Computer Old Farts Forum
 help / color / mirror / Atom feed
* [COFF] [TUHS] Will pdp 11/04 run unix?
@ 2020-07-31 13:56 jnc
  2020-07-31 15:32 ` lars
  2020-07-31 18:53 ` clemc
  0 siblings, 2 replies; 9+ messages in thread
From: jnc @ 2020-07-31 13:56 UTC (permalink / raw)


    > From: Angelo Papenhoff

    > Well the TV-11 is a tough question. I originally wrote an 11/05 emulator
    > because some document said it was an 11/10 (which is the same thing).
    > But other sources claimed it was an 11/20.

Hmmm. My memory was that it was an -11/05-10 (they are identical, except for
the paint on the front panel; and I don't recall it in enough detail to say),
but perhaps I'm wrong?

Or maybe it was an -11/20 early, and then it got replaced with an -11/10? (I
have a _very_ vague memory that the XGP's -11 was a /20, bur I wouldn't put
much weight on that.)

Moon or TK or someone might remember better.

     Noel



^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 13:56 [COFF] [TUHS] Will pdp 11/04 run unix? jnc
@ 2020-07-31 15:32 ` lars
  2020-07-31 18:53 ` clemc
  1 sibling, 0 replies; 9+ messages in thread
From: lars @ 2020-07-31 15:32 UTC (permalink / raw)


Noel Chiappa wrote:
> Hmmm. My memory was that it was an -11/05-10 (they are identical,
> except for the paint on the front panel; and I don't recall it in
> enough detail to say), but perhaps I'm wrong?

I couldn't say for sure.  I have seen conflicting information in sources
dating from the 70s when the TV-11 and XGP-11 were very much in use.

For several reasons, I believe the TV-11 was the first machine attached
to the 10-11 interface, and the XGP-11 came second.  This would lend
some weak support for the theory that the first would be a 11/20 and the
second a 11/10.

> Or maybe it was an -11/20 early, and then it got replaced with an
> -11/10? (I have a _very_ vague memory that the XGP's -11 was a /20,
> bur I wouldn't put much weight on that.)

Replacing or changing machines around would explain the confusion, but I
haven't seen anyone remembering such events, or any written record to
support that.

> Moon or TK or someone might remember better.

I did bring it up with TK at some point.  He said the TV-11 must have
been a 11/20, because at the time there was no other PDP-11 model
available.  But as far as I can see from file timestamps, the AI lab TV
project was started in 1973 or so when the 11/10 would have been
shipped.


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 13:56 [COFF] [TUHS] Will pdp 11/04 run unix? jnc
  2020-07-31 15:32 ` lars
@ 2020-07-31 18:53 ` clemc
  2020-07-31 19:38   ` lars
  1 sibling, 1 reply; 9+ messages in thread
From: clemc @ 2020-07-31 18:53 UTC (permalink / raw)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2831 bytes --]

a little more history about the XGPs of the 70's...

On Fri, Jul 31, 2020 at 10:16 AM Noel Chiappa <jnc at mercury.lcs.mit.edu>
wrote:

>     Or maybe it was an -11/20 early, and then it got replaced with an
> -11/10? (I
> have a _very_ vague memory that the XGP's -11 was a /20, bur I wouldn't put
> much weight on that.)
>

I was not involved when it was stood up, but FWIW the CMU mods to a Xerox
'Long Distance Xerography - LDX' (FAX) system was definitely was an 11/20
on the original one.  The XGP as it was called, was the first Xeroxgraphic
printer at 200 dpi attached as an 'output device' to the PDP-11/20.   A couple
of my friends and I did some of the programming of the graphics PDP-11 at
one point (you may remember the LDX used rolls of paper, with a razor to
cut when the page was complete.  The PDP 11 was hacked it to recognize Mike
Shamos's PPN and cut his paper every 1.5 inches, giving him strips of
output, but work fine if the same job was printed by anyone else - story
for another day as to why).  BTW, an interesting factoid about the LDX, is
that it was not a laser printer.  It used a CRT, the same idea Tektronix
would use shortly thereafter for their hardcopy printers.

Anyway, I'm pretty sure the copies at Stanford (Jan '73), and MIT (was the
3rd in the series and a little later) also used 11/20s or maybe 11/15's
which was the OEM version of the 20 as it was March '72 when the CMU XGP
was first stood up.   PARC made one too for MAXC shortly after that but I
think that had a Nova in it originally.  That said, Jim Teter
might remember but I think that only a handful of them was them
stood up, but most used Jim's interface/mod.  I think DEC one, as at least
the PDP-16/RTM handbook, was set using it/maybe a few others.   I also am
under the impression that after the original wire-wrapped prototype worked,
the DR-11C to Xerox machine driver logic was an early numbered 'Teter Toy"
( My memory is the designers went to PC board quickly because the WW board
would not let them close the LDX cabinet or a shelf or something like that
). I also have memories of soldering/assembling some sort interface board
for XGP in the summer of '78 under the watchful eye of Teter which we were
assembling for some reason (a bunch of us were working as systems
operators/programmers and tech's -- *i.e.* grunt work).

I do have copies of the pictures of Teter printing CMU diploma replicas on
toilet paper with it in the late '70s.

Also, another fun XGP story, Chuck Geschke (Wulf’s first PhD student,
founder of Adobe)  filed the first PhD printed on the XGP but the CMU
library would not accept it because they wanted the original ;-)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20200731/5b3868a6/attachment.htm>


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 18:53 ` clemc
@ 2020-07-31 19:38   ` lars
  2020-07-31 19:40     ` clemc
  2020-07-31 20:00     ` bgbaumgart
  0 siblings, 2 replies; 9+ messages in thread
From: lars @ 2020-07-31 19:38 UTC (permalink / raw)


Clem Cole wrote:
> Anyway, I'm pretty sure the [XGP] copies at Stanford (Jan '73), and
> MIT (was the 3rd in the series and a little later) also used 11/20s or
> maybe 11/15's which was the OEM version of the 20 as it was March '72
> when the CMU XGP was first stood up.

Thank you.  That's one more "vote" in favour of 11/20.  In which case
the TV-11 ought to be an 11/10 which was our original guess.  I don't
think it matters to the software; it should run just as well on either
model.

I have seen MIT files which describe the Stanford hardware, so it seems
their inspiration came from there.  The earliest timestamp is from
February 1973.

I got the impression the Stanford XGP had a PDP-6/10 IO bus interface
rather than going through a PDP-11.  I'm CC'ing Bruce Baumgart.


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 19:38   ` lars
@ 2020-07-31 19:40     ` clemc
  2020-07-31 20:00     ` bgbaumgart
  1 sibling, 0 replies; 9+ messages in thread
From: clemc @ 2020-07-31 19:40 UTC (permalink / raw)


Very possible, I never saw the insides of theirs.

On Fri, Jul 31, 2020 at 3:38 PM Lars Brinkhoff <lars at nocrew.org> wrote:

> Clem Cole wrote:
> > Anyway, I'm pretty sure the [XGP] copies at Stanford (Jan '73), and
> > MIT (was the 3rd in the series and a little later) also used 11/20s or
> > maybe 11/15's which was the OEM version of the 20 as it was March '72
> > when the CMU XGP was first stood up.
>
> Thank you.  That's one more "vote" in favour of 11/20.  In which case
> the TV-11 ought to be an 11/10 which was our original guess.  I don't
> think it matters to the software; it should run just as well on either
> model.
>
> I have seen MIT files which describe the Stanford hardware, so it seems
> their inspiration came from there.  The earliest timestamp is from
> February 1973.
>
> I got the impression the Stanford XGP had a PDP-6/10 IO bus interface
> rather than going through a PDP-11.  I'm CC'ing Bruce Baumgart.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20200731/4d8d6331/attachment.htm>


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 19:38   ` lars
  2020-07-31 19:40     ` clemc
@ 2020-07-31 20:00     ` bgbaumgart
  2020-07-31 20:05       ` clemc
  1 sibling, 1 reply; 9+ messages in thread
From: bgbaumgart @ 2020-07-31 20:00 UTC (permalink / raw)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1750 bytes --]

Lars et al
-
Lynn Quam is credited with building the XGP hardware interface to the SAIL PDP-6.


