Dutchman2000

Here is the demo that was programmed by John Seghers back in 1983. He wrote it to show Atari management that the 5200 was perfectly capable of creating 3D rotating asteroids as the 7800 was. Nothing further ever came of it. 3D Asteroids (5200).bin

Being as it’s not a 2600 game, correct.

It is listed in the collection as being a “Atari Reimagined” game so it is not 2600 code.

The new game was written using modern tools and made to look like a 2600 game. It is based on Tod Frye’s original concept and plays just like the other 3 games in the series. I was hoping there would be a comic book as well to complete the story but we didn’t get that.

I’ll share it with those who need it to help create new products.

If the arcade source is needed, I have it.

It definitely wasn’t Roklan or On Time Software. Since Keithen remembers talking to the programmer at Atari it seems unlikely it was programmed elsewhere.

He did find his stash of disks years ago but sadly 5200 ET was not among them.

Yes, they developed Super Missile Attack as a speed up kit for Missile Command. Atari promptly sued them and after the case was settled went to work for Warner Communications doing engineering work.

And some notes on the changes and why they were made: Changes to Marble Madness II Proposed 4/23/91 A. From the Focus we learned that the trakball is the more intuitive control to roll a marble, and that it is the desired control for the high-end player. But, the joystick was percieved as an easier control for a beginner to learn the game. Thus, we would like to change the trakball control to a joystick, and see if we gain a wider audience. In order to gain the variable speed of the trakball control we need to add a "turbo" button with the joystick. From in-house play and the Focus, the "turbo" was seen as a fun new feature. B. There was no objection to the "cuteness" level of the game from the Focus. But the players did feel the tranformation to Marble Man was "hokey, "stupid," and meaningless. We will remove "Marble Man" from the entire game. This means the Title screen, cabinet artwork, pinball spell outs, and goal transformation all need to be changed. There were objections to the "scream" of acid death, and the fall grunt from the marble. We would like to change these voices to something more comical rather than realistic. C. There were general elements of confusion. 1. Players did not find the power towers until many hours of play. We can modify the tower graphics by placing the animating power sphere on top of the tower. 2. Players were confused by score and time bubbles that look like marbles. Since the flags have values on them there is no need to give a score bubble, instead we will merely change the color of the flag to the player color to let him know he recieved the points. Second, we will replace the "100" flags that lead the player to the goal with miniture checkered goal flags. These new flags will have no collision and award no score value. (They will provide a clear path to the goal.) 3. It was not clear to players that after a score threshold was reached the player recieved extra time. All score flags will be worth from 1,000 to 5,000 points with the score threshold being every 5,000 points. This will provide a more immediate feedback of the "score gives time" rule. 4. There was a lack of basic instructions.(ie. We never told the player to get to the goal before the clock reaches 0.) We will add the necessary instructions to the "Start Screen." 5. Also there was confusion concerning how much time the player thought he had to complete the wave from the wave start. We will place the track messages on a black background and animate a 5 digit from the track clock to every players' clock. D. There were some specific multi-player problems mentioned at the Focus. 1. One,Players' did not like going to the trouble of breaking a tower and then losing the power to someone else. We can guarantee that the player who breaks the tower recieves the power right away. 2. Also the scrolling was annoying in multiplayer play. It was apparent that players did not want to hold others up by not being able to keep up, and they did not want to be slowed by players who could not keep up with them. If we return the scroller to the original Marble Madness type scrolling, the lead marble will always control the screen position and poorer players will be moved forward for free. (Marble Madness took 5 seconds off the player moved ahead) This means players will not be held up and will not hold up others. E. Players wish to have more powers that last longer. Specifically they wanted more Helicopter Power. We can give them more Helicopter Power for longer if we restrict the scroller. One solution is to simply not let the lead marble control the scroller if he has Helicopter power. F. The first round ramping was a significant problem in the field tests. We have far too many people quitting before they reach the first pin. We are going to redo enemies in the first round to more closely match the orignal Marble Madness design, where the difficulty was in learning terrain with an occaisional enemy. We will give the players more continue time in the 1st round. In order to get a new field test we need the following: New control panel assembly (harness, graphics, 3 joysticks, 6 buttons) New MOB/PF graphics New program/adpcm/audio Note: For every unit to be placed on field test we need the above. We estimate it will take 2 weeks to complete the program changes that are necessary to re-test the game. So, Monday, May 6, we will place a new version in the common area for 1 week of game pounding. On Monday, May 13, we will be able to place a new version out on field test.

