15

The two most important chips in a game console are the CPU and GPU. In the Atari 2600, the CPU was a cut down version of the 6502, a very common low-cost microprocessor. The GPU was the TIA, which was manufactured for Atari by several chip companies, as detailed in the answer to Who made the Atari 2600 TIA video chip?

When competitor Coleco introduced the Colecovision, they also introduced an 'adapter' to make it compatible with 2600 games, which effectively consisted of a 2600 console in an adapter form factor. Presumably for this, they purchased TIA chips from one of the manufacturers thereof.

I am surprised that Atari didn't get to object to this on the grounds that they owned the intellectual property rights to the TIA. Why not? If they didn't own those rights, how did that come about?

22

The TIA was designed by Jay Miner while he was working at Atari. I'd be very surprised if Atari did not hold whatever rights could be held.

That being said, the rights that could be held to an IC design at the time were much more limited than they are today. Before the Washington treaty of 1989 and TRIPS, ICs were protected only by patents, not by copyright, and patents by their nature were limited to new functionality, and could not be used to protect an IC as a whole.

That's how Nintendo/Ricoh were able to use a slightly modified 6502 as the CPU part of the 2A03 without licensing it from Commodore: The patent for the 6502 covered only its BCD implementation, the only new functionality on the chip. By disabling the BCD part of the chip, it was possible to legally use an otherwise exact copy of the 6502 core without infringing Commodore's rights.

The same lack of protection is what, in turn, made the large number of NES clones possible.

The Wikipedia article on Miner notes that he (together with others) held a patent for the later CTIA and ANTIC chips, but doesn't mention one for the TIA; and a quick search for his name at the EPO only gives a patent for a "data processing system with programmable graphics generator" (i.e., the 2600 as a whole, not the TIA specifically) in the relevant time frame. Combining this with Stephen Kitt's answer to the question linked to in your question, where he notes that "not all of these chips were manufactured under contract with Atari", I'd assume that the legal protection available at the time was simply not strong enough for Atari to do anything about the cloned TIAs.

1
  • The answer refers to "cloned TIAs", while the question text implies it was actual TIAs from the TIA manufacturers being used. Which was it?
    – T.E.D.
    Nov 6 '20 at 15:49
2

The wiki explains it in detail ...

The expansion module prompted legal action from Atari. Coleco and Atari settled out of court with Coleco becoming licensed under Atari's patents. The royalty based license also applied to Coleco's Gemini game system, a stand-alone clone of the 2600

1
  • 3
    Did Coleco's clones behave identically to the TIA in undocumented edge cases? I found find it very surprising if anyone who hadn't studied the circuit design of the TIA could produce a chip which behaved identically in all cases involving things like stores to RESP0 which are separated by 5 or 6 CPU cycles, or the effects of changing NUSIZx at different times after a sprite triggers. I remember in 1994 I noticed that repeated RESP0 behaved oddly, but it wasn't until I examined the schematics that I would have been able to characterize what was happening.
    – supercat
    Nov 7 '20 at 21:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.