Subject:
Re: [discuss] licence problems
From:
Rick Moen ####@####.####
Date:
30 Sep 2008 20:00:33 +0100
Message-Id: <20080930190028.GE1041@linuxmafia.com>
Quoting Jean-Daniel Dodin ####@####.####
> Rick Moen a écrit :
>
> > You write and then upload a HOWTO whose pages include your copyright
> > statement and your explicit grant of permission under GPLv2.
>
> agree. My concern is the usual case when nothing is said in the text
> (no licence quoted). The copyright law say that is there is no quoted
> licence, the complete property remains with the author.
OK, so it appears that we are probably in agreement after all.
As I was saying, caselaw in copyright cases has long established that
owners can issue valid licences through their conduct, or spoken words,
or written words. It's always useful to consider how the particular
factual situation will appear, if it gets litigated and appears in front
of a judge: The judge will look at the facts and circumstances, to
determine what the parties did and said to establish what permission was
granted.
As you say, if a Web site prominently says "By contributing to this
site, you are agree to issue your work under licence [foo], and warrant
you have the right to do so", _and_ the judge agrees that contributor
really could have expected to see the notice (and that assent was real),
_and_ the judge doesn't see anything compelling to indicate the
contrary, then the implied site licence would likely apply.
Towards that end, the judge would probably find persuasive a notice _on_
the submission page like Wikipedia's:
Content that violates any copyright will be deleted. Encyclopedic
content must be verifiable. You irrevocably agree to release your
contributions under the terms of the GFDL*.
(Footnote goes to this text: "GNU Free Documentation License, Version
1.2 or any later version published by the Free Software Foundation; with
no Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
Texts.")
As noted upthread, if, however, the contributor specifically embeds into
his/her contribution a different set of licensing terms, and especially
if he/she makes very clear that that licence, rather than the site one,
is the copyright holder's choice, then I very much doubt the sitewide
licence would prevail. (Site maintainers would need to consider whether
they wish to accept the contributor's terms. If those terms are
compatible, e.g., new-BSD, then I'd say there's no practical reason why
not.)
Jean-Daniel, I hope you see, now, why I insist on using very specific
examples, when discussing licensing. ;->
> such problem can arise on a wiki, where much time can pass before
> somebody notice there is something wrong. wiki is a "control after"
> system.
Yes, quite true -- and inevitable on all systems that permit free
(free as in libre) editing. CMSes would be different, in that respect
-- but we're not using those (and I'm thankful for that).
[Old manifesto imposed rather minimal conditions.]
> so nearly any licence written by the author is allowed with very few
> conditions (copy and distribute).
>
> in a world where all doc are reviewed this can be a problem, but
> manageable, on a wiki...
Well, there are two parts to my answer to that. (1) The
control-after-submission aspect of wikis is _inherent_ in the way the
software works. It exists no matter what's in the manifesto. It just
has to be dealt with. There's really little point in trying to tightly
control that situation from inside the manifesto's wording, because
attempting to do so simply cannot work.
(2) Fortunately, it's actually not a significant problem. The amount
and timeliness of "review" required is minuscule. If someone merely
lazily looks over new HOWTO submissions for a number of things including
peculiar licensing once every few weeks, say, that should more than meet
LDP's needs. There are actually a lot more compelling, likely, and
significant aspects of new submissions than the slender possibility of
odd licensing: That can be among the minor details noted in passing,
and then raised to the staff's attention if/when it ever comes up.
And I note that we are not, in fact, (yet) being deluged by new HOWTO
submissions to the point that _any_ "review" work is burdensome.
> good so the only problem could be if the document *clearly* define an
> other licence, not if nothing is written?
As I've stressed repeatedly, I would not necessarily call
document-specific licensing a _problem_, of any sort. It depends on
what licensing that is. Pretty much all of the common licences would
be fine. Many of them are, in fact, rather more permissive than, say,
"GNU Free Documentation License, Version 1.2 or any later version
published by the Free Software Foundation; with no Invariant Sections,
with no Front-Cover Texts, and with no Back-Cover Texts", and thus
_reduce_ LDP's problems rather than adding to them.
All I was saying, and it turns out you agree, is that the existence of a
site-default licence does not automatically and positively settle what
terms apply to a contribution.
> > copyright law, not contract law -- would be the concept of a "contract
> > of adhesion" (http://en.wikipedia.org/wiki/Contract_of_adhesion),
> > a "take it or leave it" contract,
>
> interesting. will the mail exchange that exists most of the time
> between discuss list and the author have value?
You mean, as a copyright licence? As an element of a contract?
Please pardon my being a stickler for detail, but I always find it
helpful in licensing discussions to use very specific examples,
otherwise people find themselves differing over unclear or undefined
hypothetical problems.
Any conduct, or oral wording, or written wording, _can_ be sufficient to
establish intent to form a contract -- just as it _can_ suffice to
grant permission under copyright law. There are, in countries with
English-derived legal systems (not sure about Civil Law countries like
France) a couple of categories of contract that may be formed only in
writing, notably ones involving real estate. (In the English tradition,
this rule was established by the 1677 "Statute of Frauds".)
> there are some documents in our collection where the licence is not
> clear about the modification beeing allowed. if we find the authors
> letters in the discuss archives may be we can clear the fact?
Oh, OK, yes. If we can find any permission grant, anywhere, by a
copyright holder, that is useful. Remember, the hypothetical judge
is simply going to look for the clearest available indication of what
the copyright holder's licensing was.