Some more information on the hardware: MARBLE MAN HARDWARE DESCRIPTION Rev.5 CREATED: 910915 by Tim Hubberstey REVISED: 910403 PLEASE ADVISE OF ANY ERRORS OR OMISSIONS Microprocessor: . ONE 68000 running @ 7.16 MHZ or 14.316Mhz (or a 68010 @ 7.16 only) . Program ROM: - 1 Mbytes using 1 meg Eproms including bank-switchable security area. At 14MHz one wait state is required. . Program RAM: - up to 64Kbytes of dedicated working RAM. . Interrupts: - Autovectored: Sound port - level 6 VAD interrupt - level 4 . Input/output: - JAMMA edge connector - 2 or 3 trackball inputs (center player deleted for 2) - 8 button inputs including 2 start buttons - Sound port read/write/reset/self-test (JSAIII PCB) . Watchdog: - Barks if not stroked at least once every 7 frames (117ms) . EEPROM: - 2K bytes (occupying 2K words, hi byte not there) - Must UNLOCK before writing data (to prevent power on false writes) - Must be LOCKED before reading data. (Any low byte write LOCKS) - Must wait 10mS min. between writes. . Video: - 64K or 16K bytes divided as : Scrolling playfield 8K bytes Motion Object 8K bytes Working RAM 48K bytes, only if 64K total - 2K bytes of color RAM in 4 512-byte banks (occupying 2K words, lo byte not there) Graphics: A. Display: - 336 x 240 pixels - (42 x 30 stamps) standard res. monitor. B. Playfield: - Size: 64 stamps x 64 stamps - Visible screen: 42 stamps x 30 stamps - Stamp size: 8 pixels x 8 pixels x 8 bits/pixel - Independent smooth scrolling horizontally & vertically. - 16K stamps Picture: D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ | H | PIC # | +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ Playfield 1 picture 7D8000-7D9FFF H = Horizontal flip (can ba a picture bit if required) PIC # = Picture number. (15 bits). PLAYFIELD PICTURE COORDINATES: PLAYFIELD HORZ. SCROLL REGISTER. | pixels--> 0 | 1FF | |<--------------------|-------------------------------->| | | | | v | 0 --- ----------------------|---------------------------------- ^ |__|<--7D8000 | 7D9F80-->|__| ^ | |__|<--7D8002 | 7D9F82-->|__| | | | | | | | | | | | VERT. | | v<----42 STAMPS----> | 64 -------|-------------------->|------------------ | STAMPS SCROLL | | ^ | | | | | | | | | | | REG | | 30 | | | | | | STAMPS | SCREEN | | | | | | | | | | | | V |_________________| | | | |__ __| | v |__|<--7D807C 7D9FFC-->|__| v 1FF ___ |__|<--7D807E_______________________________7D9FFE-->|__| <----------------------64 STAMPS------------------------> C. Motion Objects: - 56 M.O.'s (1 stamp wide) visible on any scan line - M.O.s can be up to 8 x 8 Stamps (64 x 64 pixels) - M.O. processing is pipelined to allow more than 56 stamps worth to be processed per line. - Last in highest priority motion object line buffer. - "Linked" M.O.'s - A new starting link pointer (SLIP) for every 8 scan lines of the playfield - H and V positions relative to a variable reference point. - 1024 motion object descriptors - Stamp size: 8 pixels x 8 pixels x 4 bits/pixel - 16K stamps. Each motion object is specified by 4 words in the motion object RAM as follows: ADDRESS |D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ 7DAxx0 | X X X X X X | LINK | +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ 7DAxx2 |HF | PIC # | +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ 7DAxx4 | HPOS | x x x |PRI| PALETTE | +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ 7DAxx6 | VPOS | H SIZE | x | V SIZE | +---+---+---+---.---+---+---+---.---+---+---+---.---+---+---+---+ HF: Horizontal Flip H SIZE: Number of stamps wide - 1 (0=1 stamp,...,7=8 stamps) V SIZE: Number of stamps tall - 1 PALETTE: Color palette select (0 - 7) PRI: Motion object to playfield priority. 