[TUHS] v6 archaeology continued - dds?

[Will Senn via TUHS]

One wrinkle I hit in the /cubic -/ path is that, after creating the 
pipe, forking, and doing the close/dup setup, the surviving cubic side 
executes three consecutive sys 0 instructions before closing fd 0 and 
opening /dev/ttyc. Under V3 and later, sys 0 is the indirect-syscall 
mechanism, so the sequence makes no sense and immediately produces Bad 
system call and dumps core.

In V1, however, syscall 0 was rele, essentially a yield/relinquish 
operation. In context, three rele calls after the fork/pipe setup and 
before attaching /dev/ttyc look plausibly like crude synchronization. 
Effectively giving the other process a few chances to run before cubic 
continues. Does that pattern ring a bell with anyone, or does anyone 
know whether this DDS path dates back to the V1/early-V2 syscall ABI?

Thanks,

Will


On 9/20/26 7:55 PM, Will Senn wrote:
>
> Hi All,
>
> After getting graphics and speech working, I turned to games and 
> seeing what graphics were in play with the distribution beyond 
> character graphics. Seems like Spacewar! worked on PDP1, surely some 
> the distributed games would do graphics. Well, I pulled up pristine v6 
> and looked for game source - nowhere to be found of course, until that 
> is, the excavation of the DMR tapes, where most of the games have a 
> version of source as close to original as could be desired. Sadly, not 
> graphics, per se.
>
> Unfortunately, I wasn't able to find source for cubic, so I turned to 
> ai for a insight and had it disassemble the code and found that in 
> addition to the text output that it was also optionally opening 
> /dev/ttyc and exec'ing /usr/bin/dds. These are interfaces to a 
> picturephone, as mentioned in the Ritchie, Thomposon article:
>
> There are also several one-of-a-kind devices including a Picturephone® 
> interface, a voice response unit, a voice synthesizer, a 
> phototypesetter, a digital switching network, and a satellite 
> PDP-11/20 which generates vectors, curves, and characters on a 
> Tektronix 611 storage-tube display.
>
> Wanna see one (scroll down to pg. 141):
>
> https://www.telephonecollectors.info/index.php/browse/document-repository/catalogs-manuals/bell-system-we/pubs-docs/blr/11616-69jun-blr-p138-picturephone/file
>
> Apparently, by V3, this was has been tech (from ddsput(3)):
> These routines *used* to provide an interface to the Display Data Set, 
> a peculiar device which can be called by Picturephone sets and which 
> will display some of the ASCII character set and certain other 
> graphics on the Picturephone screen.
>
> DDSPUT and DDSINIT are the referenced routines.
>
> Anyhow, it was disappointing to find that it wasn't putting any 
> graphics to the picturephone, just a text representation. However, it 
> did print the full full 4x4x4 board to the display, vs the paltry 3 
> digits we get in the terminal, something along the lines of:
>
>   1111 2222 3333 4444
>   1234 1234 1234 1234
>
> 1 --x- ---- ---- ----
> 2 ---- -o-- ---- ----
> 3 ---- ---- x--- --x-
> 4 ---- ---- ---- o---
>
> vs
>
> 113
> 222
> 331
> 433
> 441
>
> The digits correspond to plane, row, column:
> * first digit = which 4×4 plane/block (1–4)
> * second digit = displayed row (1–4)
> * third digit = position within that block (1–4)
>
> My current thinking is to get cubic working again by taking 
> essentially the same approach I used for the V6 speech and graphics 
> work: preserve the old UNIX-facing interface and replace the vanished 
> hardware beyond it. Rather than trying to reproduce an actual 
> Picturephone DDS, I would recreate the /usr/bin/dds -> /dev/ttyc path 
> and translate the DDS character/cursor operations onto the same SIMH 
> display machinery I am already using for the vt work. That would put 
> cubic back on a persistent graphical display, using the same general 
> graphics path that is already supporting BASIC and FORTRAN through the 
> V6 vt implementation.
>
> What I would ask of y'all is - do you remember the display, have 
> source, know provenance, have interesting story, etc?
>
> Regards,
>
> Will
>
>
>