From michael_davidson at pacbell.net  Wed Jan  1 02:02:37 2003
From: michael_davidson at pacbell.net (Michael Davidson)
Date: Tue, 31 Dec 2002 08:02:37 -0800
Subject: [pups] V7 setjmp/longjmp
References: <200212311307.gBVD7p600218@bsd.korb>
Message-ID: <3E11BF9D.5010903@pacbell.net>

Wolfgang Helbig wrote:

>Michael Davidson distrusted:
>
>>Note that this means that r4, r3 and r2 get restored to the values
>>that they had at the time that the next stack frame was established
>>(ie when the "next" function down was called from the original context)
>>and that these are not necessarily the values that they had at the
>>moment when the original call to setjnp returned. Hence the various
>>caveats about not relying on the value of any register variables
>>after returning via longjmp.
>>
>
>You can trust register variables when setjmp() returns the second time. Then
>the registers are restored to the values they had when the "next" function was 
>called, that is the "values as of the time longjmp() was called" (quoted from
>longjmp(3)'s man page. Thus any variable behaves the same, regardless of its
>storage class.
>
Yes, you are right - V7 restores the register variables to a state which
is consistent with the other auto variables in the function - ie the value
which they had when longjmp was called.

The caveats about not relying on register variables applied to V6.




From rlhamil at mindwarp.smart.net  Mon Jan  6 23:55:01 2003
From: rlhamil at mindwarp.smart.net (Richard L. Hamilton)
Date: Mon, 6 Jan 2003 08:55:01 -0500 (EST)
Subject: [pups] V7 RL02 driver questions
Message-ID: <200301061355.h06Dt1601978@mindwarp.smart.net>

The files

http://minnie.tuhs.org/UnixTree/V7/usr/sys/dev/rl.c
http://minnie.tuhs.org/UnixTree/V7/usr/sys/conf/mkconf.c

don't seem to match - mkconf.c wants to generate references to
rlopen() and rlclose(), which don't exist in that version of rl.c.

What I'm really looking for is the right version of driver(s) etc.
so that I can regenerate the rl2unix kernel in the V7 image
available for use with simh, so that I can add or tweak drivers, eventually
adding the dz driver so I can run multiple "terminals" with telnet access.

-- 
mailto:rlhamil at mindwarp.smart.net  http://www.smart.net/~rlhamil



From helbig at Informatik.BA-Stuttgart.DE  Tue Jan  7 04:01:45 2003
From: helbig at Informatik.BA-Stuttgart.DE (Wolfgang Helbig)
Date: Mon, 6 Jan 2003 19:01:45 +0100 (MET)
Subject: [pups] Mail delivery failed: returning message to sender
Message-ID: <200301062005.h06K5ah00493@bsd.korb>

Richard L. Hamilton searches:
>The files
>
>http://minnie.tuhs.org/UnixTree/V7/usr/sys/dev/rl.c
>http://minnie.tuhs.org/UnixTree/V7/usr/sys/conf/mkconf.c
>
>don't seem to match - mkconf.c wants to generate references to
>rlopen() and rlclose(), which don't exist in that version of rl.
>What I'm really looking for is the right version of driver(s) etc.

It seems that a version of the rl driver that matches
mkconf, was never supplied by the bell labs.

As a workaround delete the rl entry in cdevsw and the line
"int rlopen,..." from c.c. This will give you a block device of the rl disk,
but no raw device. 

Greetings,

Wolfgang



From pete at dunnington.u-net.com  Tue Jan  7 08:28:09 2003
From: pete at dunnington.u-net.com (pete at dunnington.u-net.com)
Date: Mon, 6 Jan 2003 22:28:09 GMT
Subject: [pups] Mail delivery failed: returning message to sender
In-Reply-To: Wolfgang Helbig <helbig@Informatik.BA-Stuttgart.DE>
        "[pups] Mail delivery failed: returning message to sender" (Jan  6, 19:01)
References: <200301062005.h06K5ah00493@bsd.korb>
Message-ID: <10301062228.ZM25510@mindy.dunnington.u-net.com>

On Jan 6, 19:01, Wolfgang Helbig wrote:
> Richard L. Hamilton searches:
> >The files
> >
> >http://minnie.tuhs.org/UnixTree/V7/usr/sys/dev/rl.c
> >http://minnie.tuhs.org/UnixTree/V7/usr/sys/conf/mkconf.c
> >
> >don't seem to match - mkconf.c wants to generate references to
> >rlopen() and rlclose(), which don't exist in that version of rl.
> >What I'm really looking for is the right version of driver(s) etc.
>
> It seems that a version of the rl driver that matches
> mkconf, was never supplied by the bell labs.
>
> As a workaround delete the rl entry in cdevsw and the line
> "int rlopen,..." from c.c. This will give you a block device of the rl
disk,
> but no raw device.

Hmm... my 7th Edition distribution  /usr/sys/dev/rl.c  certainly *does*
have rlopen() and rlclose().  If you want them, I can send rl.c and the
matching mkconf.c  to you, or to Warren, along with the rl1conf and rl2conf
(for RL01 and RL02 disks) from  /usr/sys/conf/  and any other files needed
(if someone can tell me what might be required -- it's a long long time
since I looked at any of this.   The distribution came for an 11/23 with 2
x RL02, and it also has some special source files for 11/23, 11/34, etc
("small machines") including tests for recovery from segmentation traps in
floating point (which isn't the same on all PDP-11 models), and an RK
driver from the Boston Children's Museum which handles overlapped seeks on
multiple drives.

I'll be away for a couple of days so don't expect a prompt reply, but the
source is online here so I can easily copy a few files when I get back.

-- 
Pete						Peter Turnbull
						Network Manager
						University of York


From dfevans at bbcr.uwaterloo.ca  Thu Jan 16 06:48:31 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Wed, 15 Jan 2003 15:48:31 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
Message-ID: <20030115154831.A9065@bcr10.uwaterloo.ca>

  Howdy.

  I'd like to confirm that somebody has successfully made a TK50 boot tape
for 2.11BSD using a TK50Z attached to some other Unix box.  In my case it will
likely have to be Linux, though IRIX might be doable.  It *should* work, after
all, but, well, you know.  :)

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From jkunz at unixag-kl.fh-kl.de  Thu Jan 16 07:50:31 2003
From: jkunz at unixag-kl.fh-kl.de (Jochen Kunz)
Date: Wed, 15 Jan 2003 22:50:31 +0100
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030115154831.A9065@bcr10.uwaterloo.ca>; from dfevans@bbcr.uwaterloo.ca on Wed, Jan 15, 2003 at 21:48:31 CET
References: <20030115154831.A9065@bcr10.uwaterloo.ca>
Message-ID: <20030115225031.S57315@MissSophie.unixag-kl.fh-kl.de>

On 2003.01.15 21:48 David Evans wrote:

>   I'd like to confirm that somebody has successfully made a TK50 boot
> tape for 2.11BSD using a TK50Z attached to some other Unix box.  
Well, the TK50z sems to be rather quirky. I had not much luck with it.
The TZ30 is not much better. You don't happen to have a QBus VAX? I used
a MicroVAX 4000-200 with NetBSD to write the 2.11BSD tape. 

Use the maketape programm from the archive. dd will not work proper. 
-- 



tschüß,
         Jochen

Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/


From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 00:40:37 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 09:40:37 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030115225031.S57315@MissSophie.unixag-kl.fh-kl.de>; from jkunz@unixag-kl.fh-kl.de on Wed, Jan 15, 2003 at 10:50:31PM +0100
References: <20030115154831.A9065@bcr10.uwaterloo.ca> <20030115225031.S57315@MissSophie.unixag-kl.fh-kl.de>
Message-ID: <20030116094037.B28194@bcr10.uwaterloo.ca>

On Wed, Jan 15, 2003 at 10:50:31PM +0100, Jochen Kunz wrote:
> Well, the TK50z sems to be rather quirky.

  That's a kind way to describe it.  I made a tape last night using
"maketape" and it didn't boot.

> You don't happen to have a QBus VAX? I used
> a MicroVAX 4000-200 with NetBSD to write the 2.11BSD tape. 
> 

  I have a KA650 boardset.  It's been misbehaving, but I haven't thorougly
explored why.  Given the troubles with the TK50Z, though, I'll likely invest
some time into troubleshooting it.

> Use the maketape programm from the archive. dd will not work proper. 

  That's truly weird.

  Anyway, thanks.  :)

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From luvisi at andru.sonoma.edu  Fri Jan 17 01:53:02 2003
From: luvisi at andru.sonoma.edu (Andru Luvisi)
Date: Thu, 16 Jan 2003 07:53:02 -0800 (PST)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030116094037.B28194@bcr10.uwaterloo.ca>
Message-ID: <Pine.LNX.4.44.0301160744570.11570-100000@gladen>

On Thu, 16 Jan 2003, David Evans wrote:
[snip]
> > Use the maketape programm from the archive. dd will not work proper.
>
>   That's truly weird.

maketape will pad each of "mtboot", "mtboot", and "boot" out to an even
length of 512 bytes.  The "cat | dd" method will not pad the first two
out.

I discovered this while trying to create a bootable tape image for simh.
I needed to emulate the maketape behavior for it to work.

Andru
-- 
Andru Luvisi, Programmer/Analyst

Quote Of The Moment:
  Quidquid latine dictum sit, altum viditur.
  ( Whatever is said in Latin sounds profound. )



From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 02:06:43 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 11:06:43 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.44.0301160744570.11570-100000@gladen>; from luvisi@andru.sonoma.edu on Thu, Jan 16, 2003 at 07:53:02AM -0800
References: <20030116094037.B28194@bcr10.uwaterloo.ca> <Pine.LNX.4.44.0301160744570.11570-100000@gladen>
Message-ID: <20030116110643.B28361@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 07:53:02AM -0800, Andru Luvisi wrote:
> maketape will pad each of "mtboot", "mtboot", and "boot" out to an even
> length of 512 bytes.  The "cat | dd" method will not pad the first two
> out.
> 

  Ahhh--of course.  One would have to do a series of "dd"s in order to
get the padding happening.

> I discovered this while trying to create a bootable tape image for simh.
> I needed to emulate the maketape behavior for it to work.
> 

  That's not a bad idea.  Perhaps I should just make a minimal bootable image
using simh or such, dd it to a SCSI disk, and then attach that to my 11/73.
Might be easier than endless futzing with the TK50.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From luvisi at andru.sonoma.edu  Fri Jan 17 03:34:16 2003
From: luvisi at andru.sonoma.edu (Andru Luvisi)
Date: Thu, 16 Jan 2003 09:34:16 -0800 (PST)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030116110643.B28361@bcr10.uwaterloo.ca>
Message-ID: <Pine.LNX.4.44.0301160932070.11570-100000@gladen>

On Thu, 16 Jan 2003, David Evans wrote:
> On Thu, Jan 16, 2003 at 07:53:02AM -0800, Andru Luvisi wrote:
> > maketape will pad each of "mtboot", "mtboot", and "boot" out to an even
> > length of 512 bytes.  The "cat | dd" method will not pad the first two
> > out.
> >
>
>   Ahhh--of course.  One would have to do a series of "dd"s in order to
> get the padding happening.

But then they would end up in separate tape files.  :-(

> > I discovered this while trying to create a bootable tape image for simh.
> > I needed to emulate the maketape behavior for it to work.
> >
>
>   That's not a bad idea.  Perhaps I should just make a minimal bootable image
> using simh or such, dd it to a SCSI disk, and then attach that to my 11/73.
> Might be easier than endless futzing with the TK50.

I already made one.  It's part of
http://www.tuhs.org/Archive/PDP-11/Boot_Images/2.11_on_Simh/211bsd.tar.gz
along with the Perl script I wrote to make it.

Andru
-- 
Andru Luvisi, Programmer/Analyst

Quote Of The Moment:
  I'm not normal.  I know it.  I don't care!
       - Ace Of Base



From bqt at update.uu.se  Fri Jan 17 03:28:30 2003
From: bqt at update.uu.se (Johnny Billquist)
Date: Thu, 16 Jan 2003 18:28:30 +0100 (CET)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030116110643.B28361@bcr10.uwaterloo.ca>
Message-ID: <Pine.LNX.4.21.0301161819350.6829-100000@Tempo.Update.UU.SE>

On Thu, 16 Jan 2003, David Evans wrote:

> On Thu, Jan 16, 2003 at 07:53:02AM -0800, Andru Luvisi wrote:
> > maketape will pad each of "mtboot", "mtboot", and "boot" out to an even
> > length of 512 bytes.  The "cat | dd" method will not pad the first two
> > out.
> 
>   Ahhh--of course.  One would have to do a series of "dd"s in order to
> get the padding happening.

Probably not. Repeated "dd"s will not produce the same result. This is the
typical error when not understanding that tapes actually are not a stream
of characters. (And also a good example when the Unix paradigm fails 
pretty bad).
Each write to the tape device will create a block. You cannot do two
writes after each other to extend the block. You will instead get two
different blocks.

Boot tapes for VAXen (and PDP-11s) are expected to have 512 byte
blocks. Actually, the first block should be repeated twice, and this is
the primary boot block. After that you can put whatever you want. Most
primary boot blocks however also expects to read 512 byte blocks, which
means that the next file (secondary booter or the system itself) also most
often is written with 512 byte blocks.

> > I discovered this while trying to create a bootable tape image for simh.
> > I needed to emulate the maketape behavior for it to work.
> 
>   That's not a bad idea.  Perhaps I should just make a minimal bootable image
> using simh or such, dd it to a SCSI disk, and then attach that to my 11/73.
> Might be easier than endless futzing with the TK50.

Don't. The moment you copy a tape file to disk, you will loose meta
information unless you use a program specifically designed to preserve
that information.

	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 dfevans at bbcr.uwaterloo.ca  Fri Jan 17 03:30:51 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 12:30:51 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.44.0301160932070.11570-100000@gladen>; from luvisi@andru.sonoma.edu on Thu, Jan 16, 2003 at 09:34:16AM -0800
References: <20030116110643.B28361@bcr10.uwaterloo.ca> <Pine.LNX.4.44.0301160932070.11570-100000@gladen>
Message-ID: <20030116123051.D28733@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 09:34:16AM -0800, Andru Luvisi wrote:
> On Thu, 16 Jan 2003, David Evans wrote:
> > On Thu, Jan 16, 2003 at 07:53:02AM -0800, Andru Luvisi wrote:
> > > maketape will pad each of "mtboot", "mtboot", and "boot" out to an even
> > > length of 512 bytes.  The "cat | dd" method will not pad the first two
> > > out.
> > >
> >
> >   Ahhh--of course.  One would have to do a series of "dd"s in order to
> > get the padding happening.
> 
> But then they would end up in separate tape files.  :-(
> 

  Ahh--I was asleep at the wheel and forgot that there's no convenient way
to concatenate the files using dd.  Duh.

> I already made one.  It's part of
> http://www.tuhs.org/Archive/PDP-11/Boot_Images/2.11_on_Simh/211bsd.tar.gz
> along with the Perl script I wrote to make it.
> 

  Oh yeah--I saw that but forgot about it.

  Just not my day.  :P

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 03:33:54 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 12:33:54 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.21.0301161819350.6829-100000@Tempo.Update.UU.SE>; from bqt@update.uu.se on Thu, Jan 16, 2003 at 06:28:30PM +0100
References: <20030116110643.B28361@bcr10.uwaterloo.ca> <Pine.LNX.4.21.0301161819350.6829-100000@Tempo.Update.UU.SE>
Message-ID: <20030116123354.E28733@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 06:28:30PM +0100, Johnny Billquist wrote:
> >   Ahhh--of course.  One would have to do a series of "dd"s in order to
> > get the padding happening.
> 
> Probably not. Repeated "dd"s will not produce the same result. This is the
> typical error when not understanding that tapes actually are not a stream
> of characters. (And also a good example when the Unix paradigm fails 
> pretty bad).

  If you could do something like

(dd if=mtboot sync bs=512 |+ dd if=mtboot sync bs=512 |+ dd if=boot sync bs=512) | dd of=/dev/whatever obs=512

in order to concatenate the multiple copies of mtboot then all would be well.

> >   That's not a bad idea.  Perhaps I should just make a minimal bootable image
> > using simh or such, dd it to a SCSI disk, and then attach that to my 11/73.
> > Might be easier than endless futzing with the TK50.
> 
> Don't. The moment you copy a tape file to disk, you will loose meta
> information unless you use a program specifically designed to preserve
> that information.
> 

  I meant an actual bootable system, not the tape image.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From luvisi at andru.sonoma.edu  Fri Jan 17 03:42:14 2003
From: luvisi at andru.sonoma.edu (Andru Luvisi)
Date: Thu, 16 Jan 2003 09:42:14 -0800 (PST)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.21.0301161819350.6829-100000@Tempo.Update.UU.SE>
Message-ID: <Pine.LNX.4.44.0301160938310.11570-100000@gladen>

On Thu, 16 Jan 2003, Johnny Billquist wrote:
> On Thu, 16 Jan 2003, David Evans wrote:
[snip]
> >   That's not a bad idea.  Perhaps I should just make a minimal
> > bootable image using simh or such, dd it to a SCSI disk, and then
> > attach that to my 11/73. Might be easier than endless futzing with the
> > TK50.
>
> Don't. The moment you copy a tape file to disk, you will loose meta
> information unless you use a program specifically designed to preserve
> that information.

I misread David's post the first time.  I thought he was thinking of doing
an installation on simh (using a bootable tape image) and then dd'ing the
disk image onto a SCSI disk to boot the real PDP-11 from.  Is there any
reason this wouldn't work?

