wellsch has just uploaded a set of RX50 disk images containing
2.9BSD for the Pro 350 to the PUPS Archive. You can find them in

	Distributions/ucb/2.9bsd4pro350-kcwellsc

He says:

	I believe the RX50 is actually 80 tracks with 10 sectors per track,
	thus yielding 800 blocks per disk.  I think the first track is
	reserved and thus Venix would not let me at it.  Hopefully I have
	not also lost additional information here too.

All the 34 disk images he sent in are 790 blocks long. Can anybody
tell us if we will need to recover track 0 to make these images useful?

At the very least, I've managed to find the pcreg.h file out of the
images (cat */*.rx50 | less -B), so I'm getting closer at recompiling
the 2.9/Pro kernel.

Cheers,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA23748
	for pups-liszt; Wed, 24 Feb 1999 03:01:12 +1100 (EST)

From jesper at df.lth.se  Wed Feb 24 02:00:54 1999
From: jesper at df.lth.se (Jesper Nilsson)
Date: Tue, 23 Feb 1999 17:00:54 +0100 (CET)
Subject: SCO Source license tainting?
Message-ID: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se>

Tjo!

I have a question that I hope someone can help me with:

I'm thinking about getting myself a SCO Source license,
but I'm worried that I might get "tainted" by this
since my day job involves writing operating systems...
My employer would not appreciate getting sued because of
my hobbies...:-)

Has anyone done research about this aspect of the license?

My goals are twofold, running an older Unix version on my PDP-11's,
and of course I want to peruse the source of the classic versions.

/^JN - Jesper Nilsson
--
              I've heard of UNIseX, but I've never had it.
                  Jesper Nilsson -- jesper at df.lth.se



From wkt at henry.cs.adfa.edu.au  Wed Feb 24 07:52:15 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Wed, 24 Feb 1999 08:52:15 +1100 (EST)
Subject: SCO Source license tainting?
In-Reply-To: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se> from Jesper Nilsson at "Feb 23, 1999  5: 0:54 pm"
Message-ID: <199902232152.IAA04722@henry.cs.adfa.edu.au>

In article by Jesper Nilsson:
> I'm thinking about getting myself a SCO Source license,
> but I'm worried that I might get "tainted" by this
> since my day job involves writing operating systems...
> My employer would not appreciate getting sued because of
> my hobbies...:-)

As long as you don't reuse tainted Unix source code in your job, you will
be ok. There are so many books covering the Unix kernel: Lions, Bach,
Goodheart, Vahalia etc., that any concerns other than source code reuse
are negligible.

That's my feelings, anyway.

	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA25583
	for pups-liszt; Wed, 24 Feb 1999 11:46:21 +1100 (EST)

From grog at lemis.com  Wed Feb 24 10:46:07 1999
From: grog at lemis.com (Greg Lehey)
Date: Wed, 24 Feb 1999 11:16:07 +1030
Subject: SCO Source license tainting?
In-Reply-To: <199902232152.IAA04722@henry.cs.adfa.edu.au>; from Warren Toomey on Wed, Feb 24, 1999 at 08:52:15AM +1100
References: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se> <199902232152.IAA04722@henry.cs.adfa.edu.au>
Message-ID: <19990224111607.G93492@lemis.com>

On Wednesday, 24 February 1999 at  8:52:15 +1100, Warren Toomey wrote:
> In article by Jesper Nilsson:
>> I'm thinking about getting myself a SCO Source license,
>> but I'm worried that I might get "tainted" by this
>> since my day job involves writing operating systems...
>> My employer would not appreciate getting sued because of
>> my hobbies...:-)
>
> As long as you don't reuse tainted Unix source code in your job, you will
> be ok. There are so many books covering the Unix kernel: Lions, Bach,
> Goodheart, Vahalia etc., that any concerns other than source code reuse
> are negligible.

I think this relates to a spectre raised during the USL/BSDI wars.
Somebody suggested that anybody who had been exposed to AT&T source
code was ``tainted'' and could thus not legally develop competitive
systems.  Somewhere I have a button that somebody brought back to me
from a USENIX, with the text ``mentally contaminated''.

Jesper, I don't think you need to worry about the problem.  That kind
of restriction would be unenforceable.

Greg
--
See complete headers for address, home page and phone numbers
finger grog at lemis.com for PGP public key


From dave at fgh.geac.com.au  Mon Feb  1 21:21:28 1999
From: dave at fgh.geac.com.au (Dave Horsfall)
Date: Mon, 1 Feb 1999 22:21:28 +1100 (EST)
Subject: Old UNIX file system formats
In-Reply-To: <199902010020.TAA29097@math.uwaterloo.ca>
Message-ID: <Pine.GSO.4.03.9902012214270.10349-100000@fgh>

On Sun, 31 Jan 1999, Ken Wellsch wrote:

> I didn't see mention of the flag "HUGE" WRT the V6 file format.
> Now I may be being mislead from my memory of Venix 1.x which is
> a derivative of V6 (while Venix 2.x is SysIII I think).  If the
> HUGE bit is set in the i-node, then and only then is the 8th
> index pointer treated as the indirection variety.  Thus 8 block
> or less files I think are directly indexed.  -- Ken

I think you're confusing LARGE with HUGE.  I don't have my Edition 5
manual handy, but in my Edition 6 manual if the LARGE bit is set then
the first seven inode pointers are indirect blocks (otherwise all
eight blocks are direct), and the eighth block is a double-indirect block.

-- 
Dave Horsfall VK2KFU  dave at geac.com.au  Ph: +61 2 9978-7493 Fx: +61 2 9978-7422
Geac Computers P/L (FGH Division) 2/57 Christie St, St Leonards 2065, Australia


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id EAA29200
	for pups-liszt; Tue, 2 Feb 1999 04:57:31 +1100 (EST)

From erin at coffee.corliss.net  Tue Feb  2 04:01:14 1999
From: erin at coffee.corliss.net (Erin W. Corliss)
Date: Mon, 1 Feb 1999 10:01:14 -0800 (PST)
Subject: Old UNIX file system formats
In-Reply-To: <199902010020.TAA29097@math.uwaterloo.ca>
Message-ID: <Pine.LNX.3.96.990201095832.20104A-100000@coffee.corliss.net>

> I didn't see mention of the flag "HUGE" WRT the V6 file format.
> Now I may be being mislead from my memory of Venix 1.x which is
> a derivative of V6 (while Venix 2.x is SysIII I think).  If the
> HUGE bit is set in the i-node, then and only then is the 8th
> index pointer treated as the indirection variety.  Thus 8 block
> or less files I think are directly indexed.  -- Ken

Hmm...  I wrote a disk image editor in Visual Basic without knowing the
specs for the filesystem -- I set it up so that if the 9th pointer is zero
and the filesize is greater than one block, then it assumed the block
pointed to by the 8th pointer was a list of blocks in the file.



From mxs46 at k2.scl.cwru.edu  Tue Feb  2 15:22:10 1999
From: mxs46 at k2.scl.cwru.edu (Michael Sokolov)
Date: Tue, 2 Feb 1999 00:22:10 -0500
Subject: Hardware free to a good home
Message-ID: <199902020522.AAA07686@skybridge.scl.cwru.edu>

Dear Ladies and Gentlemen,

I have got some hardware I have to get rid of by the end of February, and it's
free to any of you guys if you are willing to come and pick it up in Cleveland,
Ohio, USA.

Last November I received a load of equipment from one company here in
Cleveland. It was a complicated network of CPUs and peripherals of all makes
and models put together by Xerox and intended to be used as a dedicated
document processing system. The CPUs included one VAX, three unidentifiable
towers, and a bunch of PCs. I'm using the VAX and all disk and tape drives
myself for my own purposes, and I'm selling the PCs, but still I've got those
three unidentifiable towers and three very funky monitors that were attached to
them. There is also a very funky laser printer attached to one of them. Given
that the VAX and all disk and tape drives have been taken out of the equation,
it's unlikely that the rest of the stuff can still be used for that dedicated
document processing whatever thing, but the towers have some apparently generic
controller boards in them (VME or something like that) and other parts that can
be raided for. Who knows, maybe even the CPUs are standard (probably some 68K),
in which case someone who knows more about this than I do (NULL) may be able to
actually use these machines for something.

The only identification on this equipment are the Xerox model numbers. One of
the towers was called NS8090 File Server. It had an external SCSI hard disk and
an Exabyte tape drive, but I've reused these for my own purposes. The other two
towers were called 6085 workstations, and they were diskless from the beginning
(as far as I can tell they don't have any mass storage controllers). All three
have monitors with very funny connectors. Aside from the Xerox model numbers
which tell me absolutely nothing, there are no hints whatsoever as to what the
CPU architecture is and all that. All towers have AUI Ethernet ports.

The laser printer is called NS8000 Laser CP, and it was attached to the tower
that was called the NS8090 File Server. The connectors are 25-pin like the
serial and parallel ones, but they have slide locks like on AUI. These slide
locks and the fact that the printer was apparently never intended to be
connected to anything except an "NS8090 File Server" suggest that the printer's
interface is not parallel or serial, but something very funny.

It has been suggested to me that I take the boxes apart, ID as many boards as
possible, and try to sell/donate them to whoever finds them useful (and the
cabinets and such would probably have to be scrapped). However, the thing is,
I don't really have time for all this, and it's naive to think that any of this
stuff has any significant cash value.

Right now I'm in the process of moving to another (cheaper) apartment in
another part of Cleveland, and really don't feel like hauling that junk around
with me. I have got these three CPU towers, three monitors, and one laser
printer, all absolutely unidentifiable, that I have to get rid of. Given what a
great job I've done at identifying and describing this stuff, it would be naive
for me to expect to sell it. Therefore, I'm giving it away for free to anyone
who is willing to come and pick it up. I have to vacate this apartment by the
end of February, and if no one picks this stuff up, I will have no choice but
to throw it in the big dumpster, which would be a great pity if this stuff is
actually useful for something.

Once again, I'm in Cleveland, Ohio, USA.

Michael Sokolov
TUHS 4BSD Coordinator
4.3BSD-* Maintainer
Quasijarus Project Principal Architect & Developer
Phone: 440-449-0299 or 216-217-2579
ARPA Internet SMTP mail: mxs46 at k2.scl.cwru.edu
TUHS WWW page: http://minnie.cs.adfa.edu.au/TUHS/
Quasijarus WWW page: http://minnie.cs.adfa.edu.au/Quasijarus/

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA03045
	for pups-liszt; Wed, 3 Feb 1999 01:53:24 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Wed Feb  3 00:52:35 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Tue, 2 Feb 1999 09:52:35 -0500 (EST)
Subject: Old UNIX file system formats
In-Reply-To: <Pine.LNX.3.96.990201095832.20104A-100000@coffee.corliss.net> from "Erin W. Corliss" at Feb 1, 99 10:01:14 am
Message-ID: <199902021452.JAA29320@math.uwaterloo.ca>

I shouldn't have posted without doing the proper research.  I took a
gander at PUPS/Tools/Filesys/traverse.c.gz which I'm quite sure is one
of the tools I wrote when I was finally able to figure out the contents
of that V6 tape I had (also with no docs - it was such irony to look
at the setup document on the tape *after* figuring the format out that
clearly describes the block layout 8-).  I notice traverse.c.gz does
indeed use the LARG flag, not HUGE.  Since few care, I'll not bother
extracting enough of Venix 1.x to see whether that is where I met the
HUGE flag or it is just my faulty memory...  -- Ken

| From owner-pups at minnie.cs.adfa.edu.au  Mon Feb  1 13:06:04 1999
| 
| Hmm...  I wrote a disk image editor in Visual Basic without knowing the
| specs for the filesystem -- I set it up so that if the 9th pointer is zero
| and the filesize is greater than one block, then it assumed the block
| pointed to by the 8th pointer was a list of blocks in the file.


From bdc at world.std.com  Thu Feb  4 05:11:56 1999
From: bdc at world.std.com (Brian D Chase)
Date: Wed, 3 Feb 1999 11:11:56 -0800 (PDT)
Subject: MicroVAX I console port question.
Message-ID: <Pine.SGI.3.95.990203104445.22510A-100000@world.std.com>

Off-topic, but maybe somewhat related to MicroPDP-11's... I've got a
MicroVAX I in a BA23 enclosure.  I'm presently a bit thrown by the serial
console port.  I'm used to the 9pin MicroVAX II ports.  From what I've
been told, the DB25-M connector for the console requires a special serial
cable for connecting the MicroVAX I up to a terminal.  A null modem cable
is not adequate. 

First, I just wanted to verify that this is correct.  So far I haven't
been able to access a console prompt using either a null modem cable, or a
straight through cable.  So either the system isn't working correctly, or
I need to get the cable right. Secondly, if it does require a special
cable, then what are the pinouts for that cable?

I'm guessing that aside from the processor itself, the MicroVAX I is
probably a lot closer in design to its contemporary MicroPDP-11 systems.
So I'm hoping that the 25pin console port is simillar to what some of you
have worked with on your Q-bus PDP-11's. 

-brian.
---
Brian "JARAI" Chase | http://world.std.com/~bdc/ | VAXZilla LIVES!!!




From cdl at mpl.ucsd.edu  Thu Feb  4 06:37:53 1999
From: cdl at mpl.ucsd.edu (Carl Lowenstein)
Date: Wed, 3 Feb 1999 12:37:53 -0800 (PST)
Subject: MicroVAX I console port question.
Message-ID: <199902032037.MAA03843@mpl.ucsd.edu>

> From owner-pups at minnie.cs.adfa.edu.au Wed Feb  3 11:35 PST 1999
> Date: Wed, 3 Feb 1999 11:11:56 -0800 (PDT)
> From: Brian D Chase <bdc at world.std.com>
> To: PUPS Mailing List <pups at minnie.cs.adfa.oz.au>
> Subject: MicroVAX I console port question.
> Mime-Version: 1.0
> 
> Off-topic, but maybe somewhat related to MicroPDP-11's... I've got a
> MicroVAX I in a BA23 enclosure.  I'm presently a bit thrown by the serial
> console port.  I'm used to the 9pin MicroVAX II ports.  From what I've
> been told, the DB25-M connector for the console requires a special serial
> cable for connecting the MicroVAX I up to a terminal.  A null modem cable
> is not adequate. 

My MicroVax (I and only) handbook vintage 1984 says the cable is a "BC22D-10".

VAX Systems and Options Catalog Oct 1984 describes the BC22D-10 as
"A fully shielded null modem cable".  Two DB25F connectors, 6 wires.

The pins in use are 1,2,3,6,7,20.  I would expect that "null modem"
means (from one end to the other) connect 2-3, 3-2, 7-7, 6-20, 20-6.

The implication is that the computer might need to see DTR asserted
before it talks to the terminal.

    carl

        carl lowenstein         marine physical lab     u.c. san diego
        {decvax|ucbvax} !ucsd!mpl!cdl                 cdl at mpl.ucsd.edu
                                                  clowenstein at ucsd.edu


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id IAA08853
	for pups-liszt; Thu, 4 Feb 1999 08:25:02 +1100 (EST)

From bdc at world.std.com  Thu Feb  4 07:24:36 1999
From: bdc at world.std.com (Brian D Chase)
Date: Wed, 3 Feb 1999 13:24:36 -0800 (PDT)
Subject: MicroVAX I console port question.
In-Reply-To: <199902032037.MAA03843@mpl.ucsd.edu>
Message-ID: <Pine.SGI.3.95.990203132057.7730C-100000@world.std.com>

On Wed, 3 Feb 1999, Carl Lowenstein wrote:

> My MicroVax (I and only) handbook vintage 1984 says the cable is a "BC22D-10".
> 
> VAX Systems and Options Catalog Oct 1984 describes the BC22D-10 as
> "A fully shielded null modem cable".  Two DB25F connectors, 6 wires.
> 
> The pins in use are 1,2,3,6,7,20.  I would expect that "null modem"
> means (from one end to the other) connect 2-3, 3-2, 7-7, 6-20, 20-6.
> 
> The implication is that the computer might need to see DTR asserted
> before it talks to the terminal.

Or given that everything seems to be fine on my end null modem cable-wise,
it's possible that something more serious is wrong with my MicroVAX I.
Does your handbook list what a flashing "1" LED error code means?  
I'll double-check my cabling as well.

-brian.
---
Brian "JARAI" Chase | http://world.std.com/~bdc/ | VAXZilla LIVES!!!


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id GAA13300
	for pups-liszt; Fri, 5 Feb 1999 06:08:49 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb  5 05:07:56 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Thu, 4 Feb 1999 20:07:56 +0100 (MET)
Subject: Memory Management
In-Reply-To: <Pine.LNX.3.96.990121163113.412A-100000@coffee.corliss.net>
Message-ID: <Pine.VUL.3.93.990204192532.4375B-100000@Zeke.Update.UU.SE>

On Thu, 21 Jan 1999, Erin W. Corliss wrote:

> The documentation that Warren gave me describes the memory management
> scheme.  It says that when the machine is first started, the memory
> management unit is disabled -- anyone know how to enable it, and where the
> segmentation registers are (I'm assuming they are in the 0160000-0177777
> range somewhere)?

I haven't seen anyone answering this, so here I go...

Reg.	Addr.

MMR0	777572
MMR1	777574
MMR2	777576
MMR3	772516
UIPAR	777640-777656
UDPAR	777660-777676
UIPDR	777600-777616
UDPDR	777620-777636
SIPAR	772240-772256
SDPAR	772260-772276
SIPDR	772200-772216
SDPDR	772220-772236
KIPAR	772340-772356
KDPAR	772360-772376
KIPDR	772300-772316
KDPDR	772320-772336

xy in xyP?R is:

x:	U - User
	S - Supervisor
	K - Kernel
y:	I - Instruction
	D - Data

PAR is Page Address Register
PDR is Page Description Register

Okay, so for the layout of the registers...

MMR0:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 ! ! ! !     ! ! ! \-/ ! \---/ +-- Enable relocation
 ! ! ! !     ! ! !  !  !   +------ Page number
 ! ! ! !     ! ! !  !  +---------- Page address space I/D
 ! ! ! !     ! ! !  +------------- Page mode
 ! ! ! !     ! ! +---------------- Instruction completed
 ! ! ! !     ! +------------------ Maintenance mode
 ! ! ! !     +-------------------- Enable memory management trap
 ! ! ! +-------------------------- Trap-Memory management
 ! ! +---------------------------- Abort-Read only access violation
 ! +------------------------------ Abort-Page length error
 +-------------------------------- Abort-Non resident

The page info is for when a trap/fault occurs, and tells in which page it
occured.The rest should be pretty obvious.

MMR1:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
 \-------/ \---/ \-------/ \---/
     !       !       !       +---- Register number
     !       !       +------------ Amount changed (2 compl.)
     !       +-------------------- Register numbe
     +---------------------------- Amount changed (2 compl.)

Low byte is written first, and this register tells how much registers have
changed part way through an instruction, which needs to be undone to start
the intruction again.

MMR2:

Virtual address of instruction where fault occured.

MMR3:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
                     ! !   ! ! +--- Enable user D space
                     ! !   ! +----- Enable supervisor D space
                     ! !   +------- Enable kernel D space
                     ! +----------- Enable 22-bit mapping
                     +------------- Enable unibus map

If 22-bit mapping isn't enabled, the machine will be in 18-bit addressign
when MMU is enabled. Unibus-mapping is something I'll skip for now. You
need it for DMA on a 22-bit unibus machine only.

Note that at the end of a MMU trap/abort, MMR0 bit 15-12 must be cleared
for MMR1 and MMR2 to become active again.


>From a virtual address (VA), you get to the physical address (PA) like
this:

APF=VA[15:13]
DF=VA[12:0]

PA=PAR(APF)*64+DF


The PDR looks loke this:

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
! ! ! ! ! ! ! ! ! ! ! ! ! ! ! ! !
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
   \-----------/ ! !     ! \---/
         !       ! !     !   +----- ACF
         !       ! !     +--------- ED
         !       ! +--------------- W
         !       +----------------- A
         +------------------------- PLF

ACF - Access Control Field
	000 - Non resident; abort on all accesses
	001 - Read only; abort on write attempt, memory mgmt trap on read
	010 - Read only; abort on write attempt
	011 - Unused; abort on all accesses - reserved for future use
	100 - Read/Write; memory mgmt trap upon completion of read or write
	101 - Read/Write; memory mgmt trap upon completion of write
	110 - Read/Write; no system trap/abort action
	111 - Unused; abort on all accesses - reserved for future use

A - Access to page has been made.
W - Page has been written to since PAR/PDR was loaded
ED - Expansion direction
PLF - Page length field

Now, have fun...
	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol



From mxs46 at k2.scl.cwru.edu  Mon Feb  8 06:55:35 1999
From: mxs46 at k2.scl.cwru.edu (Michael Sokolov)
Date: Sun, 7 Feb 1999 15:55:35 -0500
Subject: Special Agent Michael Sokolov has moved
Message-ID: <199902072055.PAA09064@skybridge.scl.cwru.edu>

My new address is:

13444 Euclid Ave. Apt. 215
East Cleveland, OH  44112
USA

My new phone # is 216-761-3656 (voice mail not set up yet, will be done in a
couple of days).

I'm still not quite done with all move-related work, so it will be a few more
days before I catch up with my E-mail.

My hardware is laid out a lot better at the new place than at the old one, so
when I'm done hooking everything up, I'll have much better work conditions for
my Project. Also the new place is physically closer to the building where all
Cleveland ISPs are located, reducing the cost of leased lines and increasing
the probability of me getting one some day.

With the hardware taking up most of the space, I originally thought that my
apartment would look like Agent Mulder's, but it actually ended up being more
like Agent Scully's. Oh well, her place is pretty nice too, and so is mine now.

Just a reminder to all Quasijarus Project folks living in the USA, be sure to
watch the X-Files this evening. They'll finally tell us what really happened to
Mulder's sister, who is the cigarette-smoking man, and all the other cool
stuff.

Special Agent Michael Sokolov
TUHS 4BSD Coordinator
4.3BSD-* Maintainer
Quasijarus Project Principal Architect & Developer
Phone: 216-761-3656
ARPA Internet SMTP mail: mxs46 at k2.scl.cwru.edu
TUHS WWW page: http://minnie.cs.adfa.edu.au/TUHS/
Quasijarus WWW page: http://minnie.cs.adfa.edu.au/Quasijarus/


From entropy at zippy.bernstein.com  Wed Feb 17 14:23:11 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Tue, 16 Feb 1999 23:23:11 -0500 (EST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902170423.XAA24481@zippy.bernstein.com>

I'm trying to compile a 2.9BSD kernel using the distribution from the
pups archive.

"make unix" failed:
Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.

I looked in the usr.tar from the distribution, and I don't see mbuf.h
anywhere.

Does anyone know where I can find a copy of this file?

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id QAA11005
	for pups-liszt; Wed, 17 Feb 1999 16:17:41 +1100 (EST)

From sms at moe.2bsd.com  Wed Feb 17 15:15:02 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Tue, 16 Feb 1999 21:15:02 -0800 (PST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902170515.VAA23159@moe.2bsd.com>

Hi -

> I'm trying to compile a 2.9BSD kernel using the distribution from the
> pups archive.
> 
> "make unix" failed:
> Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> 
> I looked in the usr.tar from the distribution, and I don't see mbuf.h
> anywhere.
> 
> Does anyone know where I can find a copy of this file?

	That's not _all_ your missing ;-)

	Unless you have the 1985 Seismo (or Harvard - depends where you
	got the tape from) update tape to 2.9 the networking code won't
	compile much less run.  Been there, done that.  It was a fun couple
	weeks coming to the realization that the networking code hadn't
	been fully integrated and compiled in 2.9

	I believe the 2.9-Seismo update is in the PUPS archive (should be
	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
	painful upgrade process because it changes the a.out header format
	for overlaid processes (goes from 7 to 15 overlays).  If you're not
	real careful you'll have (as I did ;-)) a real mess:  can't finish
	the upgrade because the old kernel doesn't support the new overlaid
	processes but you can't build a new kernel because doing so needs
	those processes.  Something like that.  It was "interesting" ;)

	Steven Schultz

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id QAA11030
	for pups-liszt; Wed, 17 Feb 1999 16:24:28 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Wed Feb 17 15:26:09 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Wed, 17 Feb 1999 16:26:09 +1100 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <199902170515.VAA23159@moe.2bsd.com> from "Steven M. Schultz" at "Feb 16, 1999  9:15: 2 pm"
Message-ID: <199902170526.QAA14818@henry.cs.adfa.edu.au>

In article by Steven M. Schultz:
> Hi -
> 
> > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > pups archive.
> > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > Does anyone know where I can find a copy of this file?
> 
> 	That's not _all_ your missing ;-)
> 
> 	Unless you have the 1985 Seismo (or Harvard - depends where you
> 	got the tape from) update tape to 2.9 the networking code won't
> 	compile much less run.  Been there, done that.  It was a fun couple
> 	weeks coming to the realization that the networking code hadn't
> 	been fully integrated and compiled in 2.9
> 
> 	I believe the 2.9-Seismo update is in the PUPS archive (should be
> 	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> 	painful upgrade process because it changes the a.out header format
> 	for overlaid processes (goes from 7 to 15 overlays).  If you're not
> 	real careful you'll have (as I did ;-)) a real mess:  can't finish
> 	the upgrade because the old kernel doesn't support the new overlaid
> 	processes but you can't build a new kernel because doing so needs
> 	those processes.  Something like that.  It was "interesting" ;)
> 
> 	Steven Schultz

Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro. 
I'm sure he will keep us informed :-)

'Night!

	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id RAA11134
	for pups-liszt; Wed, 17 Feb 1999 17:01:58 +1100 (EST)

From djenner at halcyon.com  Wed Feb 17 16:00:45 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Tue, 16 Feb 1999 22:00:45 -0800
Subject: 2.9BSD:  mbuf.h
References: <199902170526.QAA14818@henry.cs.adfa.edu.au>
Message-ID: <36CA5B0D.8A2B2629@halcyon.com>

Speaking of the Pro, I have one and have been trying to get Venix
to run on it.  The rub is, there are two versions: one directly from
VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Venix/Pro is freely available on the Internet at ftp.update.uu.se,
but Pro/Venix seems to be a little harder to find.  Pro/Venix is
much to be preferred because you can reconfigure the kernel (in
binary) to include different drivers, etc.

I've been able to acquire all the documentation and all (almost) the
disks for Pro/Venix 2.0.  A couple of the disks are apparently
unusable or missing in the set I have.

It seems to me that Pro/Venix is a potential candidate for the PUPS
archive, the snag being DEC/Compaq residual interests in it.  PUPS
covers the AT&T part, VenturCom has "given away" their part, and
DEC/Compaq is all that's left.

So:
	1) Could this be a PUPS addition, if a good copy be found?
	2) If someone has a copy, but worries about the DEC/Compaq
	   aspects, can a good copy of the disks I have be acquired?
	   (Anyone in this category might want to respond directly
	    to me instead of posting to the mailing lists.)  After
	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
	   objections are probably to worry about the AT&T part,
	   which the Ancient Unix license covers...

Actually, I'm amazed I've gotten as far as I have with this, because
I've been pretty passive about finding it.  It's only taken 2 years
so far.

Dave

Warren Toomey wrote:
> 
> In article by Steven M. Schultz:
> > Hi -
> >
> > > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > > pups archive.
> > > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > > Does anyone know where I can find a copy of this file?
> >
> >       That's not _all_ your missing ;-)
> >
> >       Unless you have the 1985 Seismo (or Harvard - depends where you
> >       got the tape from) update tape to 2.9 the networking code won't
> >       compile much less run.  Been there, done that.  It was a fun couple
> >       weeks coming to the realization that the networking code hadn't
> >       been fully integrated and compiled in 2.9
> >
> >       I believe the 2.9-Seismo update is in the PUPS archive (should be
> >       on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> >       painful upgrade process because it changes the a.out header format
> >       for overlaid processes (goes from 7 to 15 overlays).  If you're not
> >       real careful you'll have (as I did ;-)) a real mess:  can't finish
> >       the upgrade because the old kernel doesn't support the new overlaid
> >       processes but you can't build a new kernel because doing so needs
> >       those processes.  Something like that.  It was "interesting" ;)
> >
> >       Steven Schultz
> 
> Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro.
> I'm sure he will keep us informed :-)
> 
> 'Night!
> 
>         Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id RAA11133
	for pups-liszt; Wed, 17 Feb 1999 17:01:50 +1100 (EST)

From djenner at halcyon.com  Wed Feb 17 16:00:45 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Tue, 16 Feb 1999 22:00:45 -0800
Subject: 2.9BSD:  mbuf.h
References: <199902170526.QAA14818@henry.cs.adfa.edu.au>
Message-ID: <36CA5B0D.8A2B2629@halcyon.com>

Speaking of the Pro, I have one and have been trying to get Venix
to run on it.  The rub is, there are two versions: one directly from
VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Venix/Pro is freely available on the Internet at ftp.update.uu.se,
but Pro/Venix seems to be a little harder to find.  Pro/Venix is
much to be preferred because you can reconfigure the kernel (in
binary) to include different drivers, etc.

I've been able to acquire all the documentation and all (almost) the
disks for Pro/Venix 2.0.  A couple of the disks are apparently
unusable or missing in the set I have.

It seems to me that Pro/Venix is a potential candidate for the PUPS
archive, the snag being DEC/Compaq residual interests in it.  PUPS
covers the AT&T part, VenturCom has "given away" their part, and
DEC/Compaq is all that's left.

So:
	1) Could this be a PUPS addition, if a good copy be found?
	2) If someone has a copy, but worries about the DEC/Compaq
	   aspects, can a good copy of the disks I have be acquired?
	   (Anyone in this category might want to respond directly
	    to me instead of posting to the mailing lists.)  After
	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
	   objections are probably to worry about the AT&T part,
	   which the Ancient Unix license covers...

Actually, I'm amazed I've gotten as far as I have with this, because
I've been pretty passive about finding it.  It's only taken 2 years
so far.

Dave

Warren Toomey wrote:
> 
> In article by Steven M. Schultz:
> > Hi -
> >
> > > I'm trying to compile a 2.9BSD kernel using the distribution from the
> > > pups archive.
> > > Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.
> > > Does anyone know where I can find a copy of this file?
> >
> >       That's not _all_ your missing ;-)
> >
> >       Unless you have the 1985 Seismo (or Harvard - depends where you
> >       got the tape from) update tape to 2.9 the networking code won't
> >       compile much less run.  Been there, done that.  It was a fun couple
> >       weeks coming to the realization that the networking code hadn't
> >       been fully integrated and compiled in 2.9
> >
> >       I believe the 2.9-Seismo update is in the PUPS archive (should be
> >       on the CD but my memory isn't ECC these days ;-)).  It's a fairly
> >       painful upgrade process because it changes the a.out header format
> >       for overlaid processes (goes from 7 to 15 overlays).  If you're not
> >       real careful you'll have (as I did ;-)) a real mess:  can't finish
> >       the upgrade because the old kernel doesn't support the new overlaid
> >       processes but you can't build a new kernel because doing so needs
> >       those processes.  Something like that.  It was "interesting" ;)
> >
> >       Steven Schultz
> 
> Don't worry, Nicholas is trying to patch 2.9 to get it to run on a Pro.
> I'm sure he will keep us informed :-)
> 
> 'Night!
> 
>         Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11461
	for pups-liszt; Wed, 17 Feb 1999 19:23:34 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:22:50 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:22:50 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <36CA5B0D.8A2B2629@halcyon.com> (djenner@halcyon.com)
Message-ID: <199902170822.DAA24861@zippy.bernstein.com>

>Date: Tue, 16 Feb 1999 22:00:45 -0800
>From: "David C. Jenner" <djenner at halcyon.com>
>
>Speaking of the Pro, I have one and have been trying to get Venix
>to run on it.  The rub is, there are two versions: one directly from
>VenturCom (Venix/Pro) and one licensed through DEC (Pro/Venix).

Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
for the Pro-380.  I never heard of a distinction between Venix/Pro
vs. Pro/Venix.  Then again, I got into this game fairly late...I
bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
installed (also with original install media and docs).

>Venix/Pro is freely available on the Internet at ftp.update.uu.se,
>but Pro/Venix seems to be a little harder to find.  Pro/Venix is
>much to be preferred because you can reconfigure the kernel (in
>binary) to include different drivers, etc.
>
>I've been able to acquire all the documentation and all (almost) the
>disks for Pro/Venix 2.0.  A couple of the disks are apparently
>unusable or missing in the set I have.

I have the following archives of Venix-related stuff that I snagged
from the net a few years back.  If you think any of them might contain
what you're looking for, let me know and I'll give more detail about
their contents.

-rw-r--r--   1 entropy  user         3833 Oct 17  1997 README
-rw-r--r--   1 entropy  user          532 Oct 17  1997 README.VAX
-rw-r--r--   1 entropy  user        30819 Oct 17  1997 RX50.notes
-rw-r--r--   1 entropy  user      2530759 Oct 17  1997 Venix1.tar.Z
-rw-r--r--   1 entropy  user      2503931 Oct 17  1997 Venix2.tar.Z
-rw-r--r--   1 entropy  user        15817 Oct 17  1997 cathang.txt
-rw-r--r--   1 entropy  user       332543 Oct 17  1997 mopimage.tar.Z
-rw-r--r--   1 entropy  user          443 Oct 17  1997 nbsdrx50.readme
-rw-r--r--   1 entropy  user       897510 Oct 17  1997 nbsdrx50.zip
-rw-r--r--   1 entropy  user       155648 Oct 17  1997 pppd
-rw-r--r--   1 entropy  user       193536 Oct 17  1997 pr0801eng.sys
-rw-r--r--   1 entropy  user        14153 Oct 17  1997 raind112.zip
-rw-r--r--   1 entropy  user         6621 Oct 17  1997 rx50.zip
-rw-r--r--   1 entropy  user        81152 Oct 17  1997 teledisk.zip
-rw-r--r--   1 entropy  user        61440 Oct 17  1997 venix.tar
-rw-r--r--   1 entropy  user          116 Oct 17  1997 venix1.readme
-rw-r--r--   1 entropy  user      1119490 Oct 17  1997 venix1s.zip
-rw-r--r--   1 entropy  user      1095824 Oct 17  1997 venix1u.zip
-rw-r--r--   1 entropy  user          424 Oct 17  1997 venix2.readme
-rw-r--r--   1 entropy  user      1058970 Oct 17  1997 venix2s.zip
-rw-r--r--   1 entropy  user      1145720 Oct 17  1997 venix2u.zip
-rw-r--r--   1 entropy  user       332362 Oct 17  1997 vnx2u2u5.zip

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11492
	for pups-liszt; Wed, 17 Feb 1999 19:33:00 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:32:05 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:32:05 -0500 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <199902170515.VAA23159@moe.2bsd.com> (sms@moe.2bsd.com)
Message-ID: <199902170832.DAA24878@zippy.bernstein.com>

>Date: Tue, 16 Feb 1999 21:15:02 -0800 (PST)
>From: "Steven M. Schultz" <sms at moe.2bsd.com>
>
>	I believe the 2.9-Seismo update is in the PUPS archive (should be
>	on the CD but my memory isn't ECC these days ;-)).  It's a fairly
>	painful upgrade process because it changes the a.out header format
>	for overlaid processes (goes from 7 to 15 overlays).  If you're not
>	real careful you'll have (as I did ;-)) a real mess:  can't finish
>	the upgrade because the old kernel doesn't support the new overlaid
>	processes but you can't build a new kernel because doing so needs
>	those processes.  Something like that.  It was "interesting" ;)

Sounds like fun.  Any hints on the correct upgrade path to avoid this
lossage?

Better yet, would you be willing and able to upload a disk image or
tar file of an upgraded system to the PUPS archive (or directly to me
if it's not of general interest), so I could use that as a starting
point?

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id TAA11533
	for pups-liszt; Wed, 17 Feb 1999 19:43:19 +1100 (EST)

From entropy at zippy.bernstein.com  Wed Feb 17 18:42:40 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 03:42:40 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902170842.DAA24887@zippy.bernstein.com>


As already mentioned in previous messages, I'm working on getting
2.9BSD onto a Pro 350.  I'm using 2.9BSD as a starting point because
it claims to support machines without split i/d.  The 350 uses the
F-11 chipset, which I have read does not support split i/d.

I would prefer to use 2.11BSD because I understand it's still actively
used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
says that it needs split i/d to run.  Can anyone give me more detail
about this?  Was support for machines without split i/d removed from
the kernel, or is it just that some of the programs are too big to fit
in a single 64k segment?

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA12587
	for pups-liszt; Thu, 18 Feb 1999 01:36:09 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 00:35:30 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902170822.DAA24861@zippy.bernstein.com> from "maximum entropy" at Feb 17, 99 03:22:50 am
Message-ID: <199902171435.JAA12462@math.uwaterloo.ca>

| Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
| they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
| for the Pro-380.  I never heard of a distinction between Venix/Pro
| vs. Pro/Venix.  Then again, I got into this game fairly late...I
| bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
| installed (also with original install media and docs).

My time playing with Pro's faded out before Venix 2 was available (free)
for me to try.  I've played a fair bit with Venix 1.1 on both Pro 350's
and Pro 380's.  The Venix 1 series I feel is basically V6 derived while
I understood the Venix 2 series was derived from Sys III.

About a year ago Rick Macklem that did a port to the Pro series mailed
me his "Pro stuff" which included a tape and floppies.  I've forgotten
what all is in that stash, but taking a peek at some old mail he mentions:

> The stuff I did went out on a Usenix distribution tape in about 1983/84
> and had to be merged into a 2.9BSD distribution. I did generate floppy
> sets for a few people, because that was the only easy way to get it
> installed. (The first install here was actually done by downloading the
> kernel over the serial port talking to the PDP 11 prom (ODS?).)

I had thought his set of patches were in the PUPS archive.  In fact I
see the patches under PUPS/Distributions/ucb/2.9-pro350.

-- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id BAA12654
	for pups-liszt; Thu, 18 Feb 1999 01:42:25 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 00:42:05 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 09:42:05 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
In-Reply-To: <199902170842.DAA24887@zippy.bernstein.com> from "maximum entropy" at Feb 17, 99 03:42:40 am
Message-ID: <199902171442.JAA15916@math.uwaterloo.ca>

| I would prefer to use 2.11BSD because I understand it's still actively
| used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
| says that it needs split i/d to run.  Can anyone give me more detail
| about this?  Was support for machines without split i/d removed from
| the kernel, or is it just that some of the programs are too big to fit
| in a single 64k segment?

Have you been able to acquire the documentation for the DECNA card?  I
think that is roughly what it is called.  The Pro Ethernet card.  A few
old timers like myself and Dan Lanciani talked years ago about running
things on a Pro and no-one seems to know much about this relatively
critical bit of documentation.  Again referring to Rick Macklem's
correspondence (I believe I was asking him, again, about these docs):

> Well, the short answer is "I'm not sure what the answers are". At one
> point someone mentioned they were putting the Pro stuff into 2BSD, but
> I'm not sure if they actually did it. (The guys that used it the most
> had it running on a lab of Pro380s at Columbia U. (I think. It's the
> one right in NY city.)) His name was Charlie Kim (again, I think?) and
> did some stuff to it so that it worked reasonably well on a Pro380, but
> I have no idea how you might find him now. (It was a real dog on a Pro350
> because it didn't have separate I and D space.)

The rumors we were able to find all pointed to this place and person
WRT documentation for the ethernet card.

-- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id CAA12725
	for pups-liszt; Thu, 18 Feb 1999 02:12:16 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 01:11:36 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 10:11:36 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902171435.JAA12462@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:35:30 -0500 (EST))
Message-ID: <199902171511.KAA25176@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
>
>About a year ago Rick Macklem that did a port to the Pro series mailed
>me his "Pro stuff" which included a tape and floppies.  I've forgotten
>what all is in that stash, but taking a peek at some old mail he mentions:

Would you be able to send images (rx50 teledisk, or plain dd dumps) of
these disks to me or to the archive?

>I had thought his set of patches were in the PUPS archive.  In fact I
>see the patches under PUPS/Distributions/ucb/2.9-pro350.

Those files aren't 100% complete.  Excerpt of a mail I sent last night
to Warren Toomey:

#The instructions in boot.doc are mangled.
#The patches included are reversed, and didn't apply cleanly to one of
#the files (/usr/src/net/sys/sys/machdep.c).  Also, it looks like the
#guy that produced that set of changes forgot to include his
#modifications to /usr/src/sys/conf/config, but I managed to hack
#together something that might work.

Then there's the fact that the 2.9 distribution won't even compile,
and the 2.9 upgrade patches are a nightmare...

Maybe I'll just stick to venix :-)

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id CAA12848
	for pups-liszt; Thu, 18 Feb 1999 02:44:50 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 01:44:04 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 10:44:04 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902171435.JAA12462@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:35:30 -0500 (EST))
Message-ID: <199902171544.KAA25215@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:35:30 -0500 (EST)
>
>About a year ago Rick Macklem that did a port to the Pro series mailed
>me his "Pro stuff" which included a tape and floppies.  I've forgotten
>what all is in that stash, but taking a peek at some old mail he mentions:

Would you be able to send images (rx50 teledisk, or plain dd dumps) of
these disks to me or to the archive?

>I had thought his set of patches were in the PUPS archive.  In fact I
>see the patches under PUPS/Distributions/ucb/2.9-pro350.

