Back in the BIOS clone days, they had someone read the IBM BIOS code, and write a detailed specification from it. They had someone else never look at the IBM BIOS, read the specification, and write code to implement it. (This was the "clean room" approach - the IBM BIOS was never in the room of the implementers.)
But that's still "derived" in the sense that it implements the same functionality. But it's perfectly legal. So "derived" doesn't mean "implements the exact same functionality as the other, and we examined it in detail to make sure".
Why did they do the clean room approach? So that IBM could never claim that they had copied the IBM BIOS, even by re-typing rather than electronically copying.
Well, if you're re-implementing it in Rust instead of in C, you're not copying it, either. You're making a completely new implementation. (Rust doesn't take C code as valid syntax, so far as I know, so typing in the same code from memory wouldn't get you anywhere.)
Back in the BIOS clone days, they had someone read the IBM BIOS code, and write a detailed specification from it. They had someone else never look at the IBM BIOS, read the specification, and write code to implement it. (This was the "clean room" approach - the IBM BIOS was never in the room of the implementers.)
But that's still "derived" in the sense that it implements the same functionality. But it's perfectly legal. So "derived" doesn't mean "implements the exact same functionality as the other, and we examined it in detail to make sure".
Why did they do the clean room approach? So that IBM could never claim that they had copied the IBM BIOS, even by re-typing rather than electronically copying.
Well, if you're re-implementing it in Rust instead of in C, you're not copying it, either. You're making a completely new implementation. (Rust doesn't take C code as valid syntax, so far as I know, so typing in the same code from memory wouldn't get you anywhere.)