The key concept to take away from legal theory is that in the end, judges decide.

The uncertainty and doubt that DannyBee comments seem to instill is a bit regrettable, since its not that complicated. A derivative work is a legal concept built on a few vague concept, neither which has been clearly defined, and the precedential case law for software derivatives is almost non-existent. What does exist is an industry standard, and a defined intent written in the license text.

When a judge look at a copyright case, intent and industry standards goes a long way. Judges does not seem amused when people trying to find loop holes or game the system. Add that with the commonly commercial nature of cases that get brought to court, and I would not bet on the infringer getting a pass.

"The key concept to take away from legal theory is that in the end, judges decide. "

This depends. If whether it is a derivative work turns on a question of fact, the jury will decide it, not the judge.

"The uncertainty and doubt that DannyBee comments seem to instill is a bit regrettable, since its not that complicated. "

If you honestly believe this is "not that complicated", you have apparently found a way to solve a problem that has plagued software copyright law for 20+ years!

You then go on to describe that it is not clearly defined, and there is no case law, which seems complicated to me!

". A derivative work is a legal concept built on a few vague concept, neither which has been clearly defined, and the precedential case law for software derivatives is almost non-existent"

With you so far.

However, you then claim intent and industry standards go a long way: "When a judge look at a copyright case, intent and industry standards goes a long way."

Which is it?

Is there no precedential case law, or is the precedential software case law saying "intent about derivative works is what drives what a derivative work is". Because it can't be both!

In any case, there is literally no industry standard here, and no intent in the written license text. The GPL says nothing about it (the word "derivative work" appears twice in the GPL 2.0, and neither is a definition). Large numbers of authors using the license disagree on what it means (happy to give specific examples if you like, i deal with this literally every day).

"Judges does not seem amused when people trying to find loop holes or game the system." I agree with this, but it seems, at least to me, irrelevant to this discussion.

A judge, a jury, it depend on legal system. In the legal system that Mojan resides in, Sweden, the lowest court is made up by 3 politicians and one judge which then vote on the issues at hand. To simplify, lets call it "a judge decide", but in truth, it depend on the justice system in the nation that the case is put forth in.

> plagued software copyright law for 20+

For something to plague, it has to inflict it. The question about derivative work is not brought up in cases around copyright law. Lawyers are not arguing about it in court, judges are not deciding about it, and juries are not asked to vote on it. At most, it has plagued software legal scholars, but then that claim would have to be supported by statistics from academia.

>Is there no precedential case law...

Industry standards is not precedential case law, its just the defined status quo. If one for example read the US supreme court decision around Aereo, one find that industry standard and status quo has a considered impact on judgments. The legal system in Sweden commonly uses industry standard when defining legal theory in unprecedential areas (See the Pirate Bay trial for example).

> In any case, there is literally no industry standard here

I strongly beg the differ. Larger commercial companies both uses GPL and LGPL, and the praxis is to differentiate the licenses on linking. A company like HP and Apple will distribute GPL software, and will combine LGPL and proprietary software. Someone could argue in court that the perceived praxis do not exist, but I suspect a number Amicus curiae would then be sent in arguing the opposite.

> no intent in the written license text

"The GNU General Public License does not permit incorporating your program into proprietary programs." - GPLv2 and GPLv3

If a client of yours take GPL licensed software and incorporate it into a proprietary program, I would strong doubt arguing the definition of derivative work in court will yield successful result. Its the primary reason in my view why there is no precedential case for derivative work in GPL licensed software.

The FSF, while the license writer, is not the copyright owner, or the contracting party. Their view of what should/should not be a derivative work, listed somewhere in an FAQ nobody here has probably read, would mean precisely nothing in court :)

In fact, in the only US cases to have a say on the GPL and definition of derivative works, the FSF's view was never asked for or cared about.