Re: test modules and license

[Bruno Haible]

Karl Berry wrote:
> Meanwhile, I'd like to ask about this unchanged sentence:
> 
>     The source files always say "GPL", but the real license
>     specification is in the module description file.
> 
> I don't understand.

The README and the .texi documentation of gnulib say that the license
statement in the module description override the copyright notice of
the file. This matters only to the user of the gnulib CVS; once
gnulib-tool has installed files in a project, they carry the "real"
license statement.

> Legally, what counts is the license in the actual
> source file.  You can't point to some other file and say "but it says
> LGPL there so it's really LGPL".

Why not? If we ensure that every user of the gnulib CVS understands it,
we can. Legally, what counts is the agreement (common understanding)
between two parties. I can give you a 1-dollar bill and tell you that
with it you can buy goods for $2 from me; then what is legally binding
is what I promised you, not the writing on the bill.

But I agree that it can be confusing. Paul, the original motivation for
this system was to ensure that people who copy the files manually into,
say, coreutils have "GPL" written everywhere. Now, this situation has
changed: Nearly all gnulib users use gnulib-tool. We could therefore
invert the mechanism: for an LGPLed module,
  - put the real LGPL license statement into the lib/* files,
  - have gnulib-tool replace this license with the GPL statement,
    unless --lgpl is specified.
The practical benefit would be a reduced confusion. The practical
drawback would be that the --symlink option, in the coreutils situation,
will copy more files and symlink less files.

What do you think, Paul, Jim?

Bruno