[sf-lug] Potential GPL violation, with the potential intent to victimize our Elderly loved ones

Rick Moen rick at linuxmafia.com
Mon Nov 16 19:39:45 PST 2009

Quoting Jesse Zbikowski (embeddedlinuxguy at gmail.com):

> On Mon, Nov 16, 2009 at 12:36 PM, Rick Moen <rick at linuxmafia.com> wrote:
> > Establishing in court that a second work
> > is meaningful only in the context of an older work does _nothing_
> > towards showing copyright infringement.
> Thank you for clarifying your point.  I'd like to investigate this
> further because I would imagine there are legal opinions to the
> contrary.

Not by judges.  ;->

But go ahead and have opinions to the contrary without bothering to read
the relevant caselaw.  Why should you be different from most
computerists on the Internet?

> For instance "fan fiction" is considered derivative work in
> this country, trademark issues aside, even with no copying from the
> original.

Actually, it's (that is, typical examples are) considered derivative
work under copyright law _specifically because of copying_ of creative
elements from the original.

I warned you that copying embroiled in copyright issues can be either
literal or non-literal.  Remember?  See the Computer Associates, Inc., 
vs. Altai decision.

Unauthorised fanfic in which characters named Batman and Robin have
adventures, G-rated or not, in Gotham City are violations of DC Comics's 
copyrights because of court recognition of various expressive elements
in the way those characters were written by Bob Kane and friends --
despite lack of literal, verbatim re-use of Kane and co's panel drawings
and dialogue. 

> My interpretation is that this is because it is intimately
> tied to the protected work.

Your interpretation is mistaken.  It's not because the second work is
"meaningful only in the context of" the first one, but rather because of
unauthorised use of an excessive number of copyright-covered expressive
elements without a sufficient positive defence.

(Aqua's _Barbie Girl_ was able to win against a copyright _and_
trademark suit from Mattel, Inc., because the reuse of character elements 
was held to be lawful fair use under copyright law, and as parodic
nominative use under trademark law.  Judge Alex Kozinki threw out
Mattel's suit with the comment "The parties are advised to chill.")


> It would be great if you could cite a particular case; I gather from
> your "Licensing and Law" page the case history is quite vast, and I'm
> unlikely to finish it in time to continue this discussion.

It is not "quite vast", and the entries relevant to derivative works and
copyright law are pretty obvious.  You could bother to look.  Do you
think I'm your personal free-of-charge research librarian?

> > No, that is not what I said.
> Sure, please tell me when you say "It would get dismissed as
> irrelevant to the legal issue" what you mean by "it" and what you mean
> by "the legal issue". I was just guessing, apparently incorrectly.
> Like a binary module, the sentence only makes sense in context :-)

Sorry, insufficient time.  If you don't remember the context for your
own question, look it up in the archives.

> You're right in that we are talking about derivatives of an EXPRESSION
> when we talk about derivative works. It's hard to talk about
> expressions without mentioning ideas because If You Really Want To Get
> Into It expressions are also ideas: there's not really a firm boundary
> between them. 

Oh, for heaven's sake, man.  Look, your remark suggested that you had
the two concepts confused.  That's what I meant, and it seems to have
been justified.

> Quoting Judge Learned Hand in the summary opinion of Computer
> Associates v. Altai: "Nobody has ever been able to fix that boundary,
> and nobody ever can."

Yes, I've read CAI v. Altai, too.  You seem to have missed the point: 
It's no more a laser-bright line than any of the other key distinctions
in law, but it _is_ the entire foundation of the distinction between
copyright and patents -- and is the origin of the non-copyrightability
of elements of a creative work deemed functional or required for
compatibility as opposed to expresssive.

> Yes indeed, you can draw your own house, just as you can write your
> own kernel. That's an excellent way to avoid copyright infringement.
> Let's talk about the case where you don't want to do this; you want to
> use MY work and publish a modification to it.

Oh very well.  Architect A creates a blueprint and files it with the
county.  Architect B (that would be me, in your hypothetical) finds that
blueprint, says "Say, that's really nice", brings a copy home, feeds it
through an utterly humongous digital scanner, and somehow ray-traces it
so as to bring it functionally back into AutoCAD with very little
change.  He re-does some stuff, adds some things, files it with the
county as a plan for a new house.  Architect A (that would be you, in
your hypothetical) sees it, spots what he thinks are damning and
outrageous borrowings, and files suit for alleged tort of copyright
infringement.  Some lucky judges gets to hear the parties.

The judge would probably take strong meds for his/her resulting
migraine, then bring in some professors from the local architecture
school to give a micro-course on architectural design and use of
AutoCAD.  His Honour would be seeking to try to learn _really quickly_
what is and is not reasonably deemed a creative aspect of a building
blueprint.  Moreover and probably earlier than that, he'd go and enrich
Lexis quite a bit, finding and reading applicable precedents.  For that
matter, architects A & B's lawyers would likewise be shovelling ducats
at Lexis, each trying to argue precedents they can argue to be similar
to the way they want the case to work out.

I'm not an architect, nor am I likely to spend a small fortune
researching architecture copyright litigation just to discuss the
hypothetical that you haven't actually bothered to flesh out, so I'll
just say that the judge is going to be guided by the same leading cases
as would apply in any other copyright infringment case.  Which you can
read, if interested.

Please don't ask me to read blueprints -- or tea leaves -- to try to
figure out what are expressive elements thereof. 

> > In any event, you seemed to be arguing, to further use your analogy,
> > that as designer of the blueprint for a particularly distinctive sort of
> > house, any compatible carport blueprints that are "merely a
> > modification or elaboration" of the house design, or are designed to
> > "plugin" to it, or are designed solely to work with it, are
> > automatically derivatives of the original copyrighted blueprint.
> > But, no, that is not the case.
> Right indeed that's not the case... you are free to make a
> loosely-coupled "add-on" to my design and say "insert this between
> pages 12 and 13". You might even get away with saying "replace
> everything in the middle section of page 2 with this." But if you say
> "mine is just like his except the windows are slightly larger and the
> door is smaller, and I use red lines for supporting walls instead of
> blue" -- you run into some difficulties.

If you literally or non-literally copy an excessive number of expressive
elements of a copyright-covered creative work without permission, and
don't fit into one of the delineated exception as an affirmative
defence, then you are indeed likely to lose an infringement case.  I am
not an architect, nor have I done any reading on the law of
(specifically) _blueprint_ copyright infringement, so, if you want to
know in detail -- beyond the guidelines of leading cases -- what are
considered expressive elements in building drawings, you'll have to hire
a very specialised attorney.  Good luck with that.

> First of all how do you even make a concrete expression (blueprint in
> this case) of your idea without copying from me?

You're the guy supposedly constructing the hypothetical.  Why are you
asking _me_ questions about it?

> Yes you could publish a high-level description like the above, but if
> you actually made it into a drawing you'd have to copy.  

See above.  Free your mind of the assumption that copyright law somehow
bans anything that Jesse Zbikowski _calls_ "copying".  Read the law about
what copyright covers, and what are deemed expresive and what are not.

> Consider Torvalds' claim that you can publish the SOURCE of a Linux
> module under any non-GPL compatible license you want -- it's only the
> concrete binary instantiation that's derivative.

I have no idea why you're raising this.  But, if Torvalds said that,
then he was voicing nonsense.  He's actually said quite a bit of
nonsense about copyright over the years -- much of it stuff that
contradicts other stuff he's said at other times.  He has said less
nonsense lately, presumably on account of better access to legal advice
at Linux Foundation.

> Interesting. If it's not a derivative work, then on what grounds can I
> enjoin you from distributing it? Because I very much can under US
> copyright law, by my reading. If you deny that's a derivative work, I
> think your opinion is in conflict with 17 U.S.C. 110(11) as
> established by the previously mentioned Family Home Movie Act of 2005.

Not so fast.

You spoke of a work consisting of _just_ a series of timecodes plus
independent footage, but none of the original work.

Family Home Movie Act of 2005 (written into law as 17 U.S.C. 110
paragraph 11) says that making use of _literal copying of parts_ of the
original audio or video work to make a new audio or video work is OK if
the original copy used is otherwise authorised and the borrowed elements
are "imperceptible" and no fixed copy of the new work remains at the end
of the process.

So that is quite different from what you said.  Unless perhaps the term
"timecode" means "non-trivial portion of the original work's expressive
content, literally copied".  

> Now why is distributing a set of your own clips plus a timecode a
> violation of my copyright? BECAUSE IT'S INTIMATELY TIED TO MY

You should read 17 U.S.C. 110(11) before making dumb and sweeping claims
about it.

Anyhow, would you mind going and getting a different hobby, other than 
dreging up endless amounts of bullshit about copyleft licensing?  I need
to get some work done.

More information about the sf-lug mailing list