[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Need to Identify Contributions and the Dissident Test



On Thu, 20 Jan 2005, Glenn Maynard wrote:
> On Thu, Jan 20, 2005 at 05:09:03PM -0800, Don Armstrong wrote:
> > 1) Some sort of identification of the author of the work is required
> > in order to allow people to exercise their DFSG guaranteed freedoms
> > upon a work.
> > 
> > If we did not have some sort of identification of the copyright holder
> > of the work, the work is (probably) not properly licensed, and thus we
> > cannot make use of it at all. This seems to break most copyleft
> > schemes.[1]
> 
> Copyright notices can use aliases, right? I don't know anything
> about how enforcable that renders that person's copyright claim, but
> I don't think it renders the license invalid.

At least in the US, the copyright would still be enforceable if they
actually wrote the software, since a copyright notice is no longer
required. (Well, ignoring the effect upon statutory damages.)

However, an improper copyright + licensing notice could make the
license itself invalid (or at least questionable) since it wouldn't be
a clear statement from the copyright holder that they licensed a work
appropriately.

> > 2) The purpose (as I understand it) of the dissident test is to
> > point out licenses which require disclosure of information to
> > inviduals to whom the software has not actually been distributed.
>
> I believe "don't make me identify myself" is part of the dissident
> test: a dissident identifying himself as the author of something can
> put him in personal danger; he should be able to modify and use the
> software without violating the license (eg. so, once he moves to
> Canada and identifies himself as the author of those changes, he
> doesn't find himself being sued for copyright violation).

Absolutely. In this case, the author doesn't have to disclose
information to anyone to whom the software hasn't been distributed. If
you don't distribute the software, no one should need to know that
you've modified it.

However, when you actually distribute the modifications that you've
made, some sort of disclosure seems necessary so that other people can
build upon your modifications. If this disclosure isn't provided,
you've effectively restricted people from using your modifications at
all, even though that seems central to the Free Software community.

The situation that keeps appearing to me is the following:

I release a Free work. Someone else comes along, takes the work, makes
useful modifications to it, and sells it to companies and distributes
it to the world at large. However, they fail to identify themselves.
Thus, I have no way of knowing who actually wrote the improvements to
the work or finding out if the copyright behind them is actually
sound. In fact, for all anyone knows, the "someone else" could have
been a person working while legally contracted to a company, and all
of that work, even though released under a pseudonym, is owned by the
contracting company.

In this way, all of the improvements made to the software are
unavailable to the Free Software community because no one can either
sue the copyright holder of the improvements to cause them to comply
with the GPL, or worse, incorporate the improvements back into the
work.

> > 3) GNU GPL 2a) obstensibly requires this very same thing:[2]
> > 
> > 	You must cause the modified files to carry prominent notices
> >         stating that you changed the files and the date of any change.
> 
> > 2: Others will probably argue that it doesn't, since 'stating that you
> > changed the files' doesn't necessarily mean that you actually have to
> > give your name.
> 
> I don't think there's any debate here. You don't have to give your
> name. I've never seen a serious argument to the contrary.

This is somewhat of an open question, especially as the typical way to
satisfy this clause means identifying the source of the changes.

At least, that's how I read it.


Don Armstrong

-- 
<Clint> why the hell does kernel-source-2.6.3 depend on xfree86-common?
<infinity> It... Doesn't?
<Clint> good point

http://www.donarmstrong.com              http://rzlab.ucr.edu

Attachment: signature.asc
Description: Digital signature


Reply to: