Raul Miller wrote:
> On Tue, Feb 15, 2000 at 12:02:31PM -0500, Andreas Pour wrote:
> > I think this is where you went off-track. Section 2 only refers to
> > source code distributions (as it requires the modifications to be
> > distributed under Section 1 and Section 1 deals only with source
> > code).
>
> I disagree.
>
> Section 0 includes the text:
> <quote>
> (Hereinafter, translation is included without limitation in
> the term "modification".)
> </quote>
>
> Section 2 begins:
> <quote>
> 2. You may modify your copy or copies of the Program or any portion
> of it, thus forming a work based on the Program, and copy and
> distribute such modifications or work under the terms of Section 1
> above, provided ...
> </quote>
>
> Since executables are a translation of the source, the provisions of
> section 2 also apply to executables.
It's obvious that a binary is a derived work of the Program, but that
does not mean that every single provision of the GPL applies to it. You
can only apply what makes sense to apply. Section 2 requires
distribution under Section 1 -- which explicitly refers only to source
code. If you can explain to me how you can distribute a binary under
Section 1, then we can explore the conditions to such distribution
(found in 2(a) - 2(c)). Note that Section 1 specifically says:
You may copy and distribute verbatim copies of the
Program's *source code* as you receive it.
It could have left out "source code" if it was to mean what you say it
means. But since those words are there, we must give them meaning.
Since w/out the terms "source code" there, a binary might be covered by
Section 1, one can only conclude that a binary does not fall under
Section 1.
Since the conditions to Section 2 apply only to distributions of
modified works under Section 1, and since Section 1 does not apply to
binaries, the conditions of Section 2 do not apply to binaries.
Isn't logic wonderful?
Incidentally, it is my reading that translation does not refer to
compilations into binary. I think it refers to translation into another
human-readable language -- like translating C++ into C, or C into Perl,
or C into assembler. That way use of the terms makes sense -- Section 3
deals with binaries and Section 2 with source code (including
translations into other languages). Also, it is not common usage to
refer to a binary as a "language" -- rather, a binary is referred to as
a set of"instructions". Binary is really not a language -- nobody can
read it :-). For dictionaries noting the difference b/w computer
languages and machine code, see, e.g.,
http://www.instantweb.com/~foldoc/foldoc.cgi?machine+code and
http://www.instantweb.com/~foldoc/foldoc.cgi?programming+language.
Incidentally, if notwithstanding the very clear language you believe
that Section 2 does apply to a binary, Section 2(a) says:
(a) You must cause the modified files to carry prominent
notices stating that you changed the files and the date
of any change
If you can explain to me how Debian does that with the binaries it
distributes, I would consider myself enlightened.
Ciao,
Andreas