A few lines from
Version #1 of Quam’s file RESUME[DOC,PDQ] say

< quote >

Nov. 1972 to
Feb. 1973\jFull-time research associate in computer science.
Received a grant from the NASA Viking Mission thru 
Cornell University for the analysis of candidate landing
sites for the Viking mission.\.

\jDesign and debugging of an interface between a PDP-10 (PDP-6)
and and a Xerox Graphics Printer (XGP).\.

< Unquote />

Prior to the Stanford interface,
Quam built a Nova to XGP interface at Xerox Parc
As a part time employee while also working at SAIL.
-
Bruce



p.s. Lynn Quam’s log in code is PDQ


> On 31 Jul 2020, at 12:38 PM, Lars Brinkhoff <lars at nocrew.org> wrote:
> 
> Clem Cole wrote:
>> Anyway, I'm pretty sure the [XGP] copies at Stanford (Jan '73), and
>> MIT (was the 3rd in the series and a little later) also used 11/20s or
>> maybe 11/15's which was the OEM version of the 20 as it was March '72
>> when the CMU XGP was first stood up.
> 
> Thank you.  That's one more "vote" in favour of 11/20.  In which case
> the TV-11 ought to be an 11/10 which was our original guess.  I don't
> think it matters to the software; it should run just as well on either
> model.
> 
> I have seen MIT files which describe the Stanford hardware, so it seems
> their inspiration came from there.  The earliest timestamp is from
> February 1973.
> 
> I got the impression the Stanford XGP had a PDP-6/10 IO bus interface
> rather than going through a PDP-11.  I'm CC'ing Bruce Baumgart.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20200731/070aa95e/attachment.htm>


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
  2020-07-31 20:00     ` bgbaumgart
@ 2020-07-31 20:05       ` clemc
  0 siblings, 0 replies; 9+ messages in thread
From: clemc @ 2020-07-31 20:05 UTC (permalink / raw)


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2093 bytes --]

OK, that makes sense.  As I said, I thought MAXC was a Nova.  So,
Stanford used a direct interface to the 10 and the CMU, DEC and MIT ones
used an 11/20
I don't know how many others were stood up or how.

On Fri, Jul 31, 2020 at 4:00 PM Bruce Baumgart <bgbaumgart at mac.com> wrote:

> Lars et al
> -
> Lynn Quam is credited with building the XGP hardware interface to the SAIL
> PDP-6.
>
>
> A few lines from
> Version #1 of Quam’s file RESUME[DOC,PDQ] say
>
> < quote >
>
> Nov. 1972 to
> Feb. 1973\jFull-time research associate in computer science.
> Received a grant from the NASA Viking Mission thru
> Cornell University for the analysis of candidate landing
> sites for the Viking mission.\.
>
> \jDesign and debugging of an interface between a PDP-10 (PDP-6)
> and and a Xerox Graphics Printer (XGP).\.
>
> < Unquote />
>
> Prior to the Stanford interface,
> Quam built a Nova to XGP interface at Xerox Parc
> As a part time employee while also working at SAIL.
> -
> Bruce
>
>
>
> p.s. Lynn Quam’s log in code is PDQ
>
>
> On 31 Jul 2020, at 12:38 PM, Lars Brinkhoff <lars at nocrew.org> wrote:
>
> Clem Cole wrote:
>
> Anyway, I'm pretty sure the [XGP] copies at Stanford (Jan '73), and
> MIT (was the 3rd in the series and a little later) also used 11/20s or
> maybe 11/15's which was the OEM version of the 20 as it was March '72
> when the CMU XGP was first stood up.
>
>
> Thank you.  That's one more "vote" in favour of 11/20.  In which case
> the TV-11 ought to be an 11/10 which was our original guess.  I don't
> think it matters to the software; it should run just as well on either
> model.
>
> I have seen MIT files which describe the Stanford hardware, so it seems
> their inspiration came from there.  The earliest timestamp is from
> February 1973.
>
> I got the impression the Stanford XGP had a PDP-6/10 IO bus interface
> rather than going through a PDP-11.  I'm CC'ing Bruce Baumgart.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://minnie.tuhs.org/pipermail/coff/attachments/20200731/00f3e11f/attachment-0001.htm>


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
@ 2020-07-31 21:26 jnc
  0 siblings, 0 replies; 9+ messages in thread
From: jnc @ 2020-07-31 21:26 UTC (permalink / raw)


    > From: Lars Brinkhoff

    > I have seen conflicting information in sources dating from the 70s when
    > the TV-11 and XGP-11 were very much in use.
    > For several reasons, I believe the TV-11 was the first machine attached
    > to the 10-11 interface, and the XGP-11 came second.  This would lend
    > some weak support for the theory that the first would be a 11/20 and the
    > second a 11/10.