Those files aren't 100% complete.  Excerpt of a mail I sent last night
to Warren Toomey:

#The instructions in boot.doc are mangled.d
#The patches included are reversed, and didn't apply cleanly to one of
#the files (/usr/src/net/sys/sys/machdep.c).  Also, it looks like the
#guy that produced that set of changes forgot to include his
#modifications to /usr/src/sys/conf/config, but I managed to hack
#together something that might work.

Then there's the fact that the 2.9 distribution won't even compile,
and the 2.9 upgrade patches are a nightmare...

Maybe I'll just stick to venix :-)

Cheers,
entropy

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12922
	for pups-liszt; Thu, 18 Feb 1999 03:12:05 +1100 (EST)

From entropy at zippy.bernstein.com  Thu Feb 18 02:11:24 1999
From: entropy at zippy.bernstein.com (maximum entropy)
Date: Wed, 17 Feb 1999 11:11:24 -0500 (EST)
Subject: 2.11BSD, non-split i/d issues
In-Reply-To: <199902171442.JAA15916@math.uwaterloo.ca> (message from Ken
	Wellsch on Wed, 17 Feb 1999 09:42:05 -0500 (EST))
Message-ID: <199902171611.LAA25258@zippy.bernstein.com>

>From: Ken Wellsch <kcwellsc at math.uwaterloo.ca>
>Date: Wed, 17 Feb 1999 09:42:05 -0500 (EST)
>
>Have you been able to acquire the documentation for the DECNA card?  I

I haven't looked for it.  The DECNA is optional, and my Pro doesn't
have it.  All Pro's came with an AUI port, but without the card it
doesn't do anything.

--
entropy -- it's not just a good idea, it's the second law.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12934
	for pups-liszt; Thu, 18 Feb 1999 03:12:18 +1100 (EST)

From djenner at halcyon.com  Thu Feb 18 02:11:12 1999
From: djenner at halcyon.com (David C. Jenner)
Date: Wed, 17 Feb 1999 08:11:12 -0800
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
References: <199902171435.JAA12462@math.uwaterloo.ca>
Message-ID: <36CAEA1F.D5D7C838@halcyon.com>

I haven't tried the 2.9 stuff at all on a Pro.  I have had it
running on an 11/23+ (w/binary license) for 10 years.  The
problem is the networking, as you have found.

Venix/Pro 1.1 and 2.0 run just fine on the Pro 380, and it's
pretty painless to install.  I have distribution disks for
Pro/Venix 1.1, but the install disk has apparently been
overwritten with the 2.0 installation disk.  And my distribution
for 2.0 is missing a couple of original disks; I have copies of
those disks, but they get read errors.

I guess the 2.9 stuff would be interesting if you got it to
work on the Pro, especially if you got networking to work.
I don't have any docs on the DECNA, but they must exist.  It's
probably pretty close to the DEQNA.

Dave

Ken Wellsch wrote:
> 
> | Interesting...I know there's a Venix 1.0 and a Venix 2.0.  I thought
> | they were both from Venturcom, with 1.0 being for the Pro-350 and 2.0
> | for the Pro-380.  I never heard of a distinction between Venix/Pro
> | vs. Pro/Venix.  Then again, I got into this game fairly late...I
> | bought my used Pro-350 around 1993 for US$100, with Venix 1.0 already
> | installed (also with original install media and docs).
> 
> My time playing with Pro's faded out before Venix 2 was available (free)
> for me to try.  I've played a fair bit with Venix 1.1 on both Pro 350's
> and Pro 380's.  The Venix 1 series I feel is basically V6 derived while
> I understood the Venix 2 series was derived from Sys III.
> 
> About a year ago Rick Macklem that did a port to the Pro series mailed
> me his "Pro stuff" which included a tape and floppies.  I've forgotten
> what all is in that stash, but taking a peek at some old mail he mentions:
> 
> > The stuff I did went out on a Usenix distribution tape in about 1983/84
> > and had to be merged into a 2.9BSD distribution. I did generate floppy
> > sets for a few people, because that was the only easy way to get it
> > installed. (The first install here was actually done by downloading the
> > kernel over the serial port talking to the PDP 11 prom (ODS?).)
> 
> I had thought his set of patches were in the PUPS archive.  In fact I
> see the patches under PUPS/Distributions/ucb/2.9-pro350.
> 
> -- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA12957
	for pups-liszt; Thu, 18 Feb 1999 03:17:55 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 02:15:01 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 08:15:01 -0800 (PST)
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902171615.IAA02324@moe.2bsd.com>

Hi

> Sounds like fun.  Any hints on the correct upgrade path to avoid this
> lossage?

	Oh, it's not _completely_ irrecoverable and is "fun" in a perverse
	way.

	First go thru all of the executable directories (/bin, /usr/bin,...)
	and identify all of the overlaid executables and save copies of them.
	Shouldn't be too many but the important one is 'ex'/'vi'.  A number
	of programs rely on using 'ex' scripts to edit generated files (the
	kernel makefiles are _good_ examples;)), and so on.  Having an older
	copy of 'ex'/'vi' is the main thing I remember as saving the day.

> Better yet, would you be willing and able to upload a disk image or
> tar file of an upgraded system to the PUPS archive (or directly to me

	Oh, I have no 2.9 systems - this was all done 10 years ago.  The
	systems I run now use MSCP/TMSCP devices and 2.9 lacks support
	for those.

	Steve

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA13020
	for pups-liszt; Thu, 18 Feb 1999 03:32:46 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 02:32:00 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 08:32:00 -0800 (PST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902171632.IAA02404@moe.2bsd.com>

Hi -

> From: maximum entropy <entropy at zippy.bernstein.com>
> I would prefer to use 2.11BSD because I understand it's still actively
> used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
> says that it needs split i/d to run.  Can anyone give me more detail
> about this?  Was support for machines without split i/d removed from
> the kernel, or is it just that some of the programs are too big to fit
> in a single 64k segment?

	Oh, support was NOT removed.  Non-split executables (magic number
	0407 and 0410) will still run.  

	The kernel will not fit - without split I/D it is impossible to
	create a /unix image that fits within a single 64kb (actually 48kb
	since the kernel stack takes 1 segment and the 'u' area takes
	another) address space.

	I actually went thru the exercise once (2.10 era) of creating a bare 
	bones kernel that would fit in - at least the linker said it would.
	That was only done by ripping out lots of stuff - no networking, no
	statistics gathering, almost no drivers, etc.  Never 'ran' it though
	since there seemed to be little point in such a stripped down system.

	Even V7 was hard pressed to run on a non-split machine!  In fact there
	was a paper written about shoehorning V7 onto an 11/40 and the hoops
	that needed to be jumped thru.  Not sure but that document might be
	in the /usr/doc tree of one of the PUPS Distributions hierarchy.

	Steven


From wkt at henry.cs.adfa.edu.au  Thu Feb 18 08:25:47 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Thu, 18 Feb 1999 09:25:47 +1100 (EST)
Subject: Pro/Venix
In-Reply-To: <36CA5B0D.8A2B2629@halcyon.com> from "David C. Jenner" at "Feb 16, 1999 10: 0:45 pm"
Message-ID: <199902172225.JAA16961@henry.cs.adfa.edu.au>

In article by David C. Jenner:
> It seems to me that Pro/Venix is a potential candidate for the PUPS
> archive, the snag being DEC/Compaq residual interests in it.  PUPS
> covers the AT&T part, VenturCom has "given away" their part, and
> DEC/Compaq is all that's left.
> 
> So:
> 	1) Could this be a PUPS addition, if a good copy be found?
> 	2) If someone has a copy, but worries about the DEC/Compaq
> 	   aspects, can a good copy of the disks I have be acquired?
> 	   (Anyone in this category might want to respond directly
> 	    to me instead of posting to the mailing lists.)  After
> 	   all a PUPS licensee is 99.999% covered, and DEC/Compaq
> 	   objections are probably to worry about the AT&T part,
> 	   which the Ancient Unix license covers...
> 
> Dave

If we could get DEC/Compaq to allow access to Pro/Venix by UNIX source
license holders, then yes I would certainly add it to the Archive. If
there's no source code, and SCO are happy, then it could go up for anon ftp.

Cheers,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id KAA14692
	for pups-liszt; Thu, 18 Feb 1999 10:21:58 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Thu Feb 18 09:18:40 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Thu, 18 Feb 1999 10:18:40 +1100 (EST)
Subject: Help with regs on Pro serial ports
Message-ID: <199902172318.KAA18083@henry.cs.adfa.edu.au>

I'm trying to help get the kernel for the version of 2.9BSD ported to the
Pro-350. The patches supplied by Rick Macklem are slightly incomplete, e.g
there is no config shell script which knows about the new device drivers etc.

Anyway, one vital missing file is pcreg.h, which holds the structure
describing the registers of the serial ports on the Pro-350. By perusing
the file dev/pc.c, I've worked out that the struct looks something like:


struct pcdevice {
	??? baud;
	??? cdb;
	??? csa;
	??? csb;
	??? csr;
	??? dbuf;
	??? mc0;
	??? mc1;
	??? mode;
	??? stat;
}

where the fields are not in the correct order, and I have no idea what
C type each is. If anybody can help recreate this file, could they
email me?!

I've included below the C comments at the top of dev/pc.c.

If anybody has Rick Macklem's email address, could they pass that on too?
I will email him and see if he's got a more complete set of patches somewhere.

Many thanks in advance,

	Warren

/*
 * This driver handles the two serial ports on the back of the
 * pro3xx system unit. Although not software compatible, they
 * are handled as minor device 0 & 1 respectively, for the printer
 * and communication port. Modem control is included but no sync
 * serial support for the com. port.
 * NOTE: The DSR line in the printer port is used for carrier
 * detect so terminals or modems should be cabled accordingly.
 * Local terminal cables should jumper DTR-CDT so that the carrier
 * will appear to be up or PC_SOFTCAR defined and devs or'd with 0200.
 * NOTE2: The interrupt service routines are as follows:
 *      plrint - printer port receive
 *      plxint - printer port transmit
 *      cmintr - communication port com. interrupt
 * Modem transition interrupts are NOT used.
 */

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA15025
	for pups-liszt; Thu, 18 Feb 1999 11:31:53 +1100 (EST)

From allisonp at world.std.com  Thu Feb 18 10:31:29 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Wed, 17 Feb 1999 19:31:29 -0500
Subject: 2.9BSD:  mbuf.h
Message-ID: <199902180031.AA13175@world.std.com>


<Venix/Pro is freely available on the Internet at ftp.update.uu.se,
<but Pro/Venix seems to be a little harder to find.  Pro/Venix is
<much to be preferred because you can reconfigure the kernel (in
<binary) to include different drivers, etc.

The UU.SE and gatway.dec.com version of it I ahve running on my PRO-350
for the last year or more.

