3

I was found in a situation where some author published a binary which was widely used. Unfortunately he never shared the source despite marking the program as GPL (version not specified). Despite multiple people asking he never respond to the questions about publishing the code.

Now the author vanished and people try to modify his work by binary patching. In my understanding the binary became the preferred way of editing here. For the context I should add here that the code is a Linux kernel module and is not a derivative of any existing code to my knowledge.

Is it possible to disassemble the binary, start adding symbols back and redistribute the original binary + disassembled C source?

3
  • 4
    I think the main concern is that the author may reemerge and sue people for not meeting the requirements of the GPL for his binary. However, in view of the fact that the author made that unconditionally impossible, a court might or might not allow such a suit to succeed: are the permissions moot because the requirements were never satisfied, or are the requirements void because it was never possible to satisfy them? It's probably a great jurisdiction-specific question for Law.SE, if you want a really meaningful answer, but this probably has no existing case law.
    – apsillers
    Jun 23 at 23:13
  • If the author worked with or advised others with regard to his binary, in full knowledge that they relied on his GPL permissions to modify/redistribute it, then possibly an estoppel or laches defense applies as well, since he permitted it and/or failed to bring a suit in any kind of reasonable time, despite obviously knowing about the downstream reuse. (But I am not a lawyer.)
    – apsillers
    Jun 23 at 23:16
  • "despite marking the program as GPL (version not specified)" - To license a program under GPL, the author should, at minimum, include a license statement and preferably a copy of the license itself. See How to use GNU licenses for your own software. "Marking" a program as 'gpl' (for example) in some kind of repository or package management tagging system is not enough.
    – Brandin
    Jul 20 at 9:39
2

Copyright law is interesting because it's at the boundary of criminal and civil law. But in this case, the criminal side can be safely ignored. Criminal prosecution has a higher standard of evidence, and this case is murky enough that there is insufficient public interest for the State to press charges.

This leaves us with the civil side, as apsillers already noted. The original author can sue for two main reasons: to stop an infringement, and to claim damages.

The first demand (stop infringement) is something you don't need to worry about, because that would just put you back in the default situation.

The damages claim is therefore the only one that deserves further analysis. Again, aspillers had a good comment: estoppel and laches may be available as defenses. But that is further down the line. The original author first needs to explain in the suit what action of yours was not permitted. That claim would be, in legal terms, the creation of a derivative work. Technically, that takes the shape of disassembled C code, but legally the particular shape of the derivative work does not matter.

Now the first defense of yours will be that regardless of the GPL version, you have the permission to make derivative works. Estoppel and laches are additional defenses - if you had permission anyway, there was no copyright violation. But this is a legal technicality: was the GPL offer direct permission or merely a promise thereof?

Either way, the burden of proof is on the claimant, and courts tend to take a rather dim view of claimants that caused problems for themselves. It would have been trivial not to mark the program as GPL.

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.