[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: The draft Position statement on the GFDL

@ 12/05/2004 16:12 : wrote Josh Triplett :

Humberto Massa wrote:
1. I get gcc's sources;
1. (a) I have a valid license to it;
2. I get metafont sources;
2. (a) I also have a valid license to this;
-- up to this point, no license violation
3. I make modifications to gcc and to metafont, taking care of :
3. (a) not removing any copyright (C) notices -- they are already
I don't need to put them, I received gcc under the terms of the GPL,
with the notices, and the disclaimer (as to satisfy GPL#1);
3. (b) marking the changed files as changed as to satisfy GPL#2, 'a';
3. (c) gcc does not have the announcement in GPL#2, 'c', so nothing
is required;
3. (d) I will take appropriate similar precautions stated in
license (OPL?) in making the modifications to metafont's files;
-- up to this point, no license violation
4. I will write my files needed to integrate both, taking all the
necessary precautions 3 a-d above. Notice that probably my files
completely unrelated) are derived works both of metafont and of gcc.
-- no license violation.
5. I will diff the sources from the resulting program with the

This diff is a derived work of your program and the original sources.

-- no license violation.
6. I will write a script that like this:
  mkdir ~/metagcc; chdir ~/metagcc
  tar xzvf $GCC_SOURCES
  patch -p1 ../../metagcc.patch

metagcc.patch is a derived work of your metagcc, which is a derived work
of both gcc and metafont, so you cannot distribute metagcc.patch unless
it satisfies the terms of gcc's license and metafont's license.

Even if that is not the case, wouldn't this script constitute
"contributory infringement"?
Only if this is the case (if I can't distribute metagcc.patch). I don't know about metagcc license (which I think is the OPL, but I'm not certain of it). And contributory infringement is an USofA-jurisdiction-specific thing, here in Brasil there is no such entity. And I'm sure it's the case in many places.

My primary tought is that: not containing gcc nor metafont code, and being on its entirety of my original copyright, EMPHASIS: being entirely *my* intellectual creation, I can license metafont.patch differently, I am the sole copyright holder to it, as I can license the script, but this can be wrong. I'll think a little bit more. The script does not seem to be a derived work on any of them.

Again, all this is does not apply in the case of Gentoo/OpenSSL I mentioned:

1. take a GPL'd program that uses GNUTLS.
2. make some alterations so it now is OpenSSL compatible.
3. make the thing.ebuild that will link them at a later time.

But yes, you DO have a point. Or two.


Reply to: