[cc-community] Strengthening the CC-BY-SA copyleft with respect to code

Maciej Pendolski beholder0x100 at gmail.com
Wed Dec 21 04:45:30 EST 2011

On Wed, Dec 21, 2011 at 5:47 AM, Bart Kelsey <elbarto at gmail.com> wrote:
> The answer is this:  If the media and the program are not *part of the same
> work*, then it doesn't matter.  This should apply in instances where someone
> writes a game or some other program that is *intended to load a specific
> piece of media*.  Depending on the situation, that might be an entire game
> engine or it might be just the media and game scripts that run on top of
> that game engine.

The question to me is: where could something be considered "part of
the same work"? I am not a lawyer but it seams to me that intent has
little (if any) legal effect (I might be completely wrong). No matter
how much directly is data being used/accessed, how can you tell that
game data directory is not just a "document" in an "interactive
document format"?

Then if in a "non-generic" engine there would be a game data directory
selector how would it be different from e.g. ScummVM
(emulator/replacement of a game engine used for several adventure
games)? If ScummVM would be pointing just to a specific
directory/"document"/game_content and no other game would exist for it
then would it be considered as "part of the same work" just because it
only points to a specific directory vs. allowing to select a
directory? That's a tiny difference which could change interpretation
/ perceived intent.

How much is generic generic? An universal game engine (a rather
abstract thing) would be the most generic system. FPS, RTS, ... game
engines would be generic within a genre but not overall, then we go
down until we reach some point at which we arbitrarily decide whether
it is generic enough or not (or decide based on "intent").

I guess we would need some specific example(s) of various "levels" of
being "generic" to see if there is some pattern that we can clearly
label as "generic". If "generic" means how flexible engine is then I
cannot easily tell when engine is flexible enough and when it is

It also gets back to the debate about data being functional. E.g. I
could take BZFlag (game with tanks) and make terrain look like water,
replace meshes with inflatable boats, if weapon properties were in
data files then I could turn tank cannons into high rate guns with
meshes and sounds replaced to say, micro-harpoons launchers, ... And
then it would be pretty much another game. And if someone would still
have some issues with that then a simple "document selector" should
fix the issue. And then I could think of several other data sets to
show that that game engine is flexible "enough".

Is a specific data set necessary for engine to work at all? It seams
that in this case it is not. With game engine with at least a tiny
flexibility in form of configuration file (overall physics, unit
motion, weapons, ... control) and with artwork replaced, an engine
could be proclaimed "generic".

Of course it would not hurt (at least not much) if game engine
developers who are making open source games, do it even more open by
making their engines more flexible (and to at least some elements). It
is not that much to ask for.

And we need a lawyer to tell us why we are all wrong ;)

More information about the cc-community mailing list