PDP-6 machines

[aap]

There are now two issues about information and photos/videos of ITS machines in the ITS repo. In addition @philbudne has a list of of PDP-6s here: http://www.ultimate.com/phil/pdp10/pdp6-serials.html I open this issue to have a discussion not tied to ITS about this here.

# 1

I wanted to mention that the image at the top of phil's page (of which there was a higher res version somewhere - i have it here) looks not at all like the final PDP-6. It seems to be the prototype described in this brochure. I mailed Gordon Bell about it some time ago but he said

There was no such a thing as a proto. [...] No 1 was used as our development machine.

I'm not really sure how to interpret this.

# 2

There are quite a few videos on https://www.csail.mit.edu/videoarchive/history/aifilms/ that show the MIT PDP-6 and it's robot arms and whatnot.

# 4

I would also like to add a photo of the UWA PDP-6: http://www.iucr.org/__data/assets/image/0005/57677/1967_UWA_PDP6.jpg (source) And I was sent photos of the remains of the front panel of this machine by Peter Dreisiger (photos by Alex Reid): http://pdp-6.net/images/pdp6_1.jpg http://pdp-6.net/images/pdp6_2.jpg

I find this machine particularly puzzling. The indicator panel in bay 1 (the leftmost cabinet of the APR) has a few lamps "missing". For reference the full indicator panel is shown on page 3-3 here If you take the full grid of lamps and remove the ones that are missing in the UWA machine, it doesn't really make sense, so it looks like the lamp positions were different altogether. It seems to be the same machine as this: 1 2 3 apparently still standing at DEC at this time. This photo seems to have the full lamp grid so I'd say it's probably not the UWA machine. In this context it may be noteworthy that the PDP-6 flow diagrams and schematics do not mention the A LONG flip flop at all (set when indexing or indirection happens for an instruction) while the instruction manual vol 1 does describe it. I don't know what to make of this.

# 16 2

Some more photos from Stanford: MIT: http://media.csail.mit.edu/picture.php?/3896/category/173 (are these spacewar controllers in the back?) http://media.csail.mit.edu/picture.php?/3811/category/173 http://media.csail.mit.edu/picture.php?/3839/category/173 (not much visible) http://media.csail.mit.edu/picture.php?/3885/category/173

The CHM has two front panels (hard to say where they're from...stanford? but why two?): http://www.computerhistory.org/collections/catalog/102732975 http://www.computerhistory.org/collections/catalog/102732961

[philbudne]

In addition @philbudne has a list of of PDP-6s here: http://www.ultimate.com/phil/pdp10/pdp6-serials.html

Just yesterday I found a DECUS (DEC user group) newsletter that described the locations of the first 8 systems!

I open this issue to have a discussion of this here.

# 1

There was no such a thing as a proto. [...] No 1 was used as our development machine.

The DECUS newsletter describes #1 as a prototype.

Bob Clements said it was originally built as a prototype (he called it

0) with Germanium transistors then changed over to Silicon (and then

called #1). That may just have been RCC's terminology.

# 2

There are quite a few videos on https://www.csail.mit.edu/videoarchive/history/aifilms/ that show the MIT PDP-6 and it's robot arms and whatnot.

Neat!

# 4

I would also like to add a photo of the UWA PDP-6: http://www.iucr.org/__data/assets/image/0005/57677/1967_UWA_PDP6.jpg (source) And I was sent photos of the remains of the front panel of this machine by Peter Dreisiger (photos by Alex Reid): http://pdp-6.net/images/pdp6_1.jpg http://pdp-6.net/images/pdp6_2.jpg

Thanks! I may need to move to a page per machine for supporting documentation!

I find this machine particularly puzzling. The indicator panel in bay 1 (the leftmost cabinet of the APR) has a few lamps "missing". For reference the full indicator panel is shown on page 3-3 here If you take the full grid of lamps and remove the ones that are missing in the UWA machine, it doesn't really make sense, so it looks like the lamp positions were different altogether. It seems to be the same machine as this: 1 2 3 apparently still standing at DEC at this time. This photo seems to have the full lamp grid so I'd say it's probably not the UWA machine. In this context it may be noteworthy that the PDP-6 flow diagrams and schematics do not mention the A LONG flip flop at all (set when indexing or indirection happens for an instruction) while the instruction manual vol 1 does describe it. I don't know what to make of this.