0 = MO wins over low priority PF, 1 = MO always wins PIC #: Picture #. LINK: The link register specifies the next higher priority motion object to be displayed. The M.O. that is linked (drawn) last has the highest priority. D[15:10] are spare and can be used for any purpose. VPOS: The vertical position (of the lower left corner) of the motion object relative to MOVSCROLL. HPOS: The horizontal position (of the lower left corner) of the motion object relative to MOHSCROLL. SLIP: (Starting LInk Pointer) The first object processed on every line is determined by a SLIP. The 64 slips correspond to the 64 rows of playfield stamps. SCROLL, MOB AND VAD CONFIGURATION D15 D14 D13 D12 D11 D10 D9 D8 D7 D6 D5 D4 D3 D2 D1 D0 +---+---+---+---.---+---+--+--.--+--+--+--.--+--+--+--+ | DATA |PARAM CODE | +---+---+---+---.---+---+--+--.--+--+--+--.--+--+--+--+ Motion objects scroll (universe) registers are completely seperate from playfield provided that the playfield scroll register addresses are higher than the motion object register addersses. The following address range must be used for the scroll registers and MOB configuration words : $7DBF00 - $7DBF7F. Unused words in this range MUST be 0. VAD_reg, Shadow Mnemonic Value 7CFFC0, 7DBF00 HDW_ENABLE 8000 A000 7CFFC2, 7DBF02 HDW_VSY_LD 0795 7CFFC4, 7DBF04 HDW_VBL_LD 0AEF 7CFFC6, 7DBF06 HDW_VINT 00EF 7CFFC8, 7DBF08 HDW_HSY_LD 5EEF 7CFFCA, 7DBF0A HDW_HBL_LD BEB2 7CFFCC, 7DBF0C HDW_SLIP_LD 6F05 7CFFCE, 7DBF0E HDW_APDMA_LD B301 B384 7CFFD0, 7DBF10 HDW_PMBASE_LD 9228 9229 7CFFD2, 7DBF12 HDW_ALBASE_LD 02FF 6FF 7CFFD4, 7DBF14 HDW_OPT_LD 0E50 4450 ??4e10/0e10?? 7CFFE0, 7DBF20 HDW_MOCON 5B95 5B85 ??5b95?? 7CFFE2, 7DBF22 HDW_PCON BC96 AC86 CC86 ??cc96?? 7CFFE4, 7DBF24 HDW_GRCON 3427 3027 7CFFE6, 7DBF26 MO_HSCROLL 0009 7CFFE8, 7DBF28 PF1_HSCROLL 000A 7CFFEA, 7DBF2A PF2_HSCROLL 000B 7CFFEC, 7DBF2C MO_VSCROLL 000D 7CFFEE, 7DBF2E PF1_VSCROLL 000E 7CFFF0, 7DBF30 PF2_VSCROLL 000F HDW_ENABLE = 0XC000 to disable VAD internal registers from being refreshed from the shadow locations in Alpha RAM until the next VINT_ACK. HDW_ENABLE = 0x8000 to enable the shadow registers. HDW_ENABLE = 0x0 to disable the registers in the range 7CFFC2-7CFFD4. The parameters will be read by the hardware shortly after the beginning of VBLANK D. Priority: - Motion objects at priority 1 are highest - Playfield pixels with bit 7 hi are next - Motion objects at priority 0 are next - Playfield pixels with bit 7 lo are lowest E. Color RAM: There are 4 banks of color RAM. Bank 0 is selected following reset. See LATCH for location of bank select bits. COLOR RAM ADDRESSING PLAYFIELD HI BYTE = 3C0000 + Picture bits * 4 PLAYFIELD LO BYTE = 3C0002 + Picture bits * 4 MOTION OBJECT HI BYTE = 3C0200 + Picture bits * 4 + Color palette * $40 MOTION OBJECT LO BYTE = 3C0202 + Picture bits * 4 + Color palette * $40 Color RAM data interpretation: D[15] LSB of R, G & B combined D[14:10] Red D[9:5] Green D[4:0] Blue In all cases, $0 = OFF $1F = highest intensity. NOTE: Graphics pixel data in playfield and MO roms is high true. (not inverted) MARBLE MAN VIDEO 68010 MEMORY-MAP ADDRESS R/W DATA ----------------------------------------------------------------------- Program ROM (secured) 000000~0FFFFF R D[15:0] (double-mapped, not secured) 100000~1FFFFF R D[15:0] EEPROM ++ 601001~612FFF R/W D[7:0] Watch Dog (117ms timeout) ++ 607000 W xx Video RAM (16K) 7C0000~7DFFFF R/W D[15:0] Color RAM Motion Object 7C0200~7C03FF R/W D[15:0] Color RAM Playfield 7C0400~7C05FF R/W D[15:0] Playfield picture RAM 7D8000~7D9FFF R/W D[15:0] Motion Object RAM 7DA000~7DBEFF R/W D[15:0] (Link, Picture, H-Pos, V-Pos, Link... etc.) Hardware config 7DBF00~7DBF7F R/W D[15:0] SLIP pointers 7DBF80~7DBFFF R/W M.O. link pointers Program RAM (16K) 7F8000~7FBFFF R/W D[15:0] (64K) 7F0000~7FFFFF R/W D[15:0] Note: All addresses can be accessed in byte or word mode except as noted. RAM at 7F0000 is multiple mapped at FF0000 to take advantage of 68000 "QUICK" addressing mode. Security: The security scheme is yet to be determined. It will involve bank switching the last ROM of program memory (0C0000~0FFFFF). This ROM will have 2 wait states at 14MHz (0 at 7MHz) ++ These addresses and those listed below are part of the Consistent Memory Map and as such should be defined as offsets from the base address of 600000 to allow portability. These addresses are part of the Consistent Memory Map and as such should be defined as offsets from the base address of 600000 to allow portability. CONSISTENT I/O MEMORY-MAP ADDRESS R/W DATA ----------------------------------------------------------------------- Left player input 600000.B R D[1:0] Center player input 600001.B R D[1:0] Right player input 600002.B R D[1:0] Spare input 600003.B R D[1:0] D1 : Button 1 Active lo D0 : Start Active lo NOTE: Center player removed for 2-player game When used as 2-player JAMMA with 3 buttons + start button per player: Player inputs 600000 R D[9:8],D[1:0] D9 : Left Button 1 Active lo D8 : Left Start Active lo D1 : Right Button 2 Active lo D0 : Left Button 2 Active lo Player inputs 600002 R D[9:8],D[1:0] D9 : Right Button 1 Active lo D8 : Right Start Active lo D1 : Right Button 3 Active lo D0 : Left Button 3 Active lo Status inputs 600010 R D[7:4] D7 : VBLANK Active lo D6 : Self-test Active lo D5 : Write to Audio full Active lo D4 : Read from Audio full Active lo Other inputs 600012 R D[7:4] D7 : Test Active lo D6 : Service/Test Active lo * D5 : Service/Test Active lo D4 : Service/Test Active lo * only D6 actually connected to Service switch input LETAs (LETA Mode = 0) 600020.B~600027.B R 600027.B : Left player trackball X 600026.B : not used 600025.B : Left player trackball Y 600024.B : not used 600023.B : Right player trackball X 600022.B : Center player trackball X 600021.B : Right player trackball Y 600020.B : Center player trackball Y LETA 1 (Mode = 1) 6C0020.B R D[3:0] D3 : Center player joystick down D2 : Center player joystick up D1 : Center player joystick right D0 : Center player joystick left LETA 0 (Mode = 1) 6C0021.B R D[7:0] D7 : Left player joystick down D6 : Left player joystick up D5 : Left player joystick right D4 : Left player joystick left D3 : Right player joystick down D2 : Right player joystick up D1 : Right player joystick right D0 : Right player joystick left Sound Processor Read 600030 R D[7:0] Sound Processor Write 600040 W D[7:0] LATCH (cleared by RESET) 600050 W D[10:4] D10 : Color RAM bank select - MSB D9 : Color RAM bank select - LSB D7 : Center player LETA mode: 0=sensors, 1=switches D6 : Right/left player LETA mode: 0=sensors, 1=switches D5 : LETA resolution, 0=lo D4 : Audio board reset (active lo) D1 : Left coin counter & VCR Power D0 : Right coin counter Unlock EEPROM 600060.B W xx BYTE WRITE ONLY!! Word write may not unlock Video IRQ Acknowledge 7CFFFC W 0 Writing other than 0 may cause strange problems to occur

I'm interested in a copy.

Correct.

Yes I can dump this for you no problem.

Super Cobra was programmed by Roklan Corporation for Parker Brothers and Popeye was started at Roklan but finished at On Time Software. Both of those protos were found in the archive of a retired programmer who at one time worked for both companies.

Roklan did not work on the hulk game for the 2600.

Mainly because 5200 Tempest cannot be played with digital controls. It also uses all 32k of ROM. Since it was originally meant as a 5200 exclusive we have no plans to make an XL/XE version. There is Tempest Xtreem and now Tempest Elite for XL/XE.

I’ll give it a try.

Does anyone have a WHDLoad for Mean 18? Id really like to install this on my hard drive if possible.

Correct. It came from the Roklan/On Time Software archive.

Yes, they had a few VAX 11/780’s as well as some 11/750’s.

Most likely was, this was done after the sale of Atari to Jack Tramiel so it looks like this is a list of what was on hand at that time.

Jack never programmed anything for the 2600. My bet would be he worked on the Atari 8-bit version.

I suspected he converted Joust for C64 directly from the Atari 8-bit version based on the source and data I discovered.

Your welcome.