Yeah, but Clem's note reinforces my vague memory that the XGP-11 was an
-11/20.

I wish we had a picture of the Knight TV system (the system, not a
terminal). It's a extremely significant system - I believe it my have been the
first bit-mapped computer display system ever; and thus the prototype, in some
sense, for the display of every single personal compupter (including phones)
now extant - and so there _ought_ to be a photo of it.  But looking online for
a while, I can turn up almost nothing about it! (I guess we should do a page
about it on the CHWiki...)

(Repeat my prior grump about how the AI Lab did all sorts of ground-breaking
stuff, because it was just 'tools', and not their main research focus, it's
hard to find out about a lot of it, e.g. the inter-ITS network file
system.)

But if you can find an image, even a low-res picture of that end of the AI Lab
machine room, we can tell what model the TV-11 is - early 11's had inteagrated
front panels, which are different for every model:

  http://ana-3.lcs.mit.edu/~jnc/tech/pdp11/PDP-11_Models.html

so you don't even need to be able to read anything to tell a /20 from a /10.
It was in a dual (I think - maybe triple, it's been a looooooong time :-)
rack which IIRC was along the side wall (i.e. the short building side) next to
the AI KA10 (which was sort of along the long wall, up in the corner).

I don't know if the XGP-11 code is still extant (my copy of the ITS filesystem
is offline right at the moment), but even if we look at the code, I'm not sure
we could tell; there are some _very minor_ programming differences between the
/20 and /10 (e.g. V bit on SWAB) - see the table at thd end of the PDP-11
Architecture Handbook - but I'd be aurprised if the code used any.

Surely there has to be _some_ picture of the machine room which shows it, even
if in the background.


    > I did bring it up with TK at some point.

Try RG, too.

    Noel


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [COFF] [TUHS] Will pdp 11/04 run unix?
       [not found] ` <20200731071008.GA33933@indra.papnet.eu>
@ 2020-07-31  7:57   ` lars
  0 siblings, 0 replies; 9+ messages in thread
From: lars @ 2020-07-31  7:57 UTC (permalink / raw)


Angelo Papenhoff wrote:
> Noel Chiappa wrote:
>> If the KE11 is needed to run some application on the -11/04, there
>> are KE11-B's (program compatible, but a single hex card) available,
>> ISTR.  For emulation, something (SIMH?) supports it, since the TV -11
>> on ITS (now running in emulation,I'm pretty sure) uses it.
>
> Well the TV-11 is a tough question. I originally wrote an 11/05 emulator
> because some document said it was an 11/10 (which is the same thing).
> But other sources claimed it was an 11/20.

To clarify, the emulated TV-11 is *not* in any way based on SIMH.  There
are more machines to potentially hook up to the 10-11 interface, but I'm
quite unsure if SIMH is the right vehicle for those.

But this is now clearly out of TUHS territory.  CC to coff only.


^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2020-07-31 21:26 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-07-31 13:56 [COFF] [TUHS] Will pdp 11/04 run unix? jnc
2020-07-31 15:32 ` lars
2020-07-31 18:53 ` clemc
2020-07-31 19:38   ` lars
2020-07-31 19:40     ` clemc
2020-07-31 20:00     ` bgbaumgart
2020-07-31 20:05       ` clemc
  -- strict thread matches above, loose matches on Subject: below --
2020-07-31 21:26 jnc
     [not found] <20200730213720.109C018C0A5@mercury.lcs.mit.edu>
     [not found] ` <20200731071008.GA33933@indra.papnet.eu>
2020-07-31  7:57   ` lars

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).