Andru
-- 
Andru Luvisi, Programmer/Analyst

Quote Of The Moment:
  If you tell the truth you don't have to remember anything.
                  -- Mark Twain



From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 03:45:11 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 12:45:11 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.21.0301161838280.6829-100000@Tempo.Update.UU.SE>; from bqt@update.uu.se on Thu, Jan 16, 2003 at 06:40:24PM +0100
References: <Pine.LNX.4.44.0301160938310.11570-100000@gladen> <Pine.LNX.4.21.0301161838280.6829-100000@Tempo.Update.UU.SE>
Message-ID: <20030116124511.C29094@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 06:40:24PM +0100, Johnny Billquist wrote:
> 
> Ah. Maybe I misunderstood things. Copying a bootable disk image should
> work just fine. I thought he wanted the tape image moved over.
> 
 
  No, bootable disk is what I was after.  The tape image idea floated into my
brain for a moment, but it didn't last long since I'm certain it would not
work.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From bqt at update.uu.se  Fri Jan 17 03:40:24 2003
From: bqt at update.uu.se (Johnny Billquist)
Date: Thu, 16 Jan 2003 18:40:24 +0100 (CET)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.44.0301160938310.11570-100000@gladen>
Message-ID: <Pine.LNX.4.21.0301161838280.6829-100000@Tempo.Update.UU.SE>

On Thu, 16 Jan 2003, Andru Luvisi wrote:

> On Thu, 16 Jan 2003, Johnny Billquist wrote:
> > On Thu, 16 Jan 2003, David Evans wrote:
> [snip]
> > >   That's not a bad idea.  Perhaps I should just make a minimal
> > > bootable image using simh or such, dd it to a SCSI disk, and then
> > > attach that to my 11/73. Might be easier than endless futzing with the
> > > TK50.
> >
> > Don't. The moment you copy a tape file to disk, you will loose meta
> > information unless you use a program specifically designed to preserve
> > that information.
> 
> I misread David's post the first time.  I thought he was thinking of doing
> an installation on simh (using a bootable tape image) and then dd'ing the
> disk image onto a SCSI disk to boot the real PDP-11 from.  Is there any
> reason this wouldn't work?

Ah. Maybe I misunderstood things. Copying a bootable disk image should
work just fine. I thought he wanted the tape image moved over.

	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 luvisi at andru.sonoma.edu  Fri Jan 17 04:01:35 2003
From: luvisi at andru.sonoma.edu (Andru Luvisi)
Date: Thu, 16 Jan 2003 10:01:35 -0800 (PST)
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030116123354.E28733@bcr10.uwaterloo.ca>
Message-ID: <Pine.LNX.4.44.0301160958120.11570-100000@gladen>

On Thu, 16 Jan 2003, David Evans wrote:
[snip]
>   If you could do something like
>
> (dd if=mtboot sync bs=512 |+ dd if=mtboot sync bs=512 |+ dd if=boot sync bs=512) | dd of=/dev/whatever obs=512
>
> in order to concatenate the multiple copies of mtboot then all would be well.

How about:
(
  dd if=mtboot conv=sync bs=512
  dd if=mtboot conv=sync bs=512
  dd if=boot conv=sync bs=512
) |dd of=/dev/whatever bs=512

Andru
-- 
Andru Luvisi, Programmer/Analyst

Quote Of The Moment:
  It has been said very justly that every virtue is, as it were, a
  middle course between two opposite vices. Thus, courage is midway
  between cowardice and rashness; thrift is midway between
  acquisitiveness and improvidence; morality is midway between prudery
  and licentiousness; and so forth. In like manner, justification stands
  half-way between the faults of pedantry, on the one hand, and the
  failures of ignorance, on the other.
  
  	- Nevil Maskelyne



From norman at nose.cs.utoronto.ca  Fri Jan 17 03:53:59 2003
From: norman at nose.cs.utoronto.ca (Norman Wilson)
Date: Thu, 16 Jan 2003 12:53:59 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
Message-ID: <200301161754.h0GHsbn49154@minnie.tuhs.org>

If all that is needed is to concatenate three files into a single
tape file, with each new file aligned on a block boundary, isn't

	(for f in f1 f2 f3; do dd <$f conv=sync; done) >/dev/rmt0

sufficient?

The only time I had to do anything like this, the target system was
a MicroVAX; hence the tape had to have ISO labels, and the program
to be booted a particular tape filename.  I don't remember how I
handled that; probably I used one of the several public-domain
tape-label-writing programs.  But that was a VAX; I would assume
an 11/73 doesn't have so much enforced complexity in its firmware.


From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 03:55:38 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 12:55:38 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <Pine.LNX.4.44.0301160958120.11570-100000@gladen>; from luvisi@andru.sonoma.edu on Thu, Jan 16, 2003 at 10:01:35AM -0800
References: <20030116123354.E28733@bcr10.uwaterloo.ca> <Pine.LNX.4.44.0301160958120.11570-100000@gladen>
Message-ID: <20030116125538.A29181@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 10:01:35AM -0800, Andru Luvisi wrote:
> How about:
> (
>   dd if=mtboot conv=sync bs=512
>   dd if=mtboot conv=sync bs=512
>   dd if=boot conv=sync bs=512
> ) |dd of=/dev/whatever bs=512
> 

  Will that actually work?  I'll give it a try.  I haven't had to do much with
dd in almost ten years.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From emu at ecubics.com  Fri Jan 17 04:24:25 2003
From: emu at ecubics.com (emanuel stiebler)
Date: Thu, 16 Jan 2003 11:24:25 -0700
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
References: <20030115154831.A9065@bcr10.uwaterloo.ca>
Message-ID: <3E26F8D9.3040104@ecubics.com>

David Evans wrote:
 >   Howdy.
 >
 >   I'd like to confirm that somebody has successfully made a TK50 boot 
tape
 > for 2.11BSD using a TK50Z attached to some other Unix box.

Yes. Did it to a TK50Z & TZ30 many times. Slow, but works ...

 > In my case it will
 > likely have to be Linux, though IRIX might be doable.  It *should* 
work, after
 > all, but, well, you know.  :)

What version is your "linux" ? There were problems in earlier version 
(redhat before 7.x ?) exactly with the block command on tapes, and 
that's what bites you :(

Just make sure your tape drive is clean, check if you get any errors 
during writing the tape.

And don't use any scripts, just compile the maketape.c, it worked for me 
many times.

cheers






From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 04:59:52 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 13:59:52 -0500
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <3E26F8D9.3040104@ecubics.com>; from emu@ecubics.com on Thu, Jan 16, 2003 at 11:24:25AM -0700
References: <20030115154831.A9065@bcr10.uwaterloo.ca> <3E26F8D9.3040104@ecubics.com>
Message-ID: <20030116135952.A29189@bcr10.uwaterloo.ca>

On Thu, Jan 16, 2003 at 11:24:25AM -0700, emanuel stiebler wrote:
> What version is your "linux" ? There were problems in earlier version 
> (redhat before 7.x ?) exactly with the block command on tapes, and 
> that's what bites you :(
> 

  This is a Debian setup running a 2.4.18 kernel.  I'm not really surprised
that it didn't work...Linux always seems to hose me when I try to do something
vaguely "tricky."

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From emu at ecubics.com  Fri Jan 17 05:45:56 2003
From: emu at ecubics.com (emanuel stiebler)
Date: Thu, 16 Jan 2003 12:45:56 -0700
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
References: <20030115154831.A9065@bcr10.uwaterloo.ca> <3E26F8D9.3040104@ecubics.com> <20030116135952.A29189@bcr10.uwaterloo.ca>
Message-ID: <3E270BF4.7030205@ecubics.com>

David Evans wrote:

>   This is a Debian setup running a 2.4.18 kernel.  I'm not really surprised
> that it didn't work...Linux always seems to hose me when I try to do something
> vaguely "tricky."

Just because you try to boot a 2.11BSD, the last three letters should 
tell you something. Go for FreeBSD, NetBSD, or something with a "BSD" in 
the name ... ;-)



From jkunz at unixag-kl.fh-kl.de  Fri Jan 17 03:39:13 2003
From: jkunz at unixag-kl.fh-kl.de (Jochen Kunz)
Date: Thu, 16 Jan 2003 18:39:13 +0100
Subject: [pups] Making a TK50 2.11BSD boot tape for an 11/73
In-Reply-To: <20030116110643.B28361@bcr10.uwaterloo.ca>; from dfevans@bbcr.uwaterloo.ca on Thu, Jan 16, 2003 at 17:06:43 CET
References: <20030116094037.B28194@bcr10.uwaterloo.ca> <Pine.LNX.4.44.0301160744570.11570-100000@gladen> <20030116110643.B28361@bcr10.uwaterloo.ca>
Message-ID: <20030116183913.Y57315@MissSophie.unixag-kl.fh-kl.de>

On 2003.01.16 17:06 David Evans wrote:

>   Ahhh--of course.  One would have to do a series of "dd"s in order to
> get the padding happening.
That is only half the thing. It is also important to use the proper tape
block size. maketape uses some ioctls to set this, dd don't know about
this. 

> That's not a bad idea.  Perhaps I should just make a minimal
> bootable image using simh or such, dd it to a SCSI disk, and
> then attach that to my 11/73.
A common solution. 

> Might be easier than endless futzing with the TK50.
Whimp. ;-) 
-- 



tschüß,
         Jochen

Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/


From dfevans at bbcr.uwaterloo.ca  Fri Jan 17 13:29:51 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Thu, 16 Jan 2003 22:29:51 -0500
Subject: [pups] PDP-11 happily booted
Message-ID: <20030116222951.A1396@bcr10.uwaterloo.ca>

  Yanked a SCSI disk and the penguin could manage dding the simh image on to
it.  The PDP-11 booted quite happily.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From migieger at bawue.de  Sun Jan 19 23:02:47 2003
From: migieger at bawue.de (M. Giegerich)
Date: Sun, 19 Jan 2003 18:32:47 +0530
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
Message-ID: <20030119130247.GA18906@luva.home>

I'm trying to run 2.11 BSD on the p11 emulator...
The 2.11 BSD boot image I'm using is 211_on_rl2
(which works on sim, but without network sup-
port).

