[Savannah-register-public] [task #14621] Submission of Graph Model Library

[Brook Milligan]

Follow-up Comment #2, task #14621 (project administration):

The comments by Ineiev seem to make the following three points:

- Some files lack in-file copyright notices.
- Savannah supports free software (perhaps this is not).
- A personal judgement is given regarding the contributor agreement.

Below I address each in turn.

In-file copyrights
=============

As far as I can tell, the following is the entire list of files that lack
copyright notices:

CLA-entity.pdf
CLA-individual.pdf
VERSION
docs/building_models.md
docs/class_hierarchy.md
tests/gdal/data/test_raster.tif
tests/gdal/data/tl_2010_35_county10/tl_2010_35_county10.dbf
tests/gdal/data/tl_2010_35_county10/tl_2010_35_county10.prj
tests/gdal/data/tl_2010_35_county10/tl_2010_35_county10.shp
tests/gdal/data/tl_2010_35_county10/tl_2010_35_county10.shx

The first two are LaTeX-generated files included for convenience and clarity. 
I do not believe they can have copyrights injected.

The third is a one-liner that simply states the version number.  Does this
really need a copyright notice?

The two *.md files can have copyright notices, but that injects the notices
into the Doxygen output in totally inappropriate places.  I feel it is better
to leave the notices out of these two files and preserve the clarity of the
generated documentation, rather than slavishly follow a policy that all files
should have copyright notices.

The remainder are binary files that cannot be modified to include other
elements without breaking the file formats.

Overall, given the prominence of COPYING (following GNU guidelines) and the
presence of notices in the remaining files (over 700), I feel that the copying
terms are clearly spelled out.


Free software
==========

The point of this comment is unclear, but it would seem to imply that this
software is not free software.  However, it is distributed under the GPL3,
which by definition of the Free Software Foundation makes it free software. 
Every guideline that FSF uses to define free software is concerned with the
rights of people to use the software in various ways.  The entire point of the
GPL3, as described extensively by the Free Software Foundation (e.g.,
https://www.gnu.org/philosophy/free-sw.en.html) is to protect those rights for
the users of software.  By virtue of being distributed under the GPL3, this
software cannot be more free according to these definitions.


Contributions
==========

There are many points of view regarding the nature of contributions to
software projects, even within the free software community.  It seems to me
that the freest approach is to maintain the rights of users of the software,
to maintain the rights of contributors with respect to their contributions,
and to maintain the rights of the project to advance and protect the project
as best as possible.  Anything else imposes restrictions.

The distribution of this software under the GPL3 permanently protects the
rights of users, and therefore makes this free software by definition.  The
contributor agreement is intended to maintain the other rights.

I feel that the best strategy with respect to contributions for a nascent
project is entirely unclear, because there is much conflicting evidence and
much uncertainty.  However, ambiguity is likely not among the useful
strategies.  Putting aside philosophical feelings, which cannot be evaluated
substantively, the analysis I refer to  suggests that the strategy taken here
allows contributors to retain all rights to contributions.  It also allows
"Us" to defend the project against infringement, something we would
not have otherwise.  Further, while individual contributors might have that
right under some alternative approaches, it is not at all likely that they
would prevail.  Thus, the clauses in the GPL3 that stipulate various policies
on derived works are likely to be ineffective without an interested entity. 
What good are those clauses in protecting users' rights if they cannot be
protected?

Taken together, these points suggest that the approach taken here is in fact
the most compatible with broad goals of free software: protecting users,
protecting contributors, and protecting projects.  

This project is unambiguously free software by virtue of releasing code under
the GPL3.  Whether or not it will foster a community of developers depends on
the technical details the library seeks to address and the willingness of
contributors to participate.  As a new library, it remains to be seen how this
experiment will play out.  There is, however, no single correct answer to the
question, what is likely to promote the most robust development community? 
Further, that is not a defining property of free software, nor could it be
because it is entirely speculative.

    _______________________________________________________

Reply to this item at:

  <http://savannah.gnu.org/task/?14621>

_______________________________________________
  Message sent via/by Savannah
  http://savannah.gnu.org/