Conference noted::hackers_v1

    By the way, you can also land to the left of the mountain by going
    up and over, but you must be VERY stingy with fuel.
    
    I never knew about the tunnel.  In fact I am surprised that it could
    exist.  The surface data structure, as I recall, consisted of just
    a set of heights of the surface (may have been delta height).  That
    would not allow a tunnel.  Is there special casing in the code?
    I don't remember seeing it.
    
    Burns

I think the following was lost in the transition from old to new notes.

-----

A while back in some notesfile (CTNOTES or the somewhat belly-up
HACKERS), someone was asking for the GT40 down-line loader format.
It is just the standard PDP-11 absolute loader.  However, it uses
six-bit data (ignoring random carriage-returns).

While cleaning out my office, I came upon a paper copy of the
bootstrap.  The relevant parts follow the form feed.  Note that
I'm just typing them in -- and haven't assembled or tested them.

This was taken from Jack Burness's GT40 bootstrap, "BOOTVT S09"
Copyright 1973, 1974, Digital Equipment Inc.

With hope that this reaches the requestor...

Martin.

;
; GET8 gets an eight-bit byte (absldr calls it to read binary data):
; Uses r2, r4 (permanently).  Returns the 8-bit byte in r0.
;
get8:	jsr	pc,getsix		; Get a six-bitter
	mov	r0,-(sp)		; save
	br	getp84			; and off we go
;
; GET84 is entered when the "which byte" counter overflows
get84:	clr	r2			; reset the counter
getp84:	tst	(r2)+			; next position
	add	get8tb-2(r2),pc		; off we go.
;
; Dispatch table, don't move it.
get8tb:	.word	get81-get8tb
	.word	get82-get8tb
	.word	get83-get8tb
	.word	get84-get8tb
;
get81:	jsr	pc,getsix		; get a character now
	mov	r0,r4			; save for a second
	asl	r0
	asl	r0			; shift to left of byte
	aslb	r0
	rolb	@sp			; pack them in
	aslb	r0
	rolb	@sp			; a good 8 bit thing
	mov	(sp)+,r0		; pop and return now
	rts	pc
;
get82:	asl	r0			; worst case, shift 4
	asl	r0
	aslb	r0
	rolb	r4
	aslb	r0
	rolb	r4
	aslb	r0
	rolb	r4
	mov	r4,r0
	mov	(sp)+,r4
	rts	pc
;
get83:	rol	r0
	rol	r0
	ror	r4
	rorb	r0
	ror	r4
	rorb	r0			; final character assembled
	tst	(sp)+			; fudge stack
	rts	pc
;
; Getsix gets one byte from the DL11E
;
getsix:	jsr	pc,getkb		; Any traffic to host?
	tstb	@#DLICSR		; Any traffic from host?
	bpl	getsix			; Loop until we have work to do
	jsr	r5,getbyt		; Get the byte
	  br	getsix			; Ignore <cr>, etc.
	rts	pc			; And return
;
; Getbyt is called from interrupt and non-interrupt modes to obtain
; one byte and strip off garbage.  Result in r0.  Return:
;	 (r5)	if the byte is not alphanumeric
;	2(r5)	if the byte is alphanumeric.
;
getbyt	mov	@#DLIBUF,r0		; Get the byte from the DL11E
	bic	@#177600,r0		; Remove junk and parity
	cmpb	r0,#137			; Is it too big?
	bhi	1$			; If so, return to (r5)
	cmpb	r0,#40			; Is it too small
	blo	1$			; Return to (r5) if so.
	tst	(r5)+			; Alphanumeric, return to 2(r5)
1$:	rts	r5			; Exit.
	.end

        Twas me that asked, Martin. And your response is gratefully
        appreciated. As one of my favourite SF authors would say, "You
        are a blinding flash and a deafening report". 
        
        Ta. Qz

    Re .1:  there was indeed a tunnel, but the only way I could ever
    fly all the way through was by cheating (loc 150).
    
    You can indeed land on the left of the mountain, but it's pretty
    tricky.  After many many hours of practice I was only able to do
    it successfully one out of three tries.
    
    Another game we used to play (at MR1) was to see who could land
    at McDonald's in the smallest amount of elapsed time.  A variation
    on this was to fly on instruments -- we had a piece of cardboard
    cut out to conceal the picture but show the instruments.

	Although he might have been pulling my leg, John Mucci (currently 
LDP group manager?), claims to have been the one to pay for the production 
of the "lunar lander" program. It was originally (and quite successfully) 
used to demonstrate what could be done with the GT4x hardware and later the 
VT11 and VS60. The return on investment analysis could probably be gotten 
out of John.

	The work was done as noted by the contractor in the comments and 
many others have had fun and learned by hacking at the code.

	From what I remember of the sources, the tunnel effect disappeared 
in the move from 16 to 18 bit addressing or was it from 18 to 22 bit? I 
haven't personally hacked at it since 1976 at Naval Research Labs while 
making it drive a fast graphics display (old Megatek Whizzard?).

Regards, Ira Grollman (GSG Systems Engineering)

As I remember, the tunnel was caused by a program bug - if you were going fast
enough, you could fool the collision detector. At that point MOONLANDER was
scanning through random memory interpreting the bits as a map of the terrain.
It crashed when it got into NXM territory.....

Jack Burness is at Masscomp and is the person who designed the software for
their graphics processor.
							Richie

Does anyone have the Moonlander source lying around gathering dust
on a disk somewhere?

You see, I just bought this Atari 1040ST and it has a good graphics
display, and...

Martin.

    I have a listing if you want to type it in.  You realize it is pdp-11
    macro, and uses the GT40 display processor.
    
    Is there no "product" like it for micros?
    
    								/alan

    RE: .-1
    	But there is HISTORY at stake here!

So Peter Quodling walked into my office today, coping with being
right-side up (and not in the Australian springtime) quite well.

Moonlander?  Of course I've got it.  SET HOST to Australia and
set the protection.  The blast off a 24,000 mile $COPY.

Mail to me if you want a copy.  J. F. Mucci did hack on the sources
to force a "logout" down the throat of the remote host.

Martin.

    There's a pascal version somewhere, but was written to VAX/VWS
    graphics,and runs on VAXstations.
    
    Jeff
    

A pascal version would be nice to look at -- Jack Burness' code (and the
GT40 display list hacking) isn't exactly easy to follow.

Pointers to the code would be greatly appreciated.

Martin.

    Here's a pointer to the black and white DEClander. Color was added
    by Mike (STAR::) Rosenblum, and the executable is part of the VWS
    V3 demo kit, which can be found on BULOVA::VWS$KIT:VWSDEMO030.A
    as a VMSINSTAL save set.
    
    	Jon
    
              <<< STAR::DISK$SYSKITS:[NOTES$LIBRARY]VWS.NOTE;2 >>>
                          -<  MicroVAX Workstations  >-
================================================================================
Note 975.2                    Moon Lander Program?                        2 of 2
ADVAX::C_YOST                                         3 lines   7-NOV-1986 08:20
                                     -<   >-
--------------------------------------------------------------------------------

    look in JACOB::USER1:[VAXSTATION.UVMS.VWS020.GAMES.DECLSRC]
    
    clay