[Foundation