Re: On interpreting licences (was: KDE not in Debian?)
Scripsit Lynn Winebarger <owinebar@free-expression.org>
> Secondly, the claim is not that the API is copyrighted, it's that the code
> calling it is derivative of the implementation of the API.
Right, that is the claim that was to be rejected.
> There's a difference. You'd have to do some work to show me that in all
> cases a function call is equivalent to a footnote - footnotes you don't need
> to see to understand the text, a non-standard API I need to know at least
> the documentation for the library implementation (which probably describes,
> but not defines, the API).
Uh? Where does copyright law talk about meaning and understanding? If I write
a book about (say) The Rolling Stones or Microsoft Windows one can't really
understand the text without accessing (or having had access to) their music
respectively software, but that does not give them any copyrights to my book!
> In the case of a standard, the function call gets its meaning from the
> standard's definition rather than that of any particular implementation, so
> the calling code's meaning is independent of that library implementation.
So portable code is unaffected by any implementation's copyrights but
non-portable code is affected? This would be quite interesting (and absurd).
An example that you might be more familiar with is the WWW. Think of what it
would mean if linking to a webpage would give the (intellectual) owner of that
page copyrights over that page containing the link. And of course by extension
to any page linking to that page, etc. Well I'm pretty sure it would mean I
have copyrights over the entire WWW (and so would you). Yet it remains even
more true than for function calls that one cannot determine the meaning of a
weblink without following it.
Marc van Leeuwen
Universite de Poitiers
http://wwwmathlabo.univ-poitiers.fr/~maavl/
Reply to: