Re: FRR package in Debian violates the GPL licence
On March 21, 2019 4:21:22 PM UTC, David Lamparter <equinox@diac24.net> wrote:
>You can't copyright words.
Not words, but as you say, you can copyright characters.
And characters have names.
And such names are the first expressions of the copyright, that is called
copy-right because if regulate the copy of those expression: guess what?
In a textual work, the first sign of a copyright violation is the presence
of those names.
You can name ONE character of a completely unrelated novel "Harry Potter"
(suppose no trademarks were involved), but if your novel also have an
"Hermione Granger", a "Ron Weasley", a "Draco Malfoy" and a "Tom Riddle",
it's probable that any Judge would take your argument "it's just a name" as
outrageous, an insult to their intelligence.
Now count how many names defined in GPL code are used babeld files not distributed as GPL.
How many names defined in command.h are used in babel_interface.c?
>> To be honest, the form of the text (in C, plain english, or
>> mechanically translated to an object form for mechanical
>consumption),
>> I'd say that it shouldn't change much.
>
>It changes by incorporating additional material. If a C file contains
>no include statements, only the copyright of the source applies to the
>object file. But creating an accumulative work is very much a legal
>concept (in software as well as in art.)
FRR isn't an anthology of different unrelated works.
>> All translations of a book are derivative work of the original, even
>> if they are ALSO copyright of the translators.
>
>Yes, translations of books are derivative. But a translation of a
>word,
>or even a sentence, isn't. (Caveats apply - you can probably cram
>enough content into a sentence to make it copyrightable. This boundary
>is something that courts determine on a case-by-case basis.)
And what about a sequel?
It's a totally new work and yet it is a derivative work.
Your babel_interface.c is just like a sequel of command.h
>There are actually legislations that have enshrined this in law, as a
>reaction to Microsoft's marked position in the 90ies. They explicitly
>specify that any parts of a system that are required for interoperation
>with that system can be freely copied regardless of copyright or
>license. Sometimes this also extends to reverse engineering.
This is a good objection and I'll take it at heart for my project of
conquering the world with Free Software.
But I think it would hard to sustain in a court that the internal headers
of FRR (a GPL work) marked as GPL code themselves are a system interface
needed for interoperability.
>> So why do you think that this is a "toxic precedent"?
>
>I agree that this is a toxic precedent, because in your world we need
>to reimplement everything from scratch at once.
Thanks for your opinion.
I'll seriously take it into account.
For now I don’t agree.
Note however that linking this FRR violation of section 2 of GPLv2 with the
Oracle vs Google case is convenient FUD but very arguable.
The FRR code evolved as a whole, the lib/ folder is not really a public API,
but a collection of utilities for internal use within the project.
GPLv2 section 2 explicitly exclude any section of a covereed work that is INDEPENDENT
from the GPL code, and give you the right to modify the program as a whole, adding,
changing and removing files according to the license itself.
You can add original AND independent work as new files under any license just because of it.
And you can add original BUT derivative work as new files too, but only under GPL.
By violating this second condition you terminated your license on the GPL code without affecting any third party.
You can bet Paul won't sue you.
You might be right or wrong.
You can bet nobody would help him to enforce the GPL because of fear of the API copyright issue.
I guess you are right on this: it's more likely that "free lobbist-activists" will start
to blame the victim, calling him copyright troll or something like that.
But your license has been terminated and anyone dustributing your code without changing the broken headers is going to terminate their own license too.
I still think you can practically fix the issue by changing the license of those derivative files to GPL.
But beware that a Judge might disagree and simple impose you to chease and desist any further activity on the FRR codebase.
Anyway please, stop arguing that babel_interface.c is not a derivative work of control.h.
Argue that it's fair use or something, but don't negate the evidence.
Such kind of behaviour hurts the credibility of our whole profession.
Giacomo
Reply to: