>>106348144
>>106348174
>>106348194
Are they writing a plugin using existing GCC functionality or are they modifying GCC to enable new functionality. That's the key question a court would ask.

If they are modifying GCC then that means they are subject to its license. If they are simply using existing plugin functionality that's already found in the compiler then that's not a violation.

Anyway, this whole argument is stupid. They can just change to LLVM/Clang and be completely compliant. In fact, I don't think they distribute binaries anyway, only patches, so this complaince issue is actually an issue for the end user that chooses to compile it with GCC instead of Clang.