Kr0tki

Considering that Super Breakout is a 4KB cartridge, it is more appropriate to say "It appears to be Super Breakout" 🙂 But seriously, this ROM is the final version of Super Breakout.

I've noticed it in the 8-bit final version too, that the scroll is glitchy on PAL. I guess the non-scrolling in 6-14 was not caused by intentional TV system detection, but a result of some timing bug that Mr. Dyer managed to fix (though not completely) later on. There seem to be no PAL adjustments in the game. No colour adjustments, no screen geometry fixes, everything plays 20% slower.

Actually, in 6-14-84 the bonus items appear after eating a specified amount of fruits and power pellets, same as the final version. But I guess you have figured this out already. I see that the same is stated on your review page for 3-15-84, but again, that's not what I see when playing the 1st level. Only 1-26-84 exhibits the timer bug. In 3-15-84 the bonus items appear as in the final version. Another notable difference between 1-26-84 and the later versions is ghost colours - in 1-26-84 one of the ghosts is violet but it was changed to pink in later versions. Your "Super Pac-Man (800 Repro).bin" image exhibits the violet ghost colour, the bonus timer bug, and the lack of scrolling instructions. So it is closer to the 1-26-84 5200 version, not 3-15-84. I reviewed all the EXE and disk versions of 8-bit Super Pac-Man available at AtariOnline.pl. All of them are almost 100% identical to your "Super Pac-Man (800 Repro).bin" image (with a few bytes modified here and there to, remove copy protection and allow launching the game from disk). This means that all pirate copies we had back in the day, and your repro ROM too, are based on the same version of Super Pac-Man. I don't think this version was converted from the 5200 by pirates: other 5200 conversions from back in the day did not have "#" and "*" on their title screens replaced with "Option" and "Select", and even if the prates did so, what would be the chance that they would put "Option" and "Select" on the screen at exactly the same location as in the final version? So my assumption is that an 8-bit version of Super Pac-Man was leaked from Atari which was similar to the 1-26-84 5200 proto. Dutchman2000 said that he also disabled the copy protection when compiling. And surely enough, I've tested that the final 8-bit version is the only one that does not try to modify the ROM area during gameplay. As noted above, the other difference is enabled copy protection. But there is yet another difference between 6-14-84 and the final version. When running 6-14-84 on PAL, the bottom row on the title screen remains stationary - it only moves when run in NTSC. In the final ROM it scrolls in both TV systems. This means that 6-14-84 is not simply the final version with disabled Easter egg, so there might be other differences between 6-14-84 and the final ROM. There is yet another version of 8-bit Super Pac-Man that is omitted from your analysis: the Best Electronics release, dumped by FULS and provided here. This version has the pink ghost, the correct bonus timer, and the scrolling instructions that work both in NTSC and PAL - all the same as the final version; but it also has enabled copy protection and disabled Easter egg. It would be way easier to determine all the differences between all these versions if the source code was available. @Dutchman2000 Is it possible that you would release the sources?

Maybe you are using too old conversion utlility? Support for "Standard 4KB cartridge" type was added to emulators only 11 years ago.

No - a cartridge in the right slot may show up in the address space only at $8000-$9FFF.

The ROMs for these cartridges were published on the Atariware.cl forums a few year ago, but the forums have since disappeared, unfortunately. Here are the ROMs I downloaded back then. They are indeed identical to this CM Prat Editor, with the exception of hacked titile screens. I guess MultiEditor from Multisoft was the earliest release. MultiEditor (1985)(MultiSoft)(CL).bin AtariWriter (1991)(Prismasoft)(CL).bin

I was also wondering about that "00 80 AC B8" cartridge init block being present in both ROM halves. After all, the one at $9FFA-$9FFF is not used in this cartridge, so what was the point? (It's worth noting that each of Thorn EMI's 16 KB cartridges also contains an additional cartridge init block at $9FFA-$9FFF, for no apparent reason.)

