That's what's interesting. You probably had to give code that wasn't itself licensed as GPL because that code was covered by the terms of the GPL license (through linking, i presume). That's the entire purpose of GPL! It makes-free any code that directly touches GPL code. Ya know, like Midas and gold.
That's why it's such a powerful force for software freedom if the terms hold up.
As far as I know, the effect of linking to GPL code hasn't actually been tested in court.
The license is intended to impose obligations in the case of linking, but dynamic linking does not make a copy of the library at build time; it just generates enough metadata for the program to call the library. One might reasonably argue that no derivative work is created by dynamic linking, or that it is only created when the end user runs it.
EULAs are enforceable because the program is copied into RAM at runtime (a bad precedent, I think), but the GPL is not a EULA and only imposes requirements on distributors, not end users.
> he effect of linking to GPL code hasn't actually been tested in court.
the intent of the GPL family of LICENSE is clear. There certainly will be efforts to diminish its reach by motivated parties. Tests in US courts are certainly a function of the depth of the pockets of litigants, no?
I'm not even trying to address realities like the potential unfairness of the courts here. I'm saying it would be very reasonable to conclude that linking isn't copying and a copyright license isn't required to do it.
Nah not the code in question it was a bunch of completely different subprojects in different languages, but given that the end product was massive I think they wanted to be sure they could use it. Its possible they wanted to evaluate the product.
That's why it's such a powerful force for software freedom if the terms hold up.