Before downloading 2.11_rp_unknown.gz that seems
to work for sure with p11 (at least I found enough
posts hinting at that), I would like to know if
anybody has successfully used the 211_on_rl2 distro
to boot 2.11 BSD on p11.

My results so far are p11 boots (after FIGHTING
with p11conf), recognizes the 4 rl and 5 rk disks
and images and gives me a @-prompt.
Telnetting to the defined ports won't do :-(

Any hints anybody?

-- 
M. Giegerich, mail: migieger at vsnl.com, phone: +91.(0)80.5530154


From migieger at bawue.de  Mon Jan 20 01:58:23 2003
From: migieger at bawue.de (M. Giegerich)
Date: Sun, 19 Jan 2003 21:28:23 +0530
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
In-Reply-To: <20030119130247.GA18906@luva.home>
References: <20030119130247.GA18906@luva.home>
Message-ID: <20030119155823.GA19800@luva.home>

I shouldn't do this, I know...

On Sun, Jan 19, 2003 at 06:32:47PM +0530, M. Giegerich wrote:
> I'm trying to run 2.11 BSD on the p11 emulator...
> The 2.11 BSD boot image I'm using is 211_on_rl2
> (which works on sim, but without network sup-
> port).
> 
> Before downloading 2.11_rp_unknown.gz that seems
> to work for sure with p11 (at least I found enough
> posts hinting at that), I would like to know if
> anybody has successfully used the 211_on_rl2 distro
> to boot 2.11 BSD on p11.
> 
> My results so far are p11 boots (after FIGHTING
> with p11conf), recognizes the 4 rl and 5 rk disks
> and images and gives me a @-prompt.
> Telnetting to the defined ports won't do :-(
> 
> Any hints anybody?

@17773000g did the magic... Amazing, this was really
well hidden.

The kernel on the rl images isn't set-up for networ-
king. So, I'm trying to rebuild it - getting stuck
at "make: out of memory" or "ld: too big for type
431"...
We'll see.

-- 
M. Giegerich, mail: migieger at vsnl.com, phone: +91.(0)80.5530154


From dfevans at bbcr.uwaterloo.ca  Mon Jan 20 14:18:11 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Sun, 19 Jan 2003 23:18:11 -0500
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
In-Reply-To: <20030119155823.GA19800@luva.home>; from migieger@bawue.de on Sun, Jan 19, 2003 at 09:28:23PM +0530
References: <20030119130247.GA18906@luva.home> <20030119155823.GA19800@luva.home>
Message-ID: <20030119231811.B29049@bcr4.uwaterloo.ca>

On Sun, Jan 19, 2003 at 09:28:23PM +0530, M. Giegerich wrote:
> The kernel on the rl images isn't set-up for networ-
> king. So, I'm trying to rebuild it - getting stuck
> at "make: out of memory" or "ld: too big for type
> 431"...

  For the "431" business, check out the man page for ld and read the
section on overlay sizes.  Then check in the Makefile and move some .o
files between overlays so that they are all under 8192 bytes.  That
worked for me.  A little combo of sed and awk might be handy to check
the total size of the objects included in each overlay.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From jkunz at unixag-kl.fh-kl.de  Mon Jan 20 18:07:10 2003
From: jkunz at unixag-kl.fh-kl.de (Jochen Kunz)
Date: Mon, 20 Jan 2003 09:07:10 +0100
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
In-Reply-To: <20030119231811.B29049@bcr4.uwaterloo.ca>; from dfevans@bbcr.uwaterloo.ca on Mon, Jan 20, 2003 at 05:18:11 CET
References: <20030119130247.GA18906@luva.home> <20030119155823.GA19800@luva.home> <20030119231811.B29049@bcr4.uwaterloo.ca>
Message-ID: <20030120090710.M57315@MissSophie.unixag-kl.fh-kl.de>

On 2003.01.20 05:18 David Evans wrote:

> A little combo of sed and awk might be handy to check
> the total size of the objects included in each overlay.
Been there, done that. ;-) 
Check the mail archive for a mail from me, sent at Jun 18, 2002 with
subject "Re: [pups] bsd2.11 kernel compile". It contains a Makefile
supplement that eases overly calculation. 
-- 



tschüß,
         Jochen

Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/


From dfevans at bbcr.uwaterloo.ca  Tue Jan 21 10:41:35 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Mon, 20 Jan 2003 19:41:35 -0500
Subject: [pups] The infamous 2.11BSD patch archive
Message-ID: <20030120194135.A19753@bcr10.uwaterloo.ca>

  This *must* be a FAQ, but I didn't find a real answer to it anywhere.

  What's up with the 2.11BSD patch archive?  I assume that the version
in the Unix Archive is the original release and does not have these
patches applied.  The mirror to which 2bsd.com referrs is gone and I
don't want to pound on Steve's site if I don't have to.  Any
suggestions?  Once I get the files I presume I'm just supposed to feed them
through patch in sequence (with the appropriate options, of course).

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From sms at 2BSD.COM  Tue Jan 21 11:17:09 2003
From: sms at 2BSD.COM (Steven M. Schultz)
Date: Mon, 20 Jan 2003 17:17:09 -0800 (PST)
Subject: [pups] The infamous 2.11BSD patch archive
Message-ID: <200301210117.h0L1H9o11718@moe.2bsd.com>

Hi -

> From: David Evans <dfevans at bbcr.uwaterloo.ca>
>   What's up with the 2.11BSD patch archive?  I assume that the version

	Both it and the mirror are up and running.   There are quite a few
	netblocks that have portscanned/spammed/whatever me here (dialups
	and cablemodems for the most part) and they're listed in the packet
	filters so if you're seeing timeouts, etc connecting that's probably
	the reason.

> in the Unix Archive is the original release and does not have these
> patches applied.  The mirror to which 2bsd.com referrs is gone and I

	Companies play the "musical domain name" everytime they have a
	re-organization (and, sigh, another one is in the works as I type).
	What was listed was probably a name from the last reorganization and
	of course it's very difficult to track down all the places that an
	old name was used.    

	The mirror I maintain at work is:

		ftp://ftp.cato.gd-ais.com

	(for some time yet it will also respond to 'ftp.to.gd-es.com')

	There may be other sites that also mirror

> don't want to pound on Steve's site if I don't have to.  Any

	Others have done that - usually it's not too bad, especially if I'm
	asleep at the time :)

> suggestions?  Once I get the files I presume I'm just supposed to feed them
> through patch in sequence (with the appropriate options, of course).

	Sequence is important because the patches are diffs against the last
	current version of a file.

	Each patch has rather good (if I do say so myself ;)) instructions and
	explainations of what was changed, how to apply, what to rebuild 
	immediately (and what can be deferred until it is convenient).

	Have Fun!

	Steven Schultz
	sms at 2bsd.com


From migieger at bawue.de  Tue Jan 21 12:45:06 2003
From: migieger at bawue.de (M. Giegerich)
Date: Tue, 21 Jan 2003 08:15:06 +0530
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
In-Reply-To: <20030120090710.M57315@MissSophie.unixag-kl.fh-kl.de>
References: <20030119130247.GA18906@luva.home> <20030119155823.GA19800@luva.home> <20030119231811.B29049@bcr4.uwaterloo.ca> <20030120090710.M57315@MissSophie.unixag-kl.fh-kl.de>
Message-ID: <20030121024506.GA369@luva.home>

On Mon, Jan 20, 2003 at 09:07:10AM +0100, Jochen Kunz wrote:
> On 2003.01.20 05:18 David Evans wrote:
> 
> > A little combo of sed and awk might be handy to check
> > the total size of the objects included in each overlay.
> Been there, done that. ;-) 
> Check the mail archive for a mail from me, sent at Jun 18, 2002 with
> subject "Re: [pups] bsd2.11 kernel compile". It contains a Makefile
> supplement that eases overly calculation. 

Yes, I did read this before posting...

It was easy to change the objects in the overlays
as only the first was 200 bytes too big while others
had plenty of space left. So compiling went well...

Unfortunately after installing and rebooting the
system panick'ed. After toying for too long with it,
I decided to download the rp image and give it a
try instead. Good decision as now everything is
working well (incl. networking to/fro the host).

The host is a dual P3 800 with FreeBSD-4.7R, p11-2.9
PDP 11 emulator and (the begemot 2.11bsd image on)
2.11_rp_unknown.gz.

I may post a summary on how to get this running
as I needed quite some reading in different mails
of the pups archive... (may make life easier for
others).

-- 
M. Giegerich, mail: migieger at vsnl.com, phone: +91.(0)80.5530154


From migieger at bawue.de  Wed Jan 22 12:38:37 2003
From: migieger at bawue.de (M. Giegerich)
Date: Wed, 22 Jan 2003 08:08:37 +0530
Subject: [pups] FreeBSD-4.7, p11-2.9, 2.11bsd
In-Reply-To: <20030121024506.GA369@luva.home>
References: <20030119130247.GA18906@luva.home> <20030119155823.GA19800@luva.home> <20030119231811.B29049@bcr4.uwaterloo.ca> <20030120090710.M57315@MissSophie.unixag-kl.fh-kl.de> <20030121024506.GA369@luva.home>
Message-ID: <20030122023837.GA385@luva.home>

On Tue, Jan 21, 2003 at 08:15:06AM +0530, M. Giegerich wrote:
...
> I may post a summary on how to get this running
> as I needed quite some reading in different mails
> of the pups archive... (may make life easier for
> others).