I'd like to have SLIP/PPP running on it or even be able to tweek it.

<	1) Could this be a PUPS addition, if a good copy be found?

Oh hand I'd say yes.  

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA15035
	for pups-liszt; Thu, 18 Feb 1999 11:32:11 +1100 (EST)

From allisonp at world.std.com  Thu Feb 18 10:31:41 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Wed, 17 Feb 1999 19:31:41 -0500
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902180031.AA13305@world.std.com>

<As already mentioned in previous messages, I'm working on getting
<2.9BSD onto a Pro 350.  I'm using 2.9BSD as a starting point because
<it claims to support machines without split i/d.  The 350 uses the
<F-11 chipset, which I have read does not support split i/d.

The F11 does not do I&D split but does have user/system.  

<I would prefer to use 2.11BSD because I understand it's still actively
<used, and not as buggy as 2.9.  But everything I've read about 2.11BSD
<says that it needs split i/d to run.  Can anyone give me more detail
<about this?  Was support for machines without split i/d removed from
<the kernel, or is it just that some of the programs are too big to fit
<in a single 64k segment?

It's my understanding that 2.11 will run on F11 systems (pro350 and 11/23)
if properly configured but the only binaries loose are for split I&D.
So if properly configured you can get 2.11 to utilize the user/system
spaces.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA15404
	for pups-liszt; Thu, 18 Feb 1999 12:32:52 +1100 (EST)

From sms at moe.2bsd.com  Thu Feb 18 11:25:53 1999
From: sms at moe.2bsd.com (Steven M. Schultz)
Date: Wed, 17 Feb 1999 17:25:53 -0800 (PST)
Subject: 2.11BSD, non-split i/d issues
Message-ID: <199902180125.RAA05895@moe.2bsd.com>

Hi -

> From: allisonp at world.std.com (Allison J Parent)
> The F11 does not do I&D split but does have user/system.  

	Correct.  Some systems also have an 18bit only MMU which restricts
	memory to 248kb max (others have a 22bit MMU and can physically
	have more memory).

> It's my understanding that 2.11 will run on F11 systems (pro350 and 11/23)
> if properly configured but the only binaries loose are for split I&D.

	Not likely.  The kernel won't fit in 48kb that I know of.  And there
	will be no networking support since that requires supervisor mode
	which non-split I/D systems don't have.

> So if properly configured you can get 2.11 to utilize the user/system spaces.

	The skeleton of a Makefile for non-split a kernel exists but it 
	will take much work (it is essentially just a list of file that may
	or may not be 100% current) to kick into shape.  Also, remember that
	programs like 'csh', 'vi' and so on are not only split I/D but 
	overlaid - they will not run on a non-split machine. 

	Steven Schultz

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA15443
	for pups-liszt; Thu, 18 Feb 1999 12:44:04 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Thu Feb 18 11:43:27 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Wed, 17 Feb 1999 20:43:27 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <36CAEA1F.D5D7C838@halcyon.com> from "David C. Jenner" at Feb 17, 99 08:11:12 am
Message-ID: <199902180143.UAA01509@math.uwaterloo.ca>

| I don't have any docs on the DECNA, but they must exist.  It's
| probably pretty close to the DEQNA.

The DECNA uses one of the earlier Intel network chips.  It lives
on the CTI bus, a bus like no other.  I believe the DEQNA is T-11
based and lives on the vastly better known Q-bus...  -- Ken

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id FAA18371
	for pups-liszt; Fri, 19 Feb 1999 05:47:10 +1100 (EST)

From kcwellsc at math.uwaterloo.ca  Fri Feb 19 04:46:35 1999
From: kcwellsc at math.uwaterloo.ca (Ken Wellsch)
Date: Thu, 18 Feb 1999 13:46:35 -0500 (EST)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <01BE5B64.56247680@SONAR> from "James Lothian" at Feb 18, 99 01:14:51 pm
Message-ID: <199902181846.NAA05766@math.uwaterloo.ca>

I'm going to give up as I seem to remember nothing anymore... sigh.
Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
over three years and certainly miss being able to pick up the part in
question before foaming at the mouth spouting nonsense. Many apologies
for suggesting such major inaccuracies.  -- Ken

P.S.  Allison describe the DEQNA as a state-driven device with PALs
      (I think) and that "big F" may the the gate array also mentioned.

| From simul8 at simul8.demon.co.uk  Thu Feb 18 12:27:23 1999
| 
| Just for the sake of being picky... the DEQNA is based on an Intel 
| microcontroller chip (something 8085-ish, I think). The ethernet chipset
| seems to be Fairchild (it's certainly got a big F on it.)
| 
| James


From erin at coffee.corliss.net  Fri Feb 19 06:39:35 1999
From: erin at coffee.corliss.net (Erin W. Corliss)
Date: Thu, 18 Feb 1999 12:39:35 -0800 (PST)
Subject: VaxMate
Message-ID: <Pine.LNX.3.96.990218123109.5732A-100000@coffee.corliss.net>


My local computer junk store has a VaxMate for sale.  I'm not sure of the
model -- It has a DB-25 serial port, 10-base-2 ethernet, and a phone-jack
like printer port on the back, as well as an internal ST-225 hard drive
and a 5.25 inch floppy drive.  

Anyway, when I turn it on it tries to boot up -- the graphical slider
thing on the screen gets about 90% of the way across and it displays the
number 83, which I assume is an eeror code since the number changes if you
boot it up with no keyboard.  Anyone know what the 83 means or where I can
get a list of VaxMate error codes?  Also, how intelligent is this machine
compared to a terminal?  Will it actually run a Vax operating system or
does it need a server?

--------------------------------------------------------
"...color flashing thunder crashing dynamite machine..."


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id KAA19267
	for pups-liszt; Fri, 19 Feb 1999 10:25:25 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 09:24:43 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 00:24:43 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902181846.NAA05766@math.uwaterloo.ca>
Message-ID: <Pine.VUL.3.93.990219002351.1502D-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Ken Wellsch wrote:

> I'm going to give up as I seem to remember nothing anymore... sigh.
> Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
> I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
> over three years and certainly miss being able to pick up the part in
> question before foaming at the mouth spouting nonsense. Many apologies
> for suggesting such major inaccuracies.  -- Ken
> 
> P.S.  Allison describe the DEQNA as a state-driven device with PALs
>       (I think) and that "big F" may the the gate array also mentioned.

You might not be totally out. I also thought the DEQNA was T-11 based,
since the DEUNA is. :-)

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19430
	for pups-liszt; Fri, 19 Feb 1999 11:22:47 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 10:22:33 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 19:22:33 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190022.AA25325@world.std.com>

<> I'm going to give up as I seem to remember nothing anymore... sigh.
<> Allison also sent e-mail saying the DEQNA is not T-11 based.  I guess
<> I'm thinking of an RQDX3.  I've had no place to unpack my old iron in
<> over three years and certainly miss being able to pick up the part in
<> question before foaming at the mouth spouting nonsense. Many apologies
<> for suggesting such major inaccuracies.  -- Ken
<> 
<> P.S.  Allison describe the DEQNA as a state-driven device with PALs
<>       (I think) and that "big F" may the the gate array also mentioned.
<
<You might not be totally out. I also thought the DEQNA was T-11 based,
<since the DEUNA is. :-)

I have a DEQNA in front of me.  There is a micro and that is a 8751 8bitter.
The big chip is a LSI ASIC that is a linked list DMA controller.  No t-11.
The RQDXn(n={1,2,3} uses a t-11.  The DELQA also does not use a T-11.
Both use lots of logic in PALs and ASICs to perform several state machines
needed for eithenet.  At the time of development there were few complete
and fast enough chipsets for eithernet.  The DEQNA is mid 80s design and
quite old.

The DEUNA is quite different.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19476
	for pups-liszt; Fri, 19 Feb 1999 11:40:24 +1100 (EST)

From johnh at psych.usyd.edu.au  Fri Feb 19 10:40:17 1999
From: johnh at psych.usyd.edu.au (johnh at psych.usyd.edu.au)
Date: Fri, 19 Feb 1999 11:40:17 +1100 (EST)
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190040.LAA27443@psychwarp.psych.usyd.edu.au>


| Just for the sake of being picky... the DEQNA is based on an Intel
| microcontroller chip (something 8085-ish, I think). The ethernet chipset
| seems to be Fairchild (it's certainly got a big F on it.)
|

The DEQNA uses a Intel 8751 (an EPROM version of 8051 family). I suspect that
it may deal with the programming protocol and the ring buffers. The
chip with the F (with bars top and bottom of the letter) is probably
Fujitsu.

These boards had a fairly bad reputation for lockups and dropped packets.
There was a 20+ wire ECO along with a PAL chip (with 8 of the pins cut off)
soldered on top of another chip.

The replacement ethernet controller was the DELQA, which was a complete
redesign and used a 68000 processor.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA19489
	for pups-liszt; Fri, 19 Feb 1999 11:41:28 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 10:41:03 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 01:41:03 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190022.AA25325@world.std.com>
Message-ID: <Pine.VUL.3.93.990219013735.1502E-100000@Zeke.Update.UU.SE>

Hi, Alison.

> <You might not be totally out. I also thought the DEQNA was T-11 based,
> <since the DEUNA is. :-)
> 
> I have a DEQNA in front of me.  There is a micro and that is a 8751 8bitter.
> The big chip is a LSI ASIC that is a linked list DMA controller.  No t-11.

I'm sorry. I didn't mean to imply that you were wrong, just that I was.

> The RQDXn(n={1,2,3} uses a t-11.  The DELQA also does not use a T-11.

Never looked carefully at RQDX?, but the DELQA uses an M68K, that much I
*do* know. (As do the DELUA)

> Both use lots of logic in PALs and ASICs to perform several state machines
> needed for eithenet.  At the time of development there were few complete
> and fast enough chipsets for eithernet.  The DEQNA is mid 80s design and
> quite old.

You obviously knows more about this than I do. :-)
However, as I said, atleast the DELQA have an M68K...
And the DEQNA is old, yes...

> The DEUNA is quite different.

Obviously. But it is also pretty old. Not as buggy though, which should
have been a clue. :-)

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA19674
	for pups-liszt; Fri, 19 Feb 1999 12:57:56 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 11:57:38 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 20:57:38 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190157.AA29020@world.std.com>

<I'm sorry. I didn't mean to imply that you were wrong, just that I was.

Not an argument, just posting to the group what went private by error.

