> This is highly dependent on what exactly you do to the code, and what exactly we're > asking about it. Sure, if you write some entirely new functions or blocks of code, > you could retain copyright over those. But if you're modifying code that was already > there, or copying it and then using it elsewhere with modifications, then you > wouldn't have copyright over that code.
I covered this exact point with "as long as that code is copyrightable, not all code is". Whether it's in a new function, a new block of code or in an existing loop is irrelevant.
> Certainly, you couldn't distribute the entire > file, so you wouldn't have rights to a usable program.
I didn't say that you did, I'm just saying that you retain the copyright to any copyrightable code. Unless you come to an agreement then neither party can distribute the code, because you both own copyright to various parts.
> To be extremely precise, here is what I'm saying you can't do: > 1) Take code that was released under license A. > 2) Make substantial modifications to the code, but not to the extent that it has been > entirely rewritten. > 3) Release the modified code under an incompatible license B, without permission from > the original author. > 4) Claim shared copyright over the entire code. > > If I'm understanding his arguments correctly, this is why Haze didn't want to list > his name as a copyright holder at the top of files he worked on.
You don't have a right to do point 3, but if you have made substantial modifications to the code in a way that is copyrightable then you definitely have the right to do point 4. It is quite normal to have multiple copyright assertions on a product when multiple parties have claims, individually they only have the rights to certain parts. This is no different. Violating someone elses copyright doesn't diminish your copyright.