Done before memory fades away...

If there is anybody trying this out, I would appre-
ciate some feedback on this primer.


Installing 2.11BSD on the PDP11 emulator p11 on FreeBSD
=======================================================

This relates to FreeBSD 4.7-RELEASE-p3 (the most current
4.7 RELEASE as on Jan. 19, 2003).

You'll need Harti Brandt's PDP11 emulator. I used the
latest non-beta release p11-2.9 dated Mar. 1st, 2001.
To build p11, libbegemot is needed too (libbegemot-1.9
dated Mar. 1st, 2001). Both are available at
ftp://ftp.fokus.gmd.de/pub/cats/usr/harti/p11.

116847 Mar  1  2001 libbegemot-1.9.tar.gz
298755 Mar  1  2001 p11-2.9.tar.gz
MD5 (libbegemot-1.9.tar.gz) = 92bc75e99c5bc778f59adf412ea085ea
MD5 (p11-2.9.tar.gz) = ffabf9752895d34396b2d5b38947c8dc

(My decision to use p11 was based on the fact, that
it's much less cpu hungry than Bob Supnik's sim
package).

The 2.11BSD distribution I use is 2.11_rp_unknown.gz
from http://www.tuhs.org/Archive/PDP-11/Boot_Images.

34600118 Feb  1 1998 2.11_rp_unknown.gz
MD5 (2.11_rp_unknown.gz) = d57863a498edc4598da84eb5bc88b3a2

After downloading you'll want to build libbegemot
and p11. Both is painless.

Building libbegemot
-------------------

cd work
tar xvzf ../libbegemot-1.9.tar.gz
find libbegemot-1.9 | xargs chown user.group
(May be omitted, but on my system file ownership
was messed up).
cd libbegemot-1.9
./configure
gmake
gmake install

Building p11
------------

cd work
tar xvzf ../p11-2.9.tar.gz
find p11-2.9 | xargs chown user.group
(May be omitted, but on my system file ownership
was messed up).
cd p11-2.9
./configure --with-begemot=/usr/local
gmake
gmake install

Configuring p11
---------------

The config file is in /usr/local/share/p11/p11conf.
Minimal changes are required:

Check for "define diskdir /home/private/p11/disks"
and change according to your setup. I gunzip'ed
2.11_rp_unknown.gz to /var/tmp/disks/RP_211bsd_root
to maintain the file naming of Harti. Beware, it
expands to 174 MB.

I also changed the sync time for the emulated rp
disk from 600000 to 4000 (as indicated in the docu-
mentation of p11.texi).
That is "ctrl rp 17776700 0254 5 4000" instead of
the original "ctrl rp 17776700 0254 5 600000".

(I changed the deqna's network driver from tun0
to tun3 too, as I'm using tun0 for dialup pur-
poses).

Last, I commented out the Linux tun and tap as well
as the FreeBSD tap networking entries.

My p11conf reads

/usr/local/share/p11/p11conf
++++++++++++++++++++++++++++

--- snip --- snip --- snip ---
# Copyright (c)1994-2000 Begemot Computer Associates. All rights reserved.
# See the file COPYRIGHT for details of redistribution and use.
#
# $Id: p11conf.in,v 1.6 2000/12/01 12:27:00 hbb Exp $

#
# This configuration file is somehow tailored to my setup, but it is usable
# as a fallback in any case. Make a private copy and change it as you need it.
#

#
# Documentation for the configuration syntax is not yet ready :-( so here
# is a short version. You may look also in main.c
#
# The file is read line by line. If the line is empty or starts with a #
# it is ignored. The first word of each line is a keyword and determines how
# the rest of the line is interpreted:
#
#  define NAME [VALUE]
#	Define a variable (or change it's value). Variables can be access by
#	the $(NAME) construct. Substitution takes place on VALUE, but not on
#	NAME. Referencing an undefined variable is an error unless the form
#	$(NAME?) is used, in which case an empty string is substituted.
#
#  set OPTION [VALUE]
#	Set OPTION to value and define a variable OPTION with the given VALUE.
#	There are global and controller specific options. Global options
#	should be set outside all controller declarations (outside 'ctrl'-'end'
#	blocks.
#	Subsitution is done on VALUE but not on OPTION.
#	Options are:
#
#	libdir		- this is where help and configuration files are looked
#			  for as well as rom images (global).
#	iodir		- directory for the IO programs (global).
#	clock_rate	- clock ticks per second (usually 50 or 60). (global).
#	cpu_options	- these are the jumpers on the KDJ11A (ctrl KDJ11A).
#	csw_enable	- enable the console switch register (default) (ctrl KDJ11A).
#	csw		- starting contents of the switches (ctrl KDJ11).
#	maxreq		- maximum number of pending line clock interrupts
#			  (default is 1) (ctrl KDJ11A).
#	localtime	- toy clock should report local time, not GMT (ctrl toy).
#
#  include FILE
#	Include the given FILE. Substitution takes place on FILE.
#
#  print LINE
#	Print LINE on stdout after performing substitutions.
#
#  if EXPR
#	Substitute on EXPR, evaluate it and process the following lines up to
#	a matching else, elif of endif if EXPR is non-zero. There is a limit
#	on nested if's.
#
#  elif EXPR
#	If all if-clauses where false up to now, perform substitution, compute
#	EXPR and process all lines up to the matching else, elif or endif if
#	EXPR is non-zero.
#
#  else
#	If all if-clauses where false up to now, process all lines up to the
#	matching endif.
#
#  endif
#	Ends an if construct.
#
#  ctrl NAME [ARGS]
#	Plug in the given controller into the backplane. Substitution is done
#	on ARGS. What ARGS are needed depends on the concrete controller.
#	Note, that controllers are plugged in bottom-down, i.e. the last one
#	gets nearest to memory (which is automatically inserted), which in turn
#	gets nearest to the CPU. The controller line may be followed by
#	device lines which must be followed by end 'end' line.
#
#  dev [ARGS]
#	Connect a device to the current controller. Substitution is done on
#	ARGS and if this yields empy ARGS, the line is ignored. ARGS depends
#	on the controller.
#
#  end
#	End the current controller definition.
#
#  eval NAME [EXPR]
#	Compute EXPR after performing substitutions and assign the result as
#	a signed decimal integer to NAME.
#
#
# Expressions consists of the usual +, -, *, /, %, (, ), == and !=.
# The following builtin functions are available:
#
#	acess(FILE, MODE)
#		Check whether FILE is accessable (mode may be one or more
#		of 'f', 'r', 'w' and 'x'. Returns 1 if ok.
#
#	eq(STR1,STR2)
#		Compare the two strings after stripping leading and trailing
#		space and return 1 of they are equal.
#
#	exec(COMMAND)
#		Execute COMMAND with sh -c and return the exit code of the
#		shell (this may or may not be the exit code of the command).
#

if !defined(diskdir)
  define diskdir /var/tmp/disks
endif

define RP_211_BOOT	0 $(diskdir)/RP_211bsd_root 1999
define RP_41_BOOT	0 $(diskdir)/RP_rsx_4.1boot 2000
define RP_41_SRC	1 $(diskdir)/RP_rsx_4.1src 2001

#
# standard rlv12 controller with four drives attached
# numbers are csr, vector, interrupt level and ms between msyncs
#
ctrl rl 017774400 0160 4 60000
end

#
# rp06 disk controller with 2 drives attached
# numbers are csr, vector, interrupt level and ms between msyncs
#
ctrl rp 017776700 0254 5 4000
if defined(BOOT)
 if eq($(BOOT),211)
  print Booting 2.11BSD
	dev $(RP_211_BOOT)
 elif eq($(BOOT),RSX)
  print Booting RSX4.1
	dev $(RP_41_BOOT)
	dev $(RP_41_SRC)
 endif
endif
end

#
# kl11a async interface with up to 4 lines
# kl0 is console
# numbers are csr, vectors, interrupt level, driver program and parameters
# for driver program
# drivers:
#	tty_tty [-7] device
#	tty_net [-7] [-t] [-lLogFile] port-number
#	tty_xterm [-7] [-t title] [-g geometry] [-p position] [-i] [-x xterm-args]
#	tty_fifo [-7] fifo-name
#	tty_file [-7] [-p pidfile] [-e eof-seq] [-c] file-or-prog [args...]
#
ctrl kl
	dev 017777560 060 064 4 tty_net -t -7 10000
	dev 017776500 0300 0304 4 tty_net -t -7 10001
#	dev 017776510 0310 0314 4 tty_fifo /tmp/fifo
end

#
# MRV12 rom option
# args are csr and file containing boot-rom
#
if !defined(BOOT_ROM)
  if access(boot.rom, r)
    define BOOT_ROM boot.rom
  elif access(Boot/boot.rom, r)
    define BOOT_ROM Boot/boot.rom
  else
    define BOOT_ROM $(libdir)/boot.rom
  endif
endif
ctrl mr 017777520 $(BOOT_ROM)
end

#
# parallel lp interface
# numbers are csr, vector and interrupt level
# the device line contains either a filename or a '|' which creates a pipe
# followed by the program to pipe in and optional args
#
ctrl lp 017777514 0200 4
end

#
# DEQNA ethernet controller
# I can't distribute my ROM's image. So check whether there is any. If not
# don't plug in the thing. You can create a file consisting of 0's and use
# that instead. Use either tun or tap (hey - tap let's you use MOPR/MOPDL or
# DreckNet or ...)
#
if !defined(QNA_ROM)
  if access(qna.rom, r)
    define QNA_ROM qna.rom
  elif access($(libdir)/qna.rom, r)
    define QNA_ROM $(libdir)/qna.rom
  endif
endif

if defined(QNA_ROM)
  ctrl qna 017774440 5 0x08:0x00:0x2b:0x07:0x82:0x6c 0xf8:0x7a $(QNA_ROM)
# FreeBSD:
	dev epp_tun /dev/tun3 0x08:0x00:0x2b:0x07:0x82:0x6c 0x08:0x00:0x2b:0x07:0x82:0x00
# Linux:
#	dev epp_tun /dev/net/tun 0x08:0x00:0x2b:0x07:0x82:0x6c 0x08:0x00:0x2b:0x07:0x82:0x00

# FreeBSD:
#	dev epp_tap /dev/tap0
# Linux:
#	dev epp_tap /dev/net/tun

#	eval dummy exec(ifconfig tap0 <your host> up)
#	eval dummy exec(ifconfig tap0)
  end
endif

#
# Tape controller
#
ctrl tm 017772520 0224 5
end

#
# The toy clock.
#
ctrl toy 017777526
#  set localtime yes
#  set offs_year 10
#  set settoy 12:30:00 12/31/74
end

#
# Plug in the memory board.
#
ctrl mem
end

#
# Plug in processor board
#
ctrl KDJ11A
  if defined(CSW)
    set csw $(CSW)
  endif
end
--- snip --- snip --- snip ---

Next create a 4096 byte file called qna.rom in
/usr/local/share/p11 filled with binary 0s as per
instructions in p11conf.

That's it.

Running p11
-----------

Invoke the emulator as `p11 -DBOOT=211` and you will
be presented with

--- screen output ---
# p11 -DBOOT=211
/usr/local/share/p11/p11conf,126: Booting 2.11BSD
173000 K -> 17773000 = 000005:	reset	
@
--- screen output ---

Then switch to another virtual terminal (vty) and
telnet into the p11 emulator on port 10000.

--- screen output ---
# telnet localhost 10000
Trying ::1...
telnet: connect to address ::1: Connection refused
Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.
--- screen output ---

Switch back to the p11 emulator vty and enter

--- screen output ---
@17773000g (without the @, that's p11's prompt).
--- screen output ---

And one more switch back to the 2.11bsd vty will
show you

--- screen output ---
Begemot boot PROM; V0.3
Copyright (c) 1998-2000 Begemot Computer Associates. All rights reserved.
boot>
--- screen output ---

Enter

--- screen output ---
boot>BOOT RP (again without "boot>")
--- screen output ---

and you will get

--- screen output ---
73Boot from xp(0,0,0) at 0176700
: 
--- screen output ---

Hit enter, wait a little  and you will be greeted by

--- screen output ---
: xp(0,0,0)unix
Boot: bootdev=05000 bootcsr=0176700

2.11 BSD UNIX #11: Tue Jan 6 16:57:02 MET 1998
    root at pdp11.begemot.com:/usr/src/sys/HIPPON

attaching qe0 csr 174440
qe0: DEC DEQNA addr 08:00:2b:07:82:6c
attaching lo0

phys mem  = 2097152
avail mem = 1647872
user mem  = 307200

January 20 19:51:18 init: configure system

lp 0 csr 177514 vector 200 attached
rl 0 csr 174400 vector 160 attached
tm 0 csr 172520 vector 224 attached
xp 0 csr 176700 vector 254 attached
cn 1 csr 176500 vector 300 attached
cn 2 csr 176510 vector 310 skipped:  No CSR.
cn 3 csr 176520 vector 320 skipped:  No CSR.
cn 4 csr 176530 vector 330 skipped:  No CSR.
erase, kill ^U, intr ^C
#
--- screen output ---

That's 2.11BSD's single user console.

Configure 2.11BSD for networking
--------------------------------

(Do yourself a favour and invoke tcsh).

You'll need to configure 

/etc/hosts
/etc/netstart
/etc/networks
/etc/resolv.conf

and invoke once `/etc/mkhosts /etc/hosts`.

To explain the entries in these files, I'll use
freebsd.home and 192.168.0.33 as the hostname and
the ip address of the FreeBSD host system, pdp11
and 192.168.0.45 as the hostname and ip address
of the 2.11BSD system on the p11 emulator. The
domain name for both will be home.

/etc/hosts
++++++++++

--- snip --- snip --- snip ---
127.0.0.1	localhost
192.168.0.33	freebsd.home
192.168.0.45	pdp11.home
--- snip --- snip --- snip ---

/etc/netstart
+++++++++++++

--- snip --- snip --- snip ---
#!/bin/sh -

#	@(#)netstart	5.2.1 (2.11BSD) 1996/11/16

INET=`testnet`
if [ $INET = YES ]; then
	echo Assuming NETWORKING system ...		>/dev/console 2>&1
else
	echo Assuming non-networking system ...		>/dev/console 2>&1
fi
export INET

# DON'T EVEN THINK ABOUT RUNNING ROUTED ON A PDP-11!!!  Just set routedflags
# to "NO" and set up a default gateway below.
#
routedflags=NO
rwhod=YES
named=NO

# myname is my symbolic name
# my-netmask is specified in /etc/networks
# my-default-address should be set to some machine where you
#	want to send all your packets, by default.
#

hostname=pdp11.home
hostid=2B07826C
netmask=255.255.255.0
broadcast=192.168.0.255
default=freebsd.home

hostname $hostname					>/dev/console 2>&1
hostid $hostid						>/dev/console 2>&1

if [ $INET = YES ]; then
	# ifconfig imp0 inet netmask $netmask $hostname broadcast $broadcast up -trailers >/dev/console 2>&1
	# ifconfig en0 inet netmask $netmask $hostname broadcast $broadcast up -trailers >/dev/console 2>&1
	# ifconfig ec0 inet netmask $netmask $hostname broadcast $broadcast up -trailers arp >/dev/console 2>&1
	# ifconfig il0 inet netmask $netmask $hostname broadcast $broadcast up -trailers >/dev/console 2>&1
	ifconfig qe0 inet netmask $netmask $hostname broadcast $broadcast up -trailers >/dev/console 2>&1
	arp -s hippo 08:00:2b:07:82:00
	# ifconfig sl0 inet 192.254.254.2 192.254.254.1 -arp -trailers >/dev/console 2>&1
	# slattach /dev/ttyS6 9600
	arp -s 192.168.0.45 08:00:2b:07:82:6c pub
	arp -s 192.168.0.33 00:06:29:13:7a:18 pub
	ifconfig lo0 inet localhost up -trailers	>/dev/console 2>&1

	route add $hostname localhost 0		>/dev/console 2>&1
	route add default $default 1		>/dev/console 2>&1
fi
--- snip --- snip --- snip ---

/etc/networks
+++++++++++++

--- snip --- snip --- snip ---
localnet 127
home 192.168.0
--- snip --- snip --- snip ---

/etc/resolv.conf
++++++++++++++++

--- snip --- snip --- snip ---
domain home
nameserver 192.168.0.33
--- snip --- snip --- snip ---

In /etc/netstart you need to set the arp entries
of both freebsd and pdp11 with their respective
ethernet mac addresses (pdp11's qe0 mac address
is set in p11conf). See /etc/netstart above.

Now don't forget to invoke `/etc/mkhosts /etc/hosts`.

After that (exit tcsh, if you were in and) exit csh
hitting .
The 2.11BSD system will go multi user and show you

--- screen output ---
# ^D
Fast boot ... skipping disk checks
checking quotas: done.
Assuming NETWORKING system ...
add host pdp11.home: gateway localhost
add net default: gateway freebsd.home
starting system logger
preserving editor files
clearing /tmp
standard daemons: update cron accounting.
starting network daemons: inetd rwhod printer.
starting local daemons:.
Mon Jan 20 20:27:36 PST 2003
Jan 20 20:27:36 pdp11 January 20 20:27:36 init: kernel security level changed from 0 to 1


2.11 BSD UNIX (pdp11.home) (console)

login:
--- screen output ---

You login as root with password "begemot1" (without
the "").

--- screen output ---
login: root
Password:

Last login: Mon Jan 20 20:15:16 from freebsd
2.11 BSD UNIX #11: Tue Jan 6 16:57:02 MET 1998
[1] root--> 
--- screen output ---

Before you'll be able to network to freebsd, you'll
need to setup the tun device on its side. Switch
to a vty on freebsd and as root enter `ifconfig
tun0 freebsd pdp11 up`.

