Atari Centipede’s Hidden Code Trap | Hacker News Reader

Atari Centipede’s Hidden Code Trap

arcadeblogger.com

120 pointsjasoneckert43 comments

>Simply stating that "this game copied our idea!" is all well and good; but in a court of law, the case has to be proven beyond any reasonable doubt.

In common law systems such as the US and Canada, "beyond a reasonable doubt" is generally the metric used for criminal law. Civil cases, such as this, generally use preponderance of evidence; it's a much weaker standard, essentially meaning that whichever side is more than 50% supported by the evidence will win.

Doesn't take away from the interest in this code trap, but Atari did have a lower bar for success.

Wouldn't this have been criminal copyright infringement, since it was for commercial gain and impacting the commercial prospects of Atari?

IANAL (or an expert on Canadian IP law) but while it perhaps could have been as well, it's pretty clear from TFA that this was a civil lawsuit. Use of the term "lawsuit" and "plaintiff" (the latter in the only actual document provided[1]) are dead giveaways, as is the fact that the government (who prosecutes criminal cases) does not appear to be involved.

1: https://missilecommand.files.wordpress.com/2018/12/ilovepdf_...

The article tells us the non-Atari game retained a damning Morse message in its ROM. From another source[1] (unverified), this may be the ROM data the article refers to:

    .BYTE 2, 0BB, 5A, 30 (where 0 is dot and 1 is dash and nothing for a space)
    .BYTE 5F, 0EE, 7D, 0A8 ; Morse code for "COPYRIGHT 1980 ATARI"

[1] https://www.binarni.net/2013/07/centipede/

With all the ROM images floating around in the MAME/retro scene, someone can probably verify this.

Okay, replying to my own curiosity... The above byte sequence can be found in a dump of Centipede's ROM.

    00001280  a2 13 a9 ab 5d 20 21 ca  10 fa 85 fe 60 02 bb 5a  |....] !.....`..Z|
    00001290  30 5f ee 7d a8 20 b3 21  85 ae b9 c0 21 85 b0 a9  |0_.}. .!....!...|

The entire arcade game is contained in 12 kilobytes of code, BTW. (!) The amount of imagination and fun and replayability that this era of gaming delivered under such constraints (by modern luxurious standards) is mind boggling.

Could you expand on this a little for the uninitiated amongst us?

Today, to keep people engaged in a game you need to have 50-100GB of graphics, 1GB of sound, some sort of AI, half naked cartoon ladies on the cover, 100 different weapons/races/etc.

But if you think of Centipede, Donkey Kong, PacMan, etc, their simplicity is that make them so great and timeless.

I remember I had BubbleBobble on the PC (back in the Amstrad PC 1512 days) and I would start and finish that game between 45-55 minutes no-matter-what.

Bubble Bobble's PC version was a really bad conversion though. A lot of powerups appeared under completely different conditions, white ghosts didn't shoot, even the mechanics of some levels (72 I think) were different.

I don't know how I feel about the condescending tone of this comment, especially the "half naked cartoon ladies on the cover" dig.

It's a shame your refined senses can't appreciate the fidelity to game designer's artistic visions modern technology has brought to video games, the rest of us will enjoy them in your stead.

The sexism dig isn’t about refinement. It’s something that’s rife in all popular media from music videos through to movies and computer games. However that has always been the case in gaming. There’s nothing new there. Even 8bit games for the C64/Amstrad/etc would often feature highly embellished cover art.

Things are getting better though. Before it was just the accepted status quo but at least now those clichés are being challenged.

> The sexism dig isn’t about refinement

Didn't imply it was?

Not so with Nintendo. I think I read somewhere that "if the game isn't fun with circles and squares, all the graphics in the world won't make it fun." Or something to that effect.

Exactly the kind of reason that a game like factorio can be wildly popular, it's not pretty by any means, but it does what it does very well

Spent a little time decoding this...

  02       BB       5A       30       5F       EE       7D       A8
  00000010 10111011 01011010 00110000 01011111 11101110 01111101 10101000
        C. ..O..P.. .Y...R.. I.G..H.. .T1....9 ....8... .0....A. TA.R..I.

*edit: formatting

Did you use the source code for this?

Nah, I just did a little manual pattern-matching on the bits looking for "ATARI" (morse code : "0110101000") and working backwards from there.

12 kilobytes isn't much of a constraint. Before 1979, few computers had that much room. Check out this Mario Brothers game written in only 10 lines of Atari BASIC on an Atari 8-bit computer: https://www.youtube.com/watch?v=CmpBZrfk1w8

That's super impressive!

The Mini Bros game is 2.5 kilobytes of high level source[1] so a fair resource comparison should either include FastBasic 3.4's runtime footprint (if interpreted?) and/or a compiled standalone executable size (if FastBasic is capable of it). i.e. ROM + RAM requirements to run on bare metal, including libary/runtime dependencies.

But my byte-counting pedantry aside ;-), you're exactly right to celebrate authors of such tight works that are still actually fun and attractive. I enjoyed the author's line-by-line walkthrough and dev journal. I know some modern programmers may not see the point in these apparently arbitrary competitions targeting old platforms but I really think there is wisdom and magic in this frugality!

[1] https://www.dropbox.com/sh/zj7u96etduyq6bv/AABfTzwMu3pqoNSNA...

A similar technique was used to prove that the mobile game Heroes Charge was built on stolen code https://www.facebook.com/allstarheroes.en/videos/15770877858... Except that code has become a lot more complicated since the Atari days. So, you don't need a hex dump to show the evidence. Just a gesture sequence on a certain screen and the UI pops up "Copyright: The Original Company".

Meanwhile, the original company was simultaneously being sued by Blizzard for blatantly knocking off the character designs of World of Warcraft https://blizzardwatch.com/2015/03/24/blizzard-files-lawsuit-...

”Steve decided that if a company copied the Mac ROM into their computer, he would like to be able to do a demo during the trial, where he could type a few keystokes into an unmodified infringing machine, and have a large "Stolen From Apple" icon appear on its screen.”

Ha! This makes Rap Genius's apostrophe trap even more delightful, then.

https://www.rollingstone.com/music/music-news/genius-google-...

So how was the code trap implemented?

Sounds like a coded message in the rom dump

That's where but not how.

"This trap consists of a particular data pattern. Only with the help of special code is it possible to read this pattern."

This is the key part of the story & it's totally unexplained. The blogger "Tony" just dropped it in there with no vetting at all, as if the method was self evident. Such is news in the new millennia.

The claim could just be bullcrap. Atari would have known that courts are tech-illiterate and the defendant might not have shown up for the hearings.

Well to be fair, all there is to go on at this point is that document. The blog is pitched at a wide audience and is meant to be a "light" entertaining read, rather than something in depth and technical. But agreed, it would be interesting to know the exact process of "extracting" the coded message. The point of the article is that it was there (or to your point, it was claimed in court that it was there!), rather than how it got there. Bottom line - it did the trick.

Ed Logg is alive today according to his Wikipedia entry, so there's more to go on if someone wants to do write something more detailed about this incident.

Sounded like steganography... Encoding Morse code into unused bits in the ROM

A little bit of googling shows the ROM is available. Might be a good weekend project to bust out your disassembler and find out...

Yeah, I don’t get it. You could just compare the ROMs directly. I can XOR any arbitrary secret message with a long enough key.

From the article, it wasn't a XOR, the "key" was that there was a bit sequence when interpreted as Morse code said something like "copyright ATARI". So, reasonably easy to show it wasn't just pulling the message out of thin air.

The ROMs wouldn't an exact match so you have to convince the layman jury that the differences are meaningful. The best way it to show something totally unexpected in the pirated code like your own copyright.

I wrote a few game cartridges for the Atari 400/800 computers when I was there in the early 80s. We were told to include a specific bit pattern, I think it was 5 or 8 bytes long, somewhere in our code. The idea was that the presence of a non-functional, unique bit pattern was going to be very difficult for a defendant to explain in court, since pirates mostly just copied ROMs and made minimal changes to obvious bitmaps and strings. Even if they found the mysterious bit pattern and were clever enough to (say) zero it out, they'd have to explain why most of a ROM was identical except for visible assets and that one set of bytes with no function ("What are these bytes for? Oh, you don't know? Let us show you...")

We weren't told what the pattern meant. I played with it for a couple of hours once; I knew it wasn't executable 6502 code or simple logical bit operation (NOT, reverse bits, etc.) of anything obvious. Morse code makes sense now.

Some coinop machines (Tempest, for sure) used character encodings so that the visible copyright messages were also executable code; changing the message, even slightly, would crash or have unpredictable behavior. In the case of Tempest, when the "final" ROMs were released to manufacturing, the dev team went on vacation while production ramped up. Some lawyer decided that the copyright message had to change, so they patched the ROM on the factory floor. Unfortunately this modification didn't lead to an immediate crash, but opened up a hole where the contents of the score line could do things like give you 100 extra lives (e.g., die with the low digits of your score being "50" and you'd be able to play for free for hours).

Side note here, the Atari 800 was my first computer, and I absolutely loved every minute of it. Thanks for all your work and ingenuity, they gave me a good start on a career in technology.

One question: why didn’t the Atari 800 ever take advantage of the double cartridge ports?

My understanding was that the second cartridge slot was not a very well thought out thing. I think there was one cartridge made for it (some debugging tool?)

That port cost a lot of money per unit; Atari's early efforts were not very well cost reduced. If they'd been more aggressive about cost reductions (and improving I/O speeds) they might have done much better in the market against Commodore. Instead we got more unclear hardware directions (the 1200XL's "Help" key was a prime example of the hardware group just not caring about or talking to the software folks), even more expensive machines, and a bunch of work being done by outside contractors instead of inside staff because it was easier to build empires that way. The 7200 game machine (developed by General Computer Corporation in Cambridge, IIRC) was a surprise to most of us, for instance, and not a good one.

There were a few third-party cartridges that extended Atari Basic and used the second cartridge port, the most famous being Monkey Wrench by Eastern House. (http://www.atarimania.com/utility-atari-400-800-xl-xe-monkey...).

See this post at AtariAge about the motivation (https://atariage.com/forums/topic/71760-atari-400800-left-ca...): "According to Joe Decuir, the original idea Atari had for the 800 was that software would come with the base program (LEFT cart) and the add-on content (RIGHT cart). This way you could keep expanding an educational program with new lessons, for example. Plus, having the software in ROM would reduce the RAM requirements of the computer since nothing needed to be loaded. RAM quickly came down in price and many of the advantages of such a complex cartridge arrangement went away ."

The Nintendo gameboy has this, to block unlicensed games. If the copyright message is not displayed by the cartridge at boot, the machine will refuse to play the game.

This is interesting in a couple of ways. First, I think it was ruled in court that a developer could use the Nintendo logo in their game for an unlicensed game because it was necessary to make it work. Nintendo's legal trap backfired.

In another case, someone made a knockoff system from the GameBoy which would play the same games. The games they released would show their logo, I think it was Game Fighter, but in the code it actually contained the Nintendo logo. Basically tricking the system into running a game with a different startup logo.

> Unfortunately this modification didn't lead to an immediate crash, but opened up a hole where the contents of the score line could do things like give you 100 extra lives (e.g., die with the low digits of your score being "50" and you'd be able to play for free for hours

Was this intentional is order to make the pirated game not make money at the end customer?

It would have been really rad to tie this embedded message to an input sequence. I guess in centipede that’s extra hard, but it would make for a great moment in court when the devs roll in defendants machine and moments later copyright Atari shows up..

Jarvis and DeMar did exactly this on Robotron:2084, in the hopes it would be useful in court one day:

https://youtu.be/3zr4CdUdt0w

This reminds of the story why Van Halen had a M&M clause in their contract for rock concerts.

It is a trap to ensure that the contract was fully read and complied with.

https://www.snopes.com/fact-check/brown-out/

>Simply stating that "this game copied our idea!" is all well and good; but in a court of law, the case has to be proven beyond any reasonable doubt.

This sounds like patents or copyrights can grant monopoly on an idea. As far as I know, this is not the case. (Well, at least not in principle.)

Xor ftw!

Reply on news.ycombinator.com