Subject: Re: Open Source and Contributor Agreements
From: "Philippe Verdy" <verdy_p@wanadoo.fr>
Date: Sun, 27 Nov 2005 17:33:20 +0100

From: "Brian C" <brianwc@ocf.berkeley.edu>
> That's so inaccurate it makes it appear you are trying to deliberately 
> misunderstand.
>
> If you're sincere, then answers are interspersed below:
>
> David Barrett wrote:
>> Still seeking closure on this.  Can anyone confirm or deny the following 
>> reasoning:
>>
>> - Open source has an unstated principle that all contributors are equal; 
>> nobody has any greater or lesser rights to the source than anyone else 
>> (copyright excluded).
>
> With the big proviso "copyright excluded" (and patent) this probably is a 
> principle of open source. However, it's hardly unstated. I'd say the open 
> source definition points 5, 6, and 7 are all directed at this.

There does seem to exist here a confusion between open "source" (a source 
that anybody can work on), and an open "project": open projects are in fact 
quite rare (Wikipedia being a well known exception), and require privileges 
on resources that are owned by a few participants. Maintaining the 
separation between the project and the source that it works on may help 
solve the ambiguities.

It is natural for any project to have a working committee and limitation of 
members to that working group. It is still natural for such project that 
work on that shared source that they open their source, without necessarily 
opening their working group.

Forking is still a risk, but the success of forks widely depends on the 
creation of a managed working group with its own policy, and the capacity of 
the working group to follow the advancements made in the concurrent 
"project" working on the same open "source".

"Open source" means that forking is accepted, any attempt to restrict that 
would disqualify the source as being open: users of an open source have the 
right not only to create their own modifications and publish them, but also 
have the right to group together in an alternate "project". Then it's up to 
each managed working group to decide which modification to integrate in 
their own version of the source, and it should remain possible from one 
group to benefit from the work made in another forked project: this avoids 
the forks from completely diverging, and in fact it allows better 
cooperation on the same source, for the benefit of all users.

Of cource, each managed group must take its own security mesures to avoid 
pollution of their version by copyviolating contributions. This means that 
not all contributions are acceptable in a "project", but they still remain 
usable in private works by the contributors that will assume themselves the 
copyright risk, if they choose to distribute those modifications.

For these reasons, it is essential that all contributors mark their name and 
include an history of their modification with their distribution, and that 
managed projects take preventive measures such as being allowed to reject 
insufficiently qualifying contributions.

The contributor agreement is then a separate contract, not related to the 
open-sourced software usage, but on the collaboration of that contributor 
with the managed project: it is valid for a project to fix limitations on 
such contributions, and does not change the nature of the open-sourced 
project.