[RFC] what should we do about COPYING?

[RFC] what should we do about COPYING?

[Joel Brobecker]

Michael Snyder pointed out that gdb/COPYING contains the GPLv2,
whereas GDB is now under GPLv3. I was simply going to replace the
contents of COPYING with the GPLv3, except that I noticed an
inconsistent resolution depending on the project. For instance:

  - In the root directory, we have COPYING (v2) and COPYING3 (v3);
  - In bfd, we have COPYING that contains the GPL v3.

I personally do not see the point of keeping the GPL v2 around,
and thus suggest that we replace the contents COPYING rather than
adding an extra file named COPYING3.  Note that we have a couple
of extra copies in the sim directory as well.

-- 
Joel

Re: [RFC] what should we do about COPYING?

[Stan Shebs]

Joel Brobecker wrote:
> Michael Snyder pointed out that gdb/COPYING contains the GPLv2,
> whereas GDB is now under GPLv3. I was simply going to replace the
> contents of COPYING with the GPLv3, except that I noticed an
> inconsistent resolution depending on the project. For instance:
>
>   - In the root directory, we have COPYING (v2) and COPYING3 (v3);
>   - In bfd, we have COPYING that contains the GPL v3.
>
> I personally do not see the point of keeping the GPL v2 around,
> and thus suggest that we replace the contents COPYING rather than
> adding an extra file named COPYING3.  Note that we have a couple
> of extra copies in the sim directory as well.
>   
The root dir has to cover all bases; there may be src/ projects that 
have not updated.  But gdb/ and below is all v3 (right?), there would be 
no code to which a copy of v2 would apply, and so we want just the one 
file COPYING that is v3.

Stan

Re: [RFC] what should we do about COPYING?

[Michael Snyder]

Joel Brobecker wrote:
> Michael Snyder pointed out that gdb/COPYING contains the GPLv2,
> whereas GDB is now under GPLv3. I was simply going to replace the
> contents of COPYING with the GPLv3, except that I noticed an
> inconsistent resolution depending on the project. For instance:
> 
>   - In the root directory, we have COPYING (v2) and COPYING3 (v3);
>   - In bfd, we have COPYING that contains the GPL v3.
> 
> I personally do not see the point of keeping the GPL v2 around,
> and thus suggest that we replace the contents COPYING rather than
> adding an extra file named COPYING3.  Note that we have a couple
> of extra copies in the sim directory as well.
> 

I agree.  v2 COPYING should just go.

Re: [RFC] what should we do about COPYING?

[Pedro Alves]

On Friday 18 December 2009 17:33:59, Michael Snyder wrote:
> Joel Brobecker wrote:
> > Michael Snyder pointed out that gdb/COPYING contains the GPLv2,
> > whereas GDB is now under GPLv3. I was simply going to replace the
> > contents of COPYING with the GPLv3, except that I noticed an
> > inconsistent resolution depending on the project. For instance:
> > 
> >   - In the root directory, we have COPYING (v2) and COPYING3 (v3);
> >   - In bfd, we have COPYING that contains the GPL v3.
> > 
> > I personally do not see the point of keeping the GPL v2 around,
> > and thus suggest that we replace the contents COPYING rather than
> > adding an extra file named COPYING3.  Note that we have a couple
> > of extra copies in the sim directory as well.
> > 
> 
> I agree.  v2 COPYING should just go.

Note that there is code in sim that is v2 or later, copyrighted Andrew Cagney, not FSF.

Also, in gdb/ it appears some new gplv2 or later files have crept
in:

 >grep "version 2 of" * -rn
 ChangeLog-2003:15347:   again by default on systems with version 2 of the GNU C library.
 configure:1605:                          systems with version 2 of the GNU C library (use
 configure.ac:928:  AS_HELP_STRING([--without-included-regex], [don't use included regex; this is the default on systems with version 2 of the GNU C library (use with caution on other  system)]),
 COPYING:299:    the Free Software Foundation; either version 2 of the License, or
 darwin-nat-info.c:11:   the Free Software Foundation; either version 2 of the License, or
 doc/Makefile.in:9:# the Free Software Foundation; either version 2 of the License, or
 doc/gpl.texi:359:the Free Software Foundation; either version 2 of the License, or
 doc/observer.texi:14:@c Foundation; either version 2 of the License, or (at your option)
 features/Makefile:7:# the Free Software Foundation; either version 2 of the License, or
 regformats/regdat.sh:103:   the Free Software Foundation; either version 2 of the License, or
 solib-spu.c:10:   the Free Software Foundation; either version 2 of the License, or
 spu-multiarch.c:10:   the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.ada/gnat_ada.gpr:5:--  the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.java/jprint.java:13:// the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.base/source-error.gdb:7:# the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.base/gcore-buffer-overflow.c:7:   the Free Software Foundation; either version 2 of the License, or (at
 testsuite/gdb.base/gcore-buffer-overflow.exp:5:# the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.base/watch_thread_num.c:7:   the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.mi/mi-async.exp:5:# the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.modula2/unbounded-array.exp:5:# the Free Software Foundation; either version 2 of the License, or
 testsuite/gdb.modula2/unbounded1.c:7:   the Free Software Foundation; either version 2 of the License, or

What's the story behind gpl.texi?

-- 
Pedro Alves

Re: [RFC] what should we do about COPYING?

[Joel Brobecker]

> Note that there is code in sim that is v2 or later, copyrighted Andrew
> Cagney, not FSF.

I wonder why this code is copyrighted Cagney and not the FSF, since
I thought that one requirement for pushing to the FSF tree was
assigning your changes to the FSF. Not sure what to do next...

Regardless, there is a COPYING at the root of the sources, so perhaps
we can rely on version 2 of the license being available there.

> Also, in gdb/ it appears some new gplv2 or later files have crept
> in:

Fixed most of them.  I haven't had time to look at gpl.texi yet,
so I have no idea what this file is used for yet.  I suspect that
we should have updated that file too.

-- 
Joel

Re: [RFC] what should we do about COPYING?

[Alfred M. Szmidt]

   > Also, in gdb/ it appears some new gplv2 or later files have crept
   > in:

   Fixed most of them.  I haven't had time to look at gpl.texi yet, so
   I have no idea what this file is used for yet.  I suspect that we
   should have updated that file too.p

It is used to show the license conditions of GDB in the manual.

Re: [RFC] what should we do about COPYING?

[Stan Shebs]

Joel Brobecker wrote:
>> Note that there is code in sim that is v2 or later, copyrighted Andrew
>> Cagney, not FSF.
>>     
>
> I wonder why this code is copyrighted Cagney and not the FSF, since
> I thought that one requirement for pushing to the FSF tree was
> assigning your changes to the FSF. Not sure what to do next...
>   
Back then, simulators in GDB were still somewhat of a novel concept, and 
we waffled about whether to officially consider them part of GDB, or as 
a separate library that happened to be linked in - the simulator could 
always be built as its own program, and GDB could be built without 
simulator. Plus at the time we really really really wanted a PPC 
simulator. :-)
> Regardless, there is a COPYING at the root of the sources, so perhaps
> we can rely on version 2 of the license being available there.
>   
That's good enough for me,  in general people expect that foo/COPYING 
applies to everything in that directory and its subdirs, but not to 
sibling dirs.

Stan

Re: [RFC] what should we do about COPYING?

[Joel Brobecker]

> Back then, simulators in GDB were still somewhat of a novel concept,
> and we waffled about whether to officially consider them part of
> GDB, or as a separate library that happened to be linked in - the
> simulator could always be built as its own program, and GDB could be
> built without simulator. Plus at the time we really really really
> wanted a PPC simulator. :-)

We still need the ppc simulator :) (although we have also started testing
the ppc-elf GDB over qemu, with good results too).

I don't see the rush in addressing this inconsistency with the rest
of GDB, but I have a feeling that this will need to be addressed at
some point. The argument that the sim is a library separate from GDB
might have held for the sim ppc (assuming it was published elsewhere),
but I am not sure it works for all sim implementations.

But for now, I will take a break from licensing issues :).

Thanks for the feedback!
-- 
Joel