That's correct - OpenRW is a from-scratch reimplementation of GTA3, while re3 is a manual decompilation that's capable of reproducing the original executable (or something very close to it). Both projects are very impressive in their own right.
When re3 went public, OpenRW archived their repository; when re3 was DMCA'd, OpenRW unarchived their repository; and when re3 was sued, OpenRW ceased development.
(This is based on my recollection from observing this two years ago - there may be other factors involved)
Bethesda has, in fact, taken issue with OpenMW in the past [0].
Bethesda at least had a specific reason for going after OpenMW, though. They’re okay with it now under “certain conditions” that I’m sure the OpenMW devs aren’t going to argue with.
Their main ask was that the OpenMW devs don’t advertise OpenMW as a way to play Morrowind on Android (or other platforms besides PC).
And there’s an understanding that OpenMW’s “intention” isn’t solely to be a way to play Morrowind, but also to implement a generic RPG engine and editor that uses Morrowind’s gameplay and file structure. For someone to make their own game with, not using any of Morrowind’s assets. (I’m not sure if any such game exists, though)
"And there’s an understanding that OpenMW’s “intention” isn’t solely to be a way to play Morrowind, but also to implement a generic RPG engine and editor that uses Morrowind’s gameplay and file structure. For someone to make their own game with, not using any of Morrowind’s assets. (I’m not sure if any such game exists, though)"
To be honest, every project of this kind claims that. And in reality, everyone knows 99% of usecases for such engine is to play the original game. Sometimes there are new IPs being created on such engine. For example recently there's been some "boomer shooters" coming out using engines which were made originally to play classic DOOM or Duke Nukem 3D. But such cases are rare.
Clean rooms protect from patent issues, but do little to protect against copyright or trademark actions. The fact that one creates totally new code from scratch means nothing if the end product is too similar to the existing copyrighted work.
> Clean-room design is useful as a defense against copyright infringement because it relies on independent creation. However, because independent invention is not a defense against patents, clean-room designs typically cannot be used to circumvent patent restrictions.
In code, not end product. You can reverse engineer code all day, but your end product better not end up looking identical to the protected work. I could set a thousand AI machines in a clean room generating new song lyrics. Eventually one of them will randomly come up with identical lyrics to a Taylor Swift song. That doesn't mean I now own those lyrics free and clear. "Technically, I didn't copy" isn't a defense when you try to sell an identical product. It can be a defense in patent areas where one wants to argue prior art or obviousness.
You can't copyright the functionality of a software program. Copyright is a fixed expression. You can't copyright a song about heartbreak and you can't copyright the idea of a four function calculator implemented in a GUI.
You may be able to patent the functionality of a piece of software.
The point of the clean room practice is to avoid a literal line for line copy of the original software. It would be entirely legal to say, write a song from the perspective of Taylor Swift about her feeling sad and betrayed after a breakup.
But analogies are dicey because the law treats functional software differently from literary expression.
Precedent here is vTech reimplementing the Apple II BIOS and Phoenix reimplementing the IBM BIOS and Connectix selling a PlayStation emulator for a fraction of the price of a PlayStation.
Except if you are going for obviousness you want to start with people who haven't seen the patented material. You put them in the metaphorical clean room and then argue the patent invalid because some rando people came up with the same solution to the problem. And for prior art you need a group of "clean" people who were using the process/device prior to the patent, which is especially useful when dealing with trade secret material that, fearing a leak, a competitor is trying to patent. You need people clean of any taint from that competitor.
Atari Games Corp. v. Nintendo of America Inc. 975 F.2d 832 (Fed. Cir. 1992)
"Nintendo can show copying by proving that Atari made literal copies of the NES program. Alternatively, Nintendo can show copying by proving that Atari had access to the NES program and that Atari's work — the Rabbit program — is substantially similar to Nintendo's work in ideas and the expression of those ideas."
No, clean rooms are a mechanism that allows a team to avoid copyright infringement by ensuring that the developers have never seen the copyrighted code.
The code avoids copyright, if it is actually different. The end product doesn't if it is still similar to the copyrighted work. Use a computer to rewrite Harry Potter all you want, but if it comes up with the same book it doesn't matter how you got there.
That's not a clean room implementation. A clean room would be having a description for the game of quidditch (game rules are not copyrightable just like specifications for software) and then writing a background story to explain it without ever having read a single word of Harry Potter. If that comes out anywhere similar to the Harry Potter books I'd be utterly amazed.
>> If that comes out anywhere similar to the Harry Potter books I'd be utterly amazed.
Considering that "Harry Potter" is not an uncommon name in the UK (I've met two) nor is the concept of a wizard school, or riding broomsticks, or house teams, I suspect that AI today might be able to generate something that would get Rowling's attention.
Rowling has been sued multiple times for the similarity of Harry Potter to previous books using those themes. She won every time. The bar for copyright infringement is not "a single piece of the book is similar" or even "a one sentence summary of the two is similar."
"Clean room" reverse engineering basically has you with two completely separate groups of developers. One group will decompile and analyze the target software, and build a detailed specification of how it works. The second group, with legal, sworn documents that they have never seen even the machine code of the target software - uses those specifications to build the "clone".
To be fair to Bethesda, they provide all the tooling you would need to make your own Elder Scrolls game with Morrowind, Oblivion, Skyrim, et al..
Modding a Bethesda game is literally getting into the innards of the game itself and changing or replacing bits in it as you please, at least until we get into hooking in DLLs which is a whole different story.
Of course publishers are going to keep these old IPs under lock and key, because they've now figured out they can slap new graphics on it every few years and re-sell it to us at full price.
Not really. No legal protection is able to stop an overzealous litigant from suing you. The real test is whether they win the case, not whether they can file one. Unfortunately, most people don't have the means to defend such a lawsuit, so they get bullied out of their legal rights.
These companies know very well these independent groups, even if it's on their legal bounds, can't afford a legal defense team and actively exploit that to abuse them away.
re3 is a very dubious legal case. They decompiled the game with debug symbols and then did a (aiui) non-clean room reimplementation of the code, based on the decompilation.
It is protected legal activity to decompile a binary and study the source.
It is protected legal activity to study some source code, describe it, and have someone who never saw the source reimplement the code described.
It is straightforwardly illegal to decompile source code, recompile it, and the distribute the recompiled binary.
re3 is somewhere between the latter two, and (again aiui from looking into it 2 years ago) they didn't have a clean room step.
It's probably a good thing no one will litigate it. It seems highly likely to be illegal and reasonably do. It would be a bad thing to spend money on.
> It is straightforwardly illegal to decompile source code, recompile it, and the distribute the recompiled binary.
But it is not straightforwardly illegal (according to Microsoft et al) to decompile source code, train an LLM on it, generate the source code, recompile it, and then distribute the recompiled binary.
One easy 947 quadrillion tensor-operations lifehack
"It is straightforwardly illegal to decompile source code, recompile it, and the distribute the recompiled binary."
According to what? If the rules are so poorly defined that way then maybe it's not a bad thing to disregard them.
Because simply decompiling source code and then recompiling it isn't a simple one-step thing in the way you suggest, not even close.
It's the distribution step that's illegal, since you don't own the original binary. If you can't distribute the original binary, why can you distribute a modified version of it?
If you wrote a bash script which decompiled and recompiled the original binary, that would be your property to distribute as you like
You are right, however it's likely that OpenRW was also impacted by Take-Two's actions towards Re3 given that its development also ceased around the same time.
