(I'm not trying to antagonize, I just want to extend this thought experiment.)

I'm guessing the argument is "Well a big legal entity of a company has access to lawyers and legal advice that would interpret the GPL and advise, whereas the average individual simply does not have that knowledge or resources." I think that makes sense? Not sure what happens in a court of law, obviously.

By the way, you can't redistribute copies of Windows arbitrarily either. In fact you can hardly redistribute it at all.

This is why I firmly believe the only way out of this mess is an international standard of common "service / license guarantees" and "terms and services" and hell, even "privacy policies" patterns that can be reused across software and services to immediately illuminate freedoms, limitations, and guarantees.

What if I take the source code for "ls", turn it into a library, so that I could do the same thing but more directly -- the library is obviously a derivative work, but would my program that calls it also be a derivative? (note, in this case the hypothetical library would be GPL, not LGPL).

And, if I want to link to a GPL (not LGPL) library, if I write a wrapper for it that turns it into a stand-alone executable (sort of like how you can call openssl functions, or you can run the openssl executable passing command line arguments), does one form of calling create a derivative and the other form not?

No, output of a program isn't the program itself, so it doesn't fall under copyright[1]. Why would you expect that it would be a derivative work? It should be obvious that e.g. when gcc outputs a compiled program it doesn't automatically become tied to the GPL.

The only exception - which is mentioned[2] in the GPL - is if the output of a GPL licensed program is part of the licensed program. (quines, other rare utilities)

    2. Basic Permissions.

    [...] This License explicitly affirms your unlimited permission
    to run the unmodified Program. The output from running
    a covered work is covered by this License only if the output,
    given its content, constitutes a covered work. [...]

Probably[3], but see the FAQ[4] for answers to common situations and a real lawyer for actual legal advice.

> does one form of calling create a derivative and the other form not?

Correct. They are different. The intent is that the GPL licensed code is still open. Users that receive copies of the new stand-alone executable can exercise their right to examine and modify that code. They can easily replace it with a version they have modified without touching your other programs that merely use the output.

[1] https://www.gnu.org/licenses/gpl-faq.html#GPLOutput

[2] https://www.gnu.org/licenses/gpl-3.0.en.html

[3] https://www.gnu.org/licenses/gpl-faq.html#IfLibraryIsGPL

[4] https://www.gnu.org/licenses/gpl-faq.html

My point, however, wasn't these specific scenarios, I was using them as examples as examples in response to the parent's comment inferring that you know exactly what it means. "Consult a lawyer" kind of just reinforces the point.

Of course, I do agree that most community licenses such as the GPL are infinitely more readable than most proprietary software licenses.

I don't know the law where this particular action is being brought, but speaking generally, it's not unusual for different safeguards to apply for private citizens and for businesses. For better or worse, the situation where I am tends to heavily favour the private citizen when it comes to standard form contracts or other similar agreements with businesses, while businesses (no matter their size) are basically assumed to know what they're doing so it's their own fault if they agree to something they shouldn't.

It's not a great analogy anyway, for the reasons others have already explained.

The legal argument around fair contracts is that a customer should not be tricked into agreeing to a unfair or hidden condition. For example, I can't sell a car and have a shrink wrap license that gives me permission to still use it after sale (ie, steal it back). Such contract condition would very fast be found to be illegal, and common sense say that a "sale" implies that I as a seller do no longer enjoy rights such as using the sold car after sale, regardless of trickery I might try with shrink wrap licenses.

To go back to the GPL case, its not a unfair contract condition that you need copyright permissions to distribute a copyright protected work. For the first thing there is no contract, and the second reason is that the condition is created by society and not the copyright holder.

Yup — at least if you're in the U.S.; I can't say as to other jurisdictions. (IAAL.)

Indeed, in the U.S. it's close to an iron-clad rule: You're presumed to have read the contract you signed — and even if you didn't, with rare exceptions you're still stuck with its terms.

> No "normal" people read or care about these kinds of documents, nor should they have to. Judges are starting to realize this, and react accordingly.

In the U.S., historically that's been true only in an extraordinarily-narrow set of circumstances. Look up "unconscionability" and note the distinction between procedural and substantive unconscionability [0]. I've seen no evidence of any general trend otherwise.

In summary: (A) RTFC. (B) "Ought to be" ≠ "is."

[0] See, e.g., https://www.law.cornell.edu/wex/unconscionability and https://en.wikipedia.org/wiki/Unconscionability

IANAL but can't imagine that a mouse click can be given the same weight as the usual execution of a contract, normally involving verfied ID, a notary public, and written signatures.