Flight simulator

[larsbrinkhoff]

DS; FLIGHT 306
DS; STIM 512
DS; SIM 50

Stand alone program using the 340 display.

I have assembled it, and tried running it. There's no indication it's doing anything after being started.

[larsbrinkhoff]

A video with, among other things, a flight simulator using a combination of a PDP-1 and a PDP-10.

https://www.youtube.com/watch?v=coHRMgXW0T0

[larsbrinkhoff]

The closing END directive doesn't specify a starting address.

Anyway the start is at GO. Early on, the program does CONOs to devices 420 and 460. The former could be the Spacewar consoles. But 460 is the KA IMP interface, so that seems strange.

[larsbrinkhoff]

It runs until GO+33, where there's a CONO PI which hangs.

[larsbrinkhoff]

Console commands:

• C - "SENS OF DISP"
• D - Go to DDT
• F - Go fly
• I - Restart
• ^I - Go to IDL1, resets PDL
• L - Flush all I/O
• P - Calibrate pots
• T - Test
• X - Trace scope
• Z - Clear number buffer
• 0-9 - Input number

[larsbrinkhoff]

It hangs spinning on the CONSO IMX in XPT1.

[larsbrinkhoff]

Uses device 764, what is that?

[larsbrinkhoff]

Uses IMX for input. New pot box 1-8, plus unknown channels 150-152 and 171.

[larsbrinkhoff]

With the stub IMX device, I get a display with some gauges.

[larsbrinkhoff]

[eswenson1]

Cool! What are those weird yellow diagonal lines messing up an otherwise reasonable display?

[larsbrinkhoff]

I don't know, but maybe some part of the landscape.

My IMX just returns 0 for all channels, which of course is the lowest possibe value. Perhaps this makes the software go haywire. Maybe a neutral middle value like 2048 would be better.

[larsbrinkhoff]

CC @philbudne, some 340 progress here.

[philbudne]

NEATO!!!

[larsbrinkhoff]

Device 420 indeed seems to be the Spacewar consoles. The program does CONO 420,40, just like Spacewar.

However, enabling the SIMH device makes the gauges display dissapear quickly.

[lbickley]

Hooray for 340 progress!!!

[larsbrinkhoff]

"PDP-6 COMPUTER WITH FLIGHT SIMULATOR DISPLAY", from AI memo 253, page 69.

[larsbrinkhoff]

Sorry, the picture on page 70 is much better. And clearly it's the same program!

[lbickley]

On Wed, 30 May 2018 13:16:20 -0700 Lars Brinkhoff notifications@github.com wrote:

Sorry, the picture on page 71 is much better. And clearly it's the same program!

MUCH better :)

-- 73 AF6WS Bickley Consulting West Inc. http://bickleywest.com

"Black holes are where God is dividing by zero"

[larsbrinkhoff]

Appendix D describes the flight simulator briefly and refers to AI memo 209, which is missing from MIT's website. There should also be a program listing! But it has been left out from the PDF.

[philbudne]

The device SPCWAR controllers used may have been general purpose inputs on a patch panel. And ISTR the "normal" state of all the PDP-6 spacewar switches may have been "on"

[larsbrinkhoff]

This program has some kind of interrupt problem with console input. See rcornwell/sims#68.

[larsbrinkhoff]

The program uses some IMX channels (150-152, 171) which are not documented i hardware memo 4. They are documented in AI memo 117, but I'm not sure those assignments make sense.

[larsbrinkhoff]

Device 764 is close to 760.

        Device 760 of the PDP-6 used to be used for temporary
        direct sense hookups.  (For example, the low six
        bits controlled lights on the hand of the AMF arm.)

[larsbrinkhoff]

Found a problem which made console input hang the program: The KA10 TTY device didn't properly clear its interrupt request after a DATAO.

Unfortunately, fixing this affected the display so now there's just one gague.

[larsbrinkhoff]

Maybe device 764 was used for the oculometer mentioned in AI memo 253.

[larsbrinkhoff]

I talked to David Silver. He says there should be a landscape view. Maybe it's another program, or another part of this program. Maybe one of the commands switches to this view.

He said he would look for a program listing, which would be the lastest version he did. And also AI memo 209.

[larsbrinkhoff]

It seems fixing rcornwell/sims#68 breaks ITS, when building Macsyma.

[eswenson1]

How does it break building Macsyma?

[larsbrinkhoff]

The symptom is similar to how how the build hangs on your Jenkins machine. Which I believe usually occurs somewhere in the Macsyma part?

The build log looks like this:

; Unfasl file ((DSK MUNFAS) BUILDQ UNFASL) not found.
; Unfasl file ((DSK MUNFAS) EDCTL UNFASL) not found.
; Unfasl file ((DSK MUNFAS) EDEXP UNFASL) not found.
; Unfasl file ((DSK MUNFAS) EDBUF UNFASL) not found.
; Unfasl file ((DSK MUNFAS) EDITS UNFASL) not found.
; Unfasl file ((DSK MUNFAS) EDLM UNFASL) not found.

(HYP HYPGEO APLOT2 TEKPLT IFFUN PRINT PLOT3D CPOLY ASUM SERIES NUMTH SCHATC TRG!
RED COMBIN SUPRV COMM COMM2 DSKFN MLISP MEDIT ODE2 PLOT DB DISPLA FORTRA GRAM G!
RIND COMPAR NFORMA OPTIM SCS TRIGI TRIGO CSIMP MAT MATRIX ALGSYS FACTOR RAT3A R!
AT3B RAT3C RAT3D RAT3E NRAT4 EZGCD LESFAC SPGCD SOLVE HOMOG PADE MATRUN MATCOM !
POIS2 POIS3 RATPOI FLOAT RESULT UFACT HAYAT POLYRZ NALGFA NEWDET NISIMP RATOUT !
SIMP SININT SIN RISCH ZERO INMIS LAPLAC PSOLVE OUTMIS MTRACE ARDS VT100 H19 SUS!
PEN OPERS OPS SUBLIS SETS AR MTREE CHAR DOVER TRDUMP TRUTIL TRANSS TRANSL TRANS!
1 TRANS2 TRANS3 TRANS4 TRANS5 TRANSF TROPER TRPRED TRPROP TRMODE TRDATA ACALL F!
CALL EVALW MCOMPI LJOB TRDEBG MTAGS ROMBRG INTPOL NUMER NDIFFQ MLOAD OUTEX MERR!
OR MFORMT ERMSGM LDISP MDOT SUMCON SYNEX RUNTIM UTILS MUTILS OPTION PRIMER DESC!
RI FASDMP IRINTE ASKP LIMIT TLIMIT RESIDU DEFINT SPRDET NEWINV LINNEW EEZ NEWFA!
C ALGFAC CSIMP2 LOGARC RPART APROPO ITENSR CANTEN GENER SYMTRY FILEOP RESET MST!
UFF SPECFN MMACRO BUILDQ EDCTL EDEXP EDBUF EDITS EDLM)

The last command timed out.

[eswenson1]

Not really -- on my Jenkins machine, the timeouts can occur at arbitrary places -- timeouts waiting on output.

[eswenson1]

I wonder if we should try to use XXFILE for the Macsyma build?

[larsbrinkhoff]

Yes, using XXFILE is an interesting idea.

[larsbrinkhoff]

The build breakage has been fixed.

[larsbrinkhoff]

AI memo 238:

A real time airplane simulation has been successfully developed in ITS using this feature.

Maybe Silver's FLIGHT? If so, a later version would have been ported to run under ITS.

[larsbrinkhoff]

David Silver credited at the end of this film:
https://archive.org/details/AIfilms/41-radial.mp4

Comments by Greentblatt:
http://projects.csail.mit.edu/films/aifilms/Podcasts/41-radial.mp3

[larsbrinkhoff]

See also: David Silver's "Little Robot", #1993.