<Never looked carefully at RQDX?, but the DELQA uses an M68K, that much I
<*do* know. (As do the DELUA)

Having two Qbus VAXen and several Qbus PDP-11s it's old turf.  Also I worked 
for DEC Engineering.  that and I've done a lot of hardware level work on my 
systems (repaired dead boards) so the designs are more familair.

<You obviously knows more about this than I do. :-)
<However, as I said, atleast the DELQA have an M68K...
<And the DEQNA is old, yes...

DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
(less buggy too) of the DEQNA and is largely logically the same as the 
DEQNA.

<> The DEUNA is quite different.
<
<Obviously. But it is also pretty old. Not as buggy though, which should
<have been a clue. :-)

Also The DELUA.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id MAA19690
	for pups-liszt; Fri, 19 Feb 1999 12:59:39 +1100 (EST)

From eekg at ix.netcom.com  Fri Feb 19 11:48:59 1999
From: eekg at ix.netcom.com (Eric Edwards)
Date: Thu, 18 Feb 1999 20:48:59 -0500
Subject: 2.9BSD:  mbuf.h
Message-ID: <006401be5baa$06ce3da0$33d1b7c7@eric-edwards>

I'm not sure if anyone mentioned this, but you can build a working 2.9
kernel (sans network) from the sources by just commenting out the references
to the networking include files.  I think there is an offending reference in
syslocal.c also.

Eric Edwards
eekg at ix.netcom.com
mag at csh.rit.edu

-----Original Message-----
From: maximum entropy <entropy at zippy.bernstein.com>
To: pups at minnie.cs.adfa.edu.au <pups at minnie.cs.adfa.edu.au>
Date: Tuesday, February 16, 1999 11:36 PM
Subject: 2.9BSD: mbuf.h


>"make unix" failed:
>Make:  Don't know how to make /usr/include/sys/mbuf.h.  Stop.



Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19762
	for pups-liszt; Fri, 19 Feb 1999 13:14:46 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 12:14:32 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 21:14:32 -0500
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190214.AA14211@world.std.com>


<The DEQNA uses a Intel 8751 (an EPROM version of 8051 family). I suspect th
<it may deal with the programming protocol and the ring buffers. The
<chip with the F (with bars top and bottom of the letter) is probably
<Fujitsu.

Correct on both cases.  

<These boards had a fairly bad reputation for lockups and dropped packets.
<There was a 20+ wire ECO along with a PAL chip (with 8 of the pins cut off
<soldered on top of another chip.

Actually there were revs A->n and each rev had a step.  The last one was 
N-11... it was marginal.  Good one tended to be good and the bad were PITA.
Also they tended to fail far often than MTBF predictions.

<The replacement ethernet controller was the DELQA, which was a complete
<redesign and used a 68000 processor.

The DELQA was not 68000. The board was far to small for that and had to be 
Qbus dual width and compatable with DEQNA. I have a few of them in my vaxen 
too.  The Unibus versions DEUNA and the later DELUA were 68k and very good. 
They were partly the reason why 730s and 750s were used for routers long 
after they were replaced for other tasks.

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19811
	for pups-liszt; Fri, 19 Feb 1999 13:23:12 +1100 (EST)

From allisonp at world.std.com  Fri Feb 19 12:22:56 1999
From: allisonp at world.std.com (Allison J Parent)
Date: Thu, 18 Feb 1999 21:22:56 -0500
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
Message-ID: <199902190222.AA21412@world.std.com>

<DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
<(less buggy too) of the DEQNA and is largely logically the same as the 
<DEQNA.

Memory parity exception...  Eat foot time.

DELQA M7516 is 68k and lance chip... had to pull mine to check.  The M7504
however I am correct as I pulled one down from the shelf before dining on 
foot.  Oh and the reson I forgot it's 68k, was the DELQA is far more 
reliable!  that and I only open the BA123 one a year to check the fans and 
clean dust.  It just don't break. ;)

Allison


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA19969
	for pups-liszt; Fri, 19 Feb 1999 13:36:31 +1100 (EST)

From SHOPPA at trailing-edge.com  Fri Feb 19 12:36:11 1999
From: SHOPPA at trailing-edge.com (Tim Shoppa)
Date: Thu, 18 Feb 1999 21:36:11 -0500
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
Message-ID: <990218213611.202000b8@trailing-edge.com>

><The replacement ethernet controller was the DELQA, which was a complete
><redesign and used a 68000 processor.
>The DELQA was not 68000.

Hate to turn this into a "no it isn't, yet it is" sequence, but all
my DELQA's have prominent 68000's on 'em.

> The board was far to small for that

No, it isn't.  The 68000 is the quad pack, and is smaller than either
of the two custom gate arrays that does the Q-bus handshaking.

Tim.

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id OAA20047
	for pups-liszt; Fri, 19 Feb 1999 14:07:20 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Fri Feb 19 13:06:14 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Fri, 19 Feb 1999 14:06:14 +1100 (EST)
Subject: 2.9BSD:  mbuf.h
In-Reply-To: <006401be5baa$06ce3da0$33d1b7c7@eric-edwards> from Eric Edwards at "Feb 18, 1999  8:48:59 pm"
Message-ID: <199902190306.OAA02230@henry.cs.adfa.edu.au>

In article by Eric Edwards:
> I'm not sure if anyone mentioned this, but you can build a working 2.9
> kernel (sans network) from the sources by just commenting out the references
> to the networking include files.  I think there is an offending reference in
> syslocal.c also.
> 
> Eric Edwards
> eekg at ix.netcom.com
> mag at csh.rit.edu

What is happening is that `make depend' invokes a script which finds
#includes in the source code, and builds a make dependency. However,
it's not very intelligent, and doesn't ignore:

#ifdef INET
#include <stuff>

when INET isn't defined. :-) This bites on several C files.
You just have to hand-prune the Makefile after make depend :-)
This is 2.9BSD, BTW, ignore if you're not using it.

Ciao!
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id VAA21635
	for pups-liszt; Fri, 19 Feb 1999 21:20:30 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 20:19:31 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 11:19:31 +0100 (MET)
Subject: Venix (was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190157.AA29020@world.std.com>
Message-ID: <Pine.VUL.3.93.990219111730.26158A-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Allison J Parent wrote:

> <You obviously knows more about this than I do. :-)
> <However, as I said, atleast the DELQA have an M68K...
> <And the DEQNA is old, yes...
> 
> DELQA is not 68k, The DEUNA is.  The DELQA is a cost reduced version 
> (less buggy too) of the DEQNA and is largely logically the same as the 
> DEQNA.

Really? I have a DELQA sitting right in front of me, and when I look at
it, the large chip definitely says M68000. What could that be then?

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id VAA21652
	for pups-liszt; Fri, 19 Feb 1999 21:23:04 +1100 (EST)

From bqt at Update.UU.SE  Fri Feb 19 20:22:35 1999
From: bqt at Update.UU.SE (Johnny Billquist)
Date: Fri, 19 Feb 1999 11:22:35 +0100 (MET)
Subject: DEQNA (was was Re: 2.9BSD:  mbuf.h)
In-Reply-To: <199902190214.AA14211@world.std.com>
Message-ID: <Pine.VUL.3.93.990219112030.26158B-100000@Zeke.Update.UU.SE>

On Thu, 18 Feb 1999, Allison J Parent wrote:

> <The replacement ethernet controller was the DELQA, which was a complete
> <redesign and used a 68000 processor.
> 
> The DELQA was not 68000. The board was far to small for that and had to be 
> Qbus dual width and compatable with DEQNA. I have a few of them in my vaxen 
> too.  The Unibus versions DEUNA and the later DELUA were 68k and very good. 

Hate to disagree with you, Alison. The the DELQA really is 68000, take a
peek inside yourself. It is a dual-width too...

And the DEUNA is T-11, while the DELUA is 68000.
I have never bothered plugging in any DEUNAs myself, since DELUAs are
pretty common, and they atleast are pretty good. Never had any problems
with any of them.

	Johnny

Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email: bqt at update.uu.se           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol



From g4klx at pop.agri.ch  Sun Feb 21 04:10:45 1999
From: g4klx at pop.agri.ch (Jonathan Naylor)
Date: Sat, 20 Feb 1999 19:10:45 +0100 (CET)
Subject: V7 filesystem work
Message-ID: <Pine.LNX.4.05.9902201906360.1385-100000@g4klx.agri.ch>

Hello All

A couple of weeks ago I hacked the program v7 from the bostic_tools to
work under all sorts of different Unix versions. It worked great and
allowed me to snoop around the V7 file system images from native Linux.
Anyone who wants a copy can send me an e-mail.

Anyway I had a few hours spare today, and decided to try adding the V7
filesystem to the Linux kernel. Results so far are encouraging:


g4klx:/usr/src/linux# ls -l /mnt
total 333
drwxrwxrwx   7 root     root          224 Sep 22  1988 .
drwxr-xr-x  19 root     root         1024 Feb 14 11:55 ..
drwxrwxr-x   2 3        3            2512 Sep 22  1988 bin
-rwxr-xr-x   1 3        3            8986 Jun  8  1979 boot
drwxrwxr-x   2 3        3             160 Sep 22  1988 dev
drwxrwxr-x   2 3        3             336 Sep 22  1988 etc
-rwxr-xr-x   1 daemon   daemon      53302 Jun  8  1979 hphtunix
-rwxr-xr-x   1 daemon   daemon      52850 Jun  8  1979 hptmunix
drwxrwxr-x   2 3        3             192 Sep 22  1988 lib
drwxrwxr-x   2 root     lp             96 Sep 22  1988 mdec
-rwxr-xr-x   1 root     daemon      50990 Jun  8  1979 rkunix
-rwxr-xr-x   1 root     daemon      51982 Jun  8  1979 rl2unix
-rwxr-xr-x   1 daemon   daemon      51790 Jun  8  1979 rphtunix
-rwxr-xr-x   1 daemon   daemon      51274 Jun  8  1979 rptmunix
g4klx:/usr/src/linux# df
Filesystem         1024-blocks  Used Available Capacity Mounted on
/dev/hda1            3031184 1920771   953665     67%   /
/dev/loop0              1919    1877       42     98%   /mnt
g4klx:/usr/src/linux#


I am using the loop block device to allow me to mount a file as a block
device, this saves me having to add a new partition to my disc. There
should be no reason why it won't work with a true disc partition. The V7
filesystem under Linux is read/write.

Anyone interested ?

Jonathan




From mallison at konnections.com  Sun Feb 21 12:13:38 1999
From: mallison at konnections.com (Mike Allison)
Date: Sat, 20 Feb 1999 19:13:38 -0700
Subject: V7 filesystem work
Message-ID: <199902210203.TAA18682@mail.konnections.com>

Yeah, I'm interested.  Can you write up what changes the linux port entailed???

-Mike


At 07:10 PM 2/20/99 +0100, g4klx at g4klx.demon.co.uk wrote:
>Hello All
>
>A couple of weeks ago I hacked the program v7 from the bostic_tools to
>work under all sorts of different Unix versions. It worked great and
>allowed me to snoop around the V7 file system images from native Linux.
>Anyone who wants a copy can send me an e-mail.
>
>Anyway I had a few hours spare today, and decided to try adding the V7
>filesystem to the Linux kernel. Results so far are encouraging:
>
>
>g4klx:/usr/src/linux# ls -l /mnt
>total 333
>drwxrwxrwx   7 root     root          224 Sep 22  1988 .
>drwxr-xr-x  19 root     root         1024 Feb 14 11:55 ..
>drwxrwxr-x   2 3        3            2512 Sep 22  1988 bin
>-rwxr-xr-x   1 3        3            8986 Jun  8  1979 boot
>drwxrwxr-x   2 3        3             160 Sep 22  1988 dev
>drwxrwxr-x   2 3        3             336 Sep 22  1988 etc
>-rwxr-xr-x   1 daemon   daemon      53302 Jun  8  1979 hphtunix
>-rwxr-xr-x   1 daemon   daemon      52850 Jun  8  1979 hptmunix
>drwxrwxr-x   2 3        3             192 Sep 22  1988 lib
>drwxrwxr-x   2 root     lp             96 Sep 22  1988 mdec
>-rwxr-xr-x   1 root     daemon      50990 Jun  8  1979 rkunix
>-rwxr-xr-x   1 root     daemon      51982 Jun  8  1979 rl2unix
>-rwxr-xr-x   1 daemon   daemon      51790 Jun  8  1979 rphtunix
>-rwxr-xr-x   1 daemon   daemon      51274 Jun  8  1979 rptmunix
>g4klx:/usr/src/linux# df
>Filesystem         1024-blocks  Used Available Capacity Mounted on
>/dev/hda1            3031184 1920771   953665     67%   /
>/dev/loop0              1919    1877       42     98%   /mnt
>g4klx:/usr/src/linux#
>
>
>I am using the loop block device to allow me to mount a file as a block
>device, this saves me having to add a new partition to my disc. There
>should be no reason why it won't work with a true disc partition. The V7
>filesystem under Linux is read/write.
>
>Anyone interested ?
>
>Jonathan
>
>
>
>


Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id NAA02555
	for pups-liszt; Sun, 21 Feb 1999 13:29:44 +1100 (EST)

From wkt at henry.cs.adfa.edu.au  Sun Feb 21 12:29:29 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Sun, 21 Feb 1999 13:29:29 +1100 (EST)
Subject: V7 filesystem work
In-Reply-To: <Pine.LNX.4.05.9902201906360.1385-100000@g4klx.agri.ch> from Jonathan Naylor at "Feb 20, 1999  7:10:45 pm"
Message-ID: <199902210229.NAA08687@henry.cs.adfa.edu.au>

In article by Jonathan Naylor:
> Hello All
> 
> A couple of weeks ago I hacked the program v7 from the bostic_tools to
> work under all sorts of different Unix versions. It worked great and
> allowed me to snoop around the V7 file system images from native Linux.
> Anyone who wants a copy can send me an e-mail.
> 
> Anyway I had a few hours spare today, and decided to try adding the V7
> filesystem to the Linux kernel. Results so far are encouraging:

> I am using the loop block device to allow me to mount a file as a block
> device, this saves me having to add a new partition to my disc. There
> should be no reason why it won't work with a true disc partition. The V7
> filesystem under Linux is read/write.
> 
> Anyone interested ?

I'd be happy to add any changes etc. into the Tools directory in the PUPS
Archive.

It's about time Unix could read the Unix filesystem again :-)

Ciao,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id UAA05201
	for pups-liszt; Sun, 21 Feb 1999 20:23:41 +1100 (EST)

From g4klx at pop.agri.ch  Sun Feb 21 19:07:50 1999
From: g4klx at pop.agri.ch (Jonathan Naylor)
Date: Sun, 21 Feb 1999 10:07:50 +0100 (CET)
Subject: V7 filesystem work
In-Reply-To: <199902210203.TAA18682@mail.konnections.com>
Message-ID: <Pine.LNX.4.05.9902210957580.303-100000@g4klx.agri.ch>

Hello Mike and the list

On Sat, 20 Feb 1999, Mike Allison wrote:
> Yeah, I'm interested.  Can you write up what changes the linux port entailed???
> 
> -Mike

I assume you mean the standalone V7 FS program rather than the kernel V7
FS support ?

The code was written in old C, and from a modern C programmers point of
view, rather sloppily. The warnings from the compiler were terrible, so I
added function prototypes, and made the code more ANSI C like. Then I got
rid of a few bugs, in one place I remember a character pointer being
assigned to a character.

I then typedef'd the data types so I could use int8, int16 and int32 in
the code to make it more portable. I stopped using structure overlays onto
the raw data as that is messy and is not good for (a) byte ordering and
(b) structure packing. It also allowed me to stop using the original V7
file headers which would have made a public release of the code
problematic.

The data is extracted from the raw block data by using special
architecturally neutral functions into locally held structures. That is a
particular win with the block number in three bytes trick that is used in
the inode.

It has been tested on i386/Linux with both glibc 1.0 and glibc 2.0 and
Alpha/Linux, no changes were needed.

Then I added a few new commands to let me look at the superblock and
bootblocks and a few other bits.

Then I released it.

I have just sent a copy of the program to Warren for inclusion in the PUPS
tools section. Its not very big.

Work is progressing on the V7 filesystem in the Linux kernel. Anyone who
wants the patches for that should send me an e-mail. I hope to get it into
the mainstream kernel in the Linux 2.3 series.

Jonathan



From wkt at henry.cs.adfa.edu.au  Tue Feb 23 15:25:51 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Tue, 23 Feb 1999 16:25:51 +1100 (EST)
Subject: 2.9BSD & Pro: another version
Message-ID: <199902230525.QAA00483@henry.cs.adfa.edu.au>

Ken Wellsch has just uploaded a set of RX50 disk images containing
2.9BSD for the Pro 350 to the PUPS Archive. You can find them in

	Distributions/ucb/2.9bsd4pro350-kcwellsc

He says:

	I believe the RX50 is actually 80 tracks with 10 sectors per track,
	thus yielding 800 blocks per disk.  I think the first track is
	reserved and thus Venix would not let me at it.  Hopefully I have
	not also lost additional information here too.

All the 34 disk images he sent in are 790 blocks long. Can anybody
tell us if we will need to recover track 0 to make these images useful?

At the very least, I've managed to find the pcreg.h file out of the
images (cat */*.rx50 | less -B), so I'm getting closer at recompiling
the 2.9/Pro kernel.

Cheers,
	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id DAA23748
	for pups-liszt; Wed, 24 Feb 1999 03:01:12 +1100 (EST)

From jesper at df.lth.se  Wed Feb 24 02:00:54 1999
From: jesper at df.lth.se (Jesper Nilsson)
Date: Tue, 23 Feb 1999 17:00:54 +0100 (CET)
Subject: SCO Source license tainting?
Message-ID: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se>

Tjo!

I have a question that I hope someone can help me with:

I'm thinking about getting myself a SCO Source license,
but I'm worried that I might get "tainted" by this
since my day job involves writing operating systems...
My employer would not appreciate getting sued because of
my hobbies...:-)

Has anyone done research about this aspect of the license?

My goals are twofold, running an older Unix version on my PDP-11's,
and of course I want to peruse the source of the classic versions.

/^JN - Jesper Nilsson
--
              I've heard of UNIseX, but I've never had it.
                  Jesper Nilsson -- jesper at df.lth.se



From wkt at henry.cs.adfa.edu.au  Wed Feb 24 07:52:15 1999
From: wkt at henry.cs.adfa.edu.au (Warren Toomey)
Date: Wed, 24 Feb 1999 08:52:15 +1100 (EST)
Subject: SCO Source license tainting?
In-Reply-To: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se> from Jesper Nilsson at "Feb 23, 1999  5: 0:54 pm"
Message-ID: <199902232152.IAA04722@henry.cs.adfa.edu.au>

In article by Jesper Nilsson:
> I'm thinking about getting myself a SCO Source license,
> but I'm worried that I might get "tainted" by this
> since my day job involves writing operating systems...
> My employer would not appreciate getting sued because of
> my hobbies...:-)

As long as you don't reuse tainted Unix source code in your job, you will
be ok. There are so many books covering the Unix kernel: Lions, Bach,
Goodheart, Vahalia etc., that any concerns other than source code reuse
are negligible.

That's my feelings, anyway.

	Warren

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.1/8.9.1) id LAA25583
	for pups-liszt; Wed, 24 Feb 1999 11:46:21 +1100 (EST)

From grog at lemis.com  Wed Feb 24 10:46:07 1999
From: grog at lemis.com (Greg Lehey)
Date: Wed, 24 Feb 1999 11:16:07 +1030
Subject: SCO Source license tainting?
In-Reply-To: <199902232152.IAA04722@henry.cs.adfa.edu.au>; from Warren Toomey on Wed, Feb 24, 1999 at 08:52:15AM +1100
References: <Pine.LNX.4.05-df.9902231646390.8684-100000@bartlet.df.lth.se> <199902232152.IAA04722@henry.cs.adfa.edu.au>
Message-ID: <19990224111607.G93492@lemis.com>

On Wednesday, 24 February 1999 at  8:52:15 +1100, Warren Toomey wrote:
> In article by Jesper Nilsson:
>> I'm thinking about getting myself a SCO Source license,
>> but I'm worried that I might get "tainted" by this
>> since my day job involves writing operating systems...
>> My employer would not appreciate getting sued because of
>> my hobbies...:-)
>
> As long as you don't reuse tainted Unix source code in your job, you will
> be ok. There are so many books covering the Unix kernel: Lions, Bach,
> Goodheart, Vahalia etc., that any concerns other than source code reuse
> are negligible.

I think this relates to a spectre raised during the USL/BSDI wars.
Somebody suggested that anybody who had been exposed to AT&T source
code was ``tainted'' and could thus not legally develop competitive
systems.  Somewhere I have a button that somebody brought back to me
from a USENIX, with the text ``mentally contaminated''.

Jesper, I don't think you need to worry about the problem.  That kind
of restriction would be unenforceable.

Greg
--
See complete headers for address, home page and phone numbers
finger grog at lemis.com for PGP public key