Go back to the pdp11 vty and try

--- screen output ---
[1] root--> ping freebsd.home
PING freebsd.home (192.168.0.33): 56 data bytes
64 bytes from 192.168.0.33: icmp_seq=0 ttl=64 time=40 ms
64 bytes from 192.168.0.33: icmp_seq=1 ttl=64 time=40 ms
64 bytes from 192.168.0.33: icmp_seq=2 ttl=64 time=40 ms
64 bytes from 192.168.0.33: icmp_seq=3 ttl=64 time=40 ms
^C
--- freebsd.home ping statistics ---
4 packets transmitted, 4 packets received, 0% packet loss
round-trip min/avg/max = 40/0/40 ms
[2] root-->
--- screen output ---

Shutting down
-------------

To shut down pdp11 issue `halt`:

--- screen output ---
[2] root--> halt
Jan 20 20:29:12 pdp11 January 20 20:29:12 halt: -; halted by root
Jan 20 20:29:13 pdp11 syslogd: going down on signal 15
syncing disks... done
halting
--- screen output ---

After pdp11 is halted you switch to the p11 vty and
issue "x" (without "")

--- screen output ---
@17773000g
000014 K -> 00123214 = 012767:	mov	#1,12456
[4   - kl]
@xDCOK = 1 asserted
--- screen output ---
 ^
The emulator returns you to the command prompt of
your FreeBSD system. 
There issue `ifconfig tun0 freebsd pdp11 down` to
delete the unneeded tun interface.

Todos
-----

Explore which patches (if any) from ftp://moe.2bsd.com/
pub/2.11BSD are incorporated into this 2.11BSD distri-
bution. Maybe Steven can comment on this...

-- 
M. Giegerich, mail: migieger at vsnl.com, phone: +91.(0)80.5530154


From dfevans at bbcr.uwaterloo.ca  Thu Jan 30 14:49:25 2003
From: dfevans at bbcr.uwaterloo.ca (David Evans)
Date: Wed, 29 Jan 2003 23:49:25 -0500
Subject: [pups] DLVJ1 confusion on 2.11BSD
Message-ID: <20030129234925.C18445@bcr10.uwaterloo.ca>

  So the DLVJ1 has four serial ports.  According to the info at

http://www.ibiblio.org/pub/academic/computer-science/history/pdp-11/hardware/DLVJ1.info

they essentially appear as four separate gizmos, with independant CSRs
and vectors.  So should I therefore set NKL to 5 (the console plus
these four) and then have appropriate entries for devices 1 through 4
in /etc/dtab?  Sounds like it.

  My DHV11 seems to work happily, so I already have rediculously more
serial ports than I need.

-- 
David Evans          (NeXTMail/MIME OK)             dfevans at bbcr.uwaterloo.ca
Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/
University of Waterloo         "Default is the value selected by the composer
Ontario, Canada           overridden by your command." - Roland TR-707 Manual


From helbig at Informatik.BA-Stuttgart.DE  Sat Jan  4 19:48:46 2003
From: helbig at Informatik.BA-Stuttgart.DE (Wolfgang Helbig)
Date: Sat, 4 Jan 2003 10:48:46 +0100 (MET)
Subject: [TUHS] db(I) in Unix V6
Message-ID: <200301041152.h04BqQL00209@bsd.korb>

Hi,
tinkering at V6--now I need db(I). It works fine, but three features
that are described in its manual page seem to be missing:
- The names of the general registers (r0, ...) are not build in, at least
  db only responds with `?' to `r0'. I can examine the registers
  via the `$' command though.
- The '*' feature doesn't work, i.e. `*_main' gives me a `?'.
- I can't specify a number being relocatable, i.e. '10r' gives me a `?'.

Is this an exercise to the user? Did anyone solve the exercise. Or did I
just not understand the manual page?

Greetings,

Wolfgang



From norman at nose.cs.utoronto.ca  Wed Jan  8 23:42:24 2003
From: norman at nose.cs.utoronto.ca (Norman Wilson)
Date: Wed, 08 Jan 2003 08:42:24 -0500
Subject: [TUHS] Re: Patches to improve 6th Edition
Message-ID: <200301081343.h08Dhin48421@minnie.tuhs.org>

A belated rejoinder to:

  This bug existed from (at least) the Fifth up until the Eighth
  editions from 1127 (BTL research). The Ninth edition had much
  rewrites in the kernel and the algorthim was changed at that point.

You're welcome.  I can't remember, though, whether I fixed it on
purpose or as part of the many general cleanups and restructurings
I did at that point.  And I didn't do enough of it; recently I
discovered a dumb problem that I probably introduced in the same
cleanups, wherein the i-node allocated for a file that couldn't be
created because the file system was full and the directory entry
couldn't be written was left allocated but nameless ...

Norman Wilson
Toronto ON


From helbig at Informatik.BA-Stuttgart.DE  Sat Jan 11 10:20:01 2003
From: helbig at Informatik.BA-Stuttgart.DE (Wolfgang Helbig)
Date: Sat, 11 Jan 2003 01:20:01 +0100 (MET)
Subject: [TUHS] unsigned division in V7's C
Message-ID: <200301110224.h0B2O8500309@bsd.korb>

Hi,
while wondering how unsigned integer division is implemented in C 
I found a bug in V7: a/b and a%b with b >= 2^15 does not give the expected
results, if a and b are unsigned int's.

Was this bug ever noticed or even fixed?

Greetings,

Wolfgang

Here is a program showing the bug:
main()
{
        unsigned int a, b;
        a = 60000;
        b = 40000;
        printf("a/b: %u, a%%b: %u\n", a/b, a%b);
        b = 25000;
        printf("a/b: %u, a%%b: %u\n", a/b, a%b);
}

The above program prints
a/b: 65534, a%b: 8928
a/b: 2, a%b: 10000

The first line should be of course
a/b: 1, a%b: 20000





From dmr at plan9.bell-labs.com  Sat Jan 11 17:32:41 2003
From: dmr at plan9.bell-labs.com (Dennis Ritchie)
Date: Sat, 11 Jan 2003 02:32:41 -0500
Subject: [TUHS] re: unsigned division in V7's C
Message-ID: <e16957b5aed653e277797d2d630c5b4b@plan9.bell-labs.com>

Helbig observed

  > while wondering how unsigned integer division is implemented in C
  > I found a bug in V7: a/b and a%b with b >= 2^15 does not give the expected
  > results, if a and b are unsigned int's.

Does

  http://www.cs.bell-labs.com/who/dmr/odd.html#muldiv

suffice to explain the behavior?

	Dennis



From helbig at Informatik.BA-Stuttgart.DE  Sun Jan 12 09:16:42 2003
From: helbig at Informatik.BA-Stuttgart.DE (Wolfgang Helbig)
Date: Sun, 12 Jan 2003 00:16:42 +0100 (MET)
Subject: [TUHS] unsigned division in V7's C
Message-ID: <200301120120.h0C1KrR00556@bsd.korb>

Warren replies:
>Not sure if:
>
>http://www.tuhs.org/Archive/PDP-11/Bug_Fixes/Net.v7bugs/0017
>
>is a bug fix for this.

And Dennis replies:
>Does

>  http://www.cs.bell-labs.com/who/dmr/odd.html#muldiv

>suffice to explain the behavior

Thanks for the replies, but the bug I found is about unsigned int's and
not long int's.

The V7 C code simply the DIV instruction for unsigned division. The dividend
is put in the low word of a register pair with the upper word cleared.

When the divisor is >= 2^15, DIV interprets it as a negative integer
and produces the result (N = 2^16):
	 N - a/(N-b)	instead of a/b, and
	 a%(N-b)	instead of a%b.

This explains the bug I've sent yesterday. I is easily fixed, since
when b >= N/2 and a < N, then a/b is either zero or one.

And there is another bug when using DIV for unsigned integers: If b = 1 and
a >= N/2, then a/b >= N/2, that is a signed overflow. With V7 on Bob Supniks
simulator (2.10), I get
	60000/1: 0, 60000%1: 60000
DIV did not change the dividend because of overflow.

Again a fix is easy: With unsigned ints, don't use DIV to divide by one.

In all other cases, DIV produces the correct unsigned results, if the dividend
is less N, that is its high word is zero.

I wonder why these bugs went unnoticed.

Greetings

Wolfgang



From Extern.David.Lange at gedas.com  Thu Jan 23 22:25:42 2003
From: Extern.David.Lange at gedas.com (Lange, David)
Date: Thu, 23 Jan 2003 07:25:42 -0500
Subject: [TUHS] The ^ = | ?
Message-ID: <47F5ADC94EF3C14AA51386CD1A20DDB8036D5EB6@usvwoaahs51>

Greetings,

I've been searching for some reference to the ^ symbol being the same a | in
Bourne shell. Does anyone remember seeing anything like this? I've searched
the early manpages to no avail.

Regards,

	D.


From johnh at psych.usyd.edu.au  Fri Jan 24 10:04:38 2003
From: johnh at psych.usyd.edu.au (John Holden)
Date: Fri, 24 Jan 2003 11:04:38 +1100 (EST)
Subject: [TUHS] The ^ = | ?
Message-ID: <200301240004.LAA06145@psychwarp.psych.usyd.edu.au>

> I've been searching for some reference to the ^ symbol being the same a |

Yes, pre Bourne shell, edition 6 and earlier used ^ for piped. Also chdir
insteads of the shortened 'cd'



From imp at bsdimp.com  Fri Jan 24 10:05:18 2003
From: imp at bsdimp.com (M. Warner Losh)
Date: Thu, 23 Jan 2003 17:05:18 -0700 (MST)
Subject: [TUHS] The ^ = | ?
In-Reply-To: <47F5ADC94EF3C14AA51386CD1A20DDB8036D5EB6@usvwoaahs51>
References: <47F5ADC94EF3C14AA51386CD1A20DDB8036D5EB6@usvwoaahs51>
Message-ID: <20030123.170518.49361029.imp@bsdimp.com>

In message: <47F5ADC94EF3C14AA51386CD1A20DDB8036D5EB6 at usvwoaahs51>
            "Lange, David" <Extern.David.Lange at gedas.com> writes:
: I've been searching for some reference to the ^ symbol being the same a | in
: Bourne shell. Does anyone remember seeing anything like this? I've searched
: the early manpages to no avail.