I recently noted that Alex Reid's photos show two different display locations (one in "before it left the factory" and "towards end") and the display above the switches in "on delivery". I wasn't looking at the switches at all!

Maybe you should show/compare the different lamp/switch configurations side by side on your page?

# 16

Some more photos from Stanford: http://media.csail.mit.edu/picture.php?/3896/category/173 (are these spacewar controllers in the back?) http://media.csail.mit.edu/picture.php?/3811/category/173 http://media.csail.mit.edu/picture.php?/3839/category/173 (not much visible) http://media.csail.mit.edu/picture.php?/3885/category/173

The above are MIT.

The CHM has two front panels (hard to say where they're from...stanford? but why two?): http://www.computerhistory.org/collections/catalog/102732975 http://www.computerhistory.org/collections/catalog/102732961

One of the descriptions says: The object is a DEC PDP-6 programmer's panel. Scratched on the inside of the object is "Given to Gordon Bell May 1966."

Surely CHM must have some provenance info for the items. Perhaps Al Kossow can help.

[aap]

Maybe you should show/compare the different lamp/switch configurations side by side on your page?

Will do.

The above are MIT.

Of course.... makes sense since it's an MIT server. For some reason I was under the impression this was the stanford machine. I originally found those photos somewhere else, maybe that was the reason of my confusion.

The object is a DEC PDP-6 programmer's panel. Scratched on the inside of the object is "Given to Gordon Bell May 1966."

Good catch, I missed that. I just wrote an email to ask if it's true he was given a PDP-6 panel, that's already answered and even dated then.

[aap]

Here now is a comparison of the indicator panels with the lamps removed and an explanation of the respective flip-flop: http://pdp-6.net/notes6.html The thing that irritates me is that A LONG (which was apparently added rather late, it serves no function other than to be displayed on the lamp) is there while other apparently more useful flip-flops aren't.

[aap]

Maybe this info from Gordon Bell (via email) is of interest too:

When the KA10 came out, I believe that MIT and Stanford traded in their 6s and we just left them there.

[larsbrinkhoff]

So maybe that's why the DM group got a used 6.

[philbudne]

The PDP-10 (KA) came out in 1968, so I imagine the DM group got a '6 that had been (physically) traded in.

I went through all the DECUSCOPE (DECUS newsletters) thru 1968, which include the announcement of the KA, and found: the locations for the first eight serials, and the (initial) configurations of some of them.

I now have names of 19 PDP-6 locations (counting LLL twice, and not counting DM), and serial numbers for the first 8, plus SAIL (#16). It's possible LLL's second CPU was a trade-in, but it's noted that they upgraded to KA10s, and the list of systems that were connected to OCTOPUS (the "network" that the PDP-6es controlled) they were anything but poor. Something that just struck me about the LLL photos, is that they show no DECtape units!

[aap]

@philbudne photos of the PDP-6 from Australia: http://aap.papnet.eu/pub/img/lcm/pdp-6/gallery.html Some are of the 340 display (without the tube, that's sitting on a table next to it). It was partially saved from being scrapped and is now at LCM in Seattle. And some more photos that Stephen Jones had sent to me (probably when it was still in Australia): http://aap.papnet.eu/pdp6-lcm/ CHM apparently has the Moby-memory

[philbudne]

Yow! Do they plan on trying to restore it?

I have it as S/N 4 at http://www.ultimate.com/phil/pdp10/pdp6-serials.html

[aap]

There are plans to do something with it, but most of it was actually scrapped, so it's not quite clear yet what will happen exactly.

[larsbrinkhoff]

[larsbrinkhoff]

I hijacked this thread to upload photos of the printed 3D model.

[larsbrinkhoff]

Finished. I'll call this serial number 27.