The not-quite-so-open part of opensource...
I started using OpenFOAM around 2006 and felt liberated by its openness. Unlike commercial CFD codes where you moan about things and then wait for a new version to see if it gets better, here you can actually do something about it. In my case, this liberty also felt like an obligation to give back. So not long after starting with OpenFOAM, I was posting code on cfd-online.com and sending tar files to Henry Weller (OpenCFD at that time) for incorporation in the OpenFOAM code base. This was naturally all a bit chaotic - no version control and manual code merges - but by 2008 things had finally moved to git, which made it not only much easier to contribute but also easier to track where the contributions actually came from.
Well, I suppose from the title of the posting, you have already guessed that none of the technological advances actually resulted in any transparency in the code contributions (also known as “Code Provenance”).
The unwritten policy was that any contributions should not have the author’s copyright mention, with the justification being that a single copyright makes it far easier to enforce GPL violations. However, the danger that a single copyright holder presents for any opensource software remains significant. A single copyright holder can unilaterally change to a proprietary license (although the last opensource version would remain open) or they can opt for a dual licensing model with special closed-source proprietary features being offered to their direct customers. These points are made in the presentation on opensource licenses by Bradley Kuhn (Software Freedom Conservancy) given at the OpenFOAM workshop in 2011.
I placed good faith in assurances that the record of my contributions would not be “lost”, since they are of course preserved in the git history. I understood this to mean that my code authorship would be correctly recognized and documented in due course. I have waited a long time for this to happen of its own accord and now decided that the time had come for my contributions to be properly attributed.
In the period from 2008 to 2011 (which predates my move to ESI in 2016 to work on OpenFOAM full time), I made 1300+ commits spanning 600+ files. The number of actual files modified is, of course, much higher, but we naturally ignore style and compilation-related changes, simple bug fixes and any scripted changes (for example, the change from GPL-2 to GPL-3+ licensing) etc.
I approached the OpenFOAM Foundation Ltd. (who claim sole copyright ownership and where Henry Weller is now company director) by opening a report on their issue tracker (text follows at the bottom of the post). Their response was to delete both the issue and my bug reporting account without comment.
I subsequently submitted the changes as a github pull-request, which was also immediately closed by them without comment.
I would be interested to hear other people’s experiences with contributing to opensource projects, especially with regards to German copyright law. Note that there is also an ongoing discussion on cfd-online regarding the “Original OpenFOAM author(s)”
Please re-share (also via other media).
[The following is the text of the issue filed and now deleted]
Accurate Code Provenance
During the period 2008 to 2011, I authored and contributed code to the OpenFOAM code base (as acknowledged in the release notes for versions 1.6 to 2.0), but which has never been reflected in the source code contents, and effectively suppressed in the git repository published by the Foundation.
It is to be noted that since contributions were authored within Germany the copyright is non-transferrable (as correctly noted in 2.2 of the contributor agreement).
Since I have never signed a Contributor Agreement (or equivalent) to waive my rights, I request that my rights as stipulated by section 13 of the Urheberrechtsgesetz (German Act on Copyright and Related Rights) be respected.
The author has the right to be identified as the author of the work. He may determine whether the work shall bear a designation of authorship and which designation is to be used.
The authorship designation shall have the form
Copyright (C) date-range Mark Olesen
to be located in header of the files in question. The designation shall appear in chronologically sorted order where other authorship (copyright) designations appear in the file headers to avoid an implicit devaluation or denigration of the contribution.
The text here largely corresponds to that of my similarly titled linkedIn article