OS-MP 4.3D (SunOS 4.1.3) man pages say:

     A pipeline is a sequence of one or more  commands  separated
     by  `|'  (or,  for  historical  compatibility, by `^').  The

Warner


From cdl at mpl.ucsd.edu  Fri Jan 24 12:09:23 2003
From: cdl at mpl.ucsd.edu (Carl Lowenstein)
Date: Thu, 23 Jan 2003 18:09:23 -0800 (PST)
Subject: [TUHS] The ^ = | ?
Message-ID: <200301240209.SAA20717@opihi.ucsd.edu>

> From: "Lange, David" <Extern.David.Lange at gedas.com>
> To: "'tuhs at minnie.tuhs.org'" <tuhs at minnie.tuhs.org>
> Subject: [TUHS] The ^ = | ?
> 
> Greetings,
> 
> I've been searching for some reference to the ^ symbol being the same a | in
> Bourne shell. Does anyone remember seeing anything like this? I've searched
> the early manpages to no avail.

It's here in my printed copy of the 6th Edition Unix Programmer's Manual.

Page entitled SH(I)
Section DESCRIPTION
Subsection Command lines.  One or more commands separated by '|' or '^'
constitute a chain of filters.  . . .

    carl

-- 
    carl lowenstein         marine physical lab     u.c. san diego
                                                 clowenst at ucsd.edu


From norman at nose.cs.utoronto.ca  Fri Jan 24 13:09:51 2003
From: norman at nose.cs.utoronto.ca (Norman Wilson)
Date: Thu, 23 Jan 2003 22:09:51 -0500
Subject: [TUHS] The ^ = | ?
Message-ID: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10>

The earliest UNIX Programmer's Manual to describe shell
pipelines is the Third Edition, February 1973.  It gives a
syntax quite different from the modern one:

	com1 > com2 > com3 > outfile

meant what we would now write as

	com1 | com2 | com3 > outfile

This original syntax was pretty cumbersome; pretty
obviously it was put in as a quick hack (as were many
things in those early days).  Because > and < applied
only to the following word, pipelined commands with
arguments had to be quoted:

	who > "grep ken" >/tmp/kenlogins

Even worse, the shell had no inherent way to tell whether
the final word was a file or a program; if the last element
in a pipeline was to write to standard output, you had to
say so explicitly:

	who > "grep ken" >

On the other hand the syntax was symmetric: you could
also write

	"grep ken" < who <

pipe(II) also debuted in the Third Edition.

By the Fourth Edition (November 1973) there had evidently
been more time to think about the syntax; the modern notation
is shown, except that ^ is allowed as a synonym for |.  I have
long guessed that was because in those dark days of the
past, some upper-case-only terminals (remember stty lcase?)
offered no way to type | (and perhaps likewise {}`~) but I don't
really know.  Dennis?

Norman Wilson
Toronto ON




From wkt at minnie.tuhs.org  Fri Jan 24 15:37:45 2003
From: wkt at minnie.tuhs.org (Warren Toomey)
Date: Fri, 24 Jan 2003 15:37:45 +1000 (EST)
Subject: [TUHS] The ^ = | ?
In-Reply-To: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10> from Norman Wilson at "Jan
 23, 2003 10:09:51 pm"
Message-ID: <200301240537.h0O5bjd60797@minnie.tuhs.org>

In article by Norman Wilson:
> The earliest UNIX Programmer's Manual to describe shell
> pipelines is the Third Edition.

I don't have Peter Salus' book `A Quarter Century of UNIX'
here at home, but I think this topic is discussed in the
book. Dennis Ritchie's paper:

http://cm.bell-labs.com/cm/cs/who/dmr/hist.html

talks about the change from > to `another symbol', but not explicitly
the ^.


From lars at nocrew.org  Fri Jan 24 17:07:33 2003
From: lars at nocrew.org (Lars Brinkhoff)
Date: 24 Jan 2003 08:07:33 +0100
Subject: [TUHS] The ^ = | ?
In-Reply-To: <200301240537.h0O5bjd60797@minnie.tuhs.org>
References: <200301240537.h0O5bjd60797@minnie.tuhs.org>
Message-ID: <854r7z9gay.fsf@junk.nocrew.org>

Warren Toomey <wkt at minnie.tuhs.org> writes:
> From my memory of Salus' book, the > syntax was relaced by ^, but
> when Ken went to the U.K to give a talk, he made the decision to
> switch over to the | symbol.
> 
> Anyway, I could be completely wrong about this being in Salus'
> book, so if I am, I'll have to hunt down where this info came from.

You are right.  Pages 51-53 (in the 1995 reprint).

-- 
Lars Brinkhoff          http://lars.nocrew.org/     Linux, GCC, PDP-10,
Brinkhoff Consulting    http://www.brinkhoff.se/    HTTP programming


From dmr at plan9.bell-labs.com  Fri Jan 24 17:22:12 2003
From: dmr at plan9.bell-labs.com (Dennis Ritchie)
Date: Fri, 24 Jan 2003 02:22:12 -0500
Subject: [TUHS] re: The ^ = | ?
Message-ID: <a38e20f346810e5ccfaa9b6436166506@plan9.bell-labs.com>

As Norman said, the earliest notation for
pipes used an extension (or abuse) of the semantics
of > and < .

Warren's memory of what Salus wrote (it's on p. 52-53)
is correct about the introduction of | (though I suspect
that McIlroy (whom Salus quotes) is being kind to me
when he said "he [Ken] couldn't bear to reveal my [Doug's]
ugly syntax."  Actually, I was responsible for the
particular < and > syntax as implemented, although the
whole idea came from much earlier on blackboard-only
ideas, and the blackboard was Doug's.

As to the original question: probably the ^ as an alternative
to | (which does seem to be there from the start, i.e. 4th
Edition) did have to do with character-set convenience
on upper-case-only terminals.  The TTY driver accepted
\! as an escape for |, but this was somewhat of a pain.

	Dennis



From sven.mascheck at student.uni-ulm.de  Fri Jan 24 20:32:37 2003
From: sven.mascheck at student.uni-ulm.de (Sven Mascheck)
Date: Fri, 24 Jan 2003 11:32:37 +0100
Subject: [TUHS] The ^ = | ?
In-Reply-To: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10>
References: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10>
Message-ID: <20030124103237.GA2265@wega.rz.uni-ulm.de>

Norman Wilson wrote:

> The earliest UNIX Programmer's Manual to describe shell
> pipelines is the Third Edition, February 1973.

BTW: As the manual source isn't trivial to process on many new systems,
i had converted sh(1) from V3, V4, V6, PWB and V7 into html:
<http://www.uni-ulm.de/~s_smasch/various/bourne/#predecessors>

Sven


From helbig at Informatik.BA-Stuttgart.DE  Sat Jan 25 18:46:47 2003
From: helbig at Informatik.BA-Stuttgart.DE (Wolfgang Helbig)
Date: Sat, 25 Jan 2003 09:46:47 +0100 (MET)
Subject: [TUHS] The ^ = | ?
Message-ID: <200301251051.h0PApbC00222@bsd.korb>

>I've been searching for some reference to the ^ symbol being the same a | in
>Bourne shell. Does anyone remember seeing anything like this? I've searched
>the early manpages to no avail.

Look at
	http://www.ba-stuttgart.de/~helbig/os/v6/doc/index.html
for the Unix V6 sh(I) man page. 

Greetings,

Wolfgang




From grog at lemis.com  Sun Jan 26 11:40:03 2003
From: grog at lemis.com (Greg Lehey)
Date: Sun, 26 Jan 2003 09:40:03 +0800
Subject: [TUHS] The ^ = | ?
In-Reply-To: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10>
References: <pGdcVJPx3rqv.zrrJXl9y@192.168.0.10>
Message-ID: <20030126014003.GA3818@sydney.worldwide.lemis.com>

On Thursday, 23 January 2003 at 22:09:51 -0500, Norman Wilson wrote:
>
> By the Fourth Edition (November 1973) there had evidently been more
> time to think about the syntax; the modern notation is shown, except
> that ^ is allowed as a synonym for |.  I have long guessed that was
> because in those dark days of the past, some upper-case-only
> terminals (remember stty lcase?)  offered no way to type | (and
> perhaps likewise {}`~) but I don't really know.

Certainly the [AK]SR-33 didn't have {}~.  I had this funny feeling
that they did have |, but it doesn't make sense, and I'm probably
wrong.

A thing to recall is that the shape of some of the symbols was
somewhat different in those days.  In particular, _ was a backarrow
(pointing left), and ^ was an uparrow, making it look much more like |
than it appears nowadays.

Greg
--
Finger grog at lemis.com for PGP public key
See complete headers for address and phone numbers