Your dumps of E00.02 and R01.00 are overdumps. E00.02 contains 2 copies of the same 2 KB data and R01.00 contains garbage in the first 2 KB. I've stripped off the unnecessary data. The resulting 4 KB ROM images differ only in the first 16 bytes which contain ASCII signatures - they are "E00.02-101579-02" and "R01.00-101579-02", respectively. I guess "E" means some kind of preliminary version and "R" means final release - maybe they decided at some point in time that version E0.02 is good enough to consider it final and renamed it R1.00? Later SALT 2.04 and 2.05 cartridges also contain signatures starting with "R02.0x" (curiously, 2.06 and later SALTs don't contain signatures anymore). I have also extracted a ROM image from your E00.01 file. It was easy because the binary file just loads the ROM contents into RAM at $8000-$9FFF and performs some required initialization to move the display list out of the ROM area. This 16 KB ROM is interesting as it contains 4 KB of unique data - the first 8 KB is one 2 KB block mirrored four times, and the second 8 KB is another 2 KB block mirrored four times. I suppose this cartridge originally consisted of two 2 KB ROM chips mounted on a standard 16 KB board - that would explain the mirroring of the ROM contents. But I am not sure if any 16 KB ROM boards even existed in 1979 (the first released 16 KB cartridge was "Wizard of Wor" in 1981). Maybe the ROMs were placed on two separate cartridge boards which were inserted into both Left and Right cartridge slots? That would also result in the mirroring as observed in the binary file. Anyway, the ROM image contains a signature "E00.01-092379-02" in each of the two 2KB ROMs - at $9800 and $B804. Currently emulators do not support such kind of 4 KB ROMs, so I am attaching the E00.01 image in its 16 KB form. 400-800 SALT E00.01 (1979-09-23)(Atari)(US).bin 400-800 SALT E00.02 (1979-10-15)(Atari)(US).bin 400-800 SALT R01.00 (1979-10-15)(Atari)(US).bin

Some of your ROM images are overdumps. I've stripped the ROMs from unnecessary data and I'm attaching them below. Stargate 1-13-84 is a two-chip 16KB cartridge. Millipede 1-4-84 is a one-chip 16KB cartridge. Choplifter 1-4-84 is a two-chip 16KB cartridge. Choplifter 1-13-84 is a one-chip 16KB cartridge. Some differences I've noticed: Comparing Choplifter 1-4-84 and Choplifter 1-13-84, most of the binary differences amount to switching from two-chip to one-chip cartridge during development. Basically, all ROM accesses to the $6000-$9FFF area needed to be changed to point to $8000-$BFFF instead. Other than that, there are only two other differences between the two: When pressing Reset, 1-4-84 goes to the BIOS splash screen, while 1-13-84 goes to the title screen like the final version. The CHOPLIFTER! title on the splash screen is off-center in 1-4-84, in 1-13-84 it is centered like in the final version. (Update your screenshot on the 1-13-84 page, please!) And, as you noticed, there is leftover source code garbage in 1-13-84. Comparing the "unknown" Choplifter prototype and the final version, there are two differences I have noticed: In the proto there is a little glitch on the top white bar above the points display - part of its left side is miscoloured. (Again, please fix your screenshots.) The demo mode in the "unknown" contains a custom attract mode that cycles background colours (wait a minute or so), presumably for screen-saving. The final version does not implement any custom attract mode and just uses the standard one as implemented in the BIOS, which kicks in after about 9 minutes. Note that the "unknown" prototype does not disable the standard attract mode, so after 9 minutes it kicks in and starts colour-cycling the top frame of the screen independently. The prototype (and the two earlier prototypes too) does not work with the two-port 5200 BIOS. They fixed compatibility in the final version. On Stargate 1-13-84, the game's title shows up for me on the splash screen, both in Altirra and in Atari800, same as the 1-5-84 prototype. Are you sure there wasn't some kind of a mixup? Choplifter! (1984-01-04)(Atari)(proto).bin Choplifter! (1984-01-13)(Atari)(proto).bin Millipede (1984-01-04)(Atari)(proto).bin Stargate (1984-01-13)(Atari)(proto).bin

Not at all. The files on MrFish's webpage are a mess. "Translator (NTSC) Side A (OS-B)" and "Translator (NTSC) Side B (OS-A)" are verbatim copies of the original "Atari Translator DX5063" disk as posted by bandit in this post above, but they are mislabelled. As stated numerous times in this thread, that disk contains OS-B on both sides, the difference being side A has the reset-proof version and side B the non-reset-proof one. "Translator (PAL) Side A (OS-A)" is indeed a dump of the original "PAL Translator rev # 0" disk as posted by luckybuck in this post, but "Translator (PAL) Side B (OS-B)" is not side B of that disk, but an attempt by 1050 at creating a translator disk with PAL OS B, originally posted here. @MrFish please fix your webpage. @luckybuck Have you ever managed to dump Side B of your "PAL Translator rev # 0 serial # 4" diskette, or verified that it is empty?

You're misunderstanding the issue. Only the sources of OS A, OS B, XL OS rev. 2 and CC 01r04 come from Atari. The sources for all the other revisions were my recreations based on disassemblies. The comments that you have metioned were in fact written by myself. Atari knew nothing about the bug.

Yes they were, more or less.

The "Popeye" cartridge is compatible with all computer models.

I have prepared a package of sources of all OS versions in the format accepted by the ca65 cross-assembler. It's just a directive to generate a sequence of bytes, similar to MAC/65's ".BYTE". They have used it as a way of forcing absolute addressing. "VFD 8\$8E,8\low APPMHI,8\high APPMHI" generates $8E, $0E, $00, which equals to "STX APPMHI" but using the absolute addressing instead of zero-page.

Wikipedia lists some examples older than "Bruce Lee".

JKK, could you adhere to the practice of admitting upfront that you are the author of the video you are promoting? I think we all deserve that kind of honesty. Also, how is this new video different from the one you were promoting last year? I don't think anyone would like to view the same video again if it is not at all different.

One reason Atari chose to design the controllers to require disabling debounce, was to enable reading of simultaneous pressing of multiple keys. A must-read for this topic is the "PAM Package" document, available at the old Atari Museum website. It discusses several approaches to reading the keypad.

The 16-byte CAR header contains a checksum of the following ROM data. You haven't mentioned that you were updating the checksum, so maybe this causes the modified CAR file to stop working? I'd suggest to convert the CAR image to raw binary first, via File -> Save Firmware -> Save Cartridge..., then hex-edit it, then convert it back to CAR. This way the checksum would be recomputed correctly.

I wonder if anyone noticed through all these years (I certainly haven't), but a nice high-quality photo of this prototype 130XE case is proudly presented on the front of each 130XE box and on the title page of the owner's manual. ^_^

You're hitting the right note, if only coincidentally: the preliminary specifications of Self Test contained a mock-up screen of the Audio-Visual Test on which the notes were indeed showing the theme from Close Encounters! Someone at Atari apparently intended to use that theme in Self Test before the legal team dropped in 🙂 I'm attaching all of the Sweet-16 project specs since AtariMuseum.com is offline; see sweet-16_OS-supplement-1_Self-Test.pdf. 1200XL.zip

Nah, it's probably just the date the message was written on. Or maybe I was wrong and the message is aimed at hackers - in that case an insult would be nothing out of ordinary. Yup, "RoboCop" is a well-known joke program, many Polish users have been recalling their stories of disappointment when they were buying that "game" from pirates only to find out at home what it really was. The translation: "Ha ha ha! You've been thinking it's a normal game, but a "certain" someone has played a joke on you. It's sad, but for now there's no profit in doing anything in Poland. Maybe sometime, in the future, I will create "RoboCop". So for now you have to settle for this awesome game. It's really great, because I was developing it for a lot of time, four days (2-3 hours daily with breaks for meals). It's great. And now greetings to the best computer company ........ with its head developer Boris Peacock. Have a nice fun! Bye! 26 Oct. 1991"

Alright ... "Hello you pimpley worm! Why are you sticking your stinky hands here? You wanted to steal some great Italo Disco & Marina-style music. I see you've already finished the "Bat Man" game. If so, then send the solution and a fully drawing of the map. As a reward we will send you another, world's best game, "Robocop" - 21 Dec. 1991“ Maybe it's supposed to be displayed as an ending screen.

You've got your answer in post #3.

No, all cartridge releases are the same 8K version without intermissions. The one with intermissions was released by Datasoft (and later by other publishers) only on disk and tape.

Some of that translation is incorrect, because the words were translated separately instead of as whole sentences. It should be: Выставка Компьютерной Техники -> Exhibition of Computer Hardware вход по пригласительным билетам -> entrance by invitation cards PZ Karen was a Polish company that acted as official Atari service center. The poster contains a logo for "Quasar", a brand of PC/XT compatible computers assembled in Poland by Karen.