Re: copyright notices (was: Re: Are we (nearly) ready for 3.4 yet?)

[Olaf Till]

On Tue, Jan 04, 2011 at 06:49:45PM -0500, John W. Eaton wrote:
> On 15-Dec-2010, Søren Hauberg wrote:
> 
> | >   * Update copyright years.
> | 
> | You actually need to do that? I did not know that.
> 
> Yes.
> 
> Actually, the FSF recommends updating the copyright year in all
> files in any year they are published.

Really? It seems not appropriate to me if nothing in the file has
changed. And in this Howto of the FSF:

http://www.gnu.org/licenses/gpl-howto.html

it is said: "The copyright notice should include the year in which you
finished preparing the release (so if you finished it in 1998 but
didn't post it until 1999, use 1998)."

> This now somewhat complicated by things like
> 
>   Copyright (C) 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002,
>                 2003, 2004, 2005, 2006, 2007, 2008, 2009 John W. Eaton
>   Copyright (C) 2009 Jaroslav Hajek
>   Copyright (C) 2009, 2010 VZLU Prague
> 
> (in src/data.cc) or 
> 
>   Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009 David Bateman
>   Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004 Andy Adler
>   Copyright (C) 2010 VZLU Prague
> 
> (in liboctave/Sparse.cc).
> 
> One good thing is that the FSF has now changed its position on using
> ranges of years instead of requiring that each year be enumerated
> separately, so these statements can be condensed somewhat, but that is
> only part of the problem as I see it.  The bigger problem is deciding
> exactly which of these statements should be updated with the new year.
> All of them?  That has not been done in the past, and I think it is a
> bit strange to retroactively add someone to the copyright statement
> for a prior year if the first change they made was this year (for
> example).  And in the first second case above, I think we stopped
> adding new years for Andy Adler since he was no longer working on the
> code.  But that is not consistent with what we would do if the
> copyright were held by a single person.  In that case, we would update
> the list of years regardless of whether there were any changes to the
> file made by the person listed in the copyright line.

As already said above, I'd doubt that the latter is appropriate if
nothing in the file has changed. The copyright should only be updated
for copyright holders who made changes in the respective year, even if
there is only one copyright holder ...

But if it should be chosen to update the copyright year nevertheless,
then IMO there is no reason to exclude any copyright holder from this
update.

And surely there should be no copyright assignment for years in which
a person has not (yet started to) change(d) the file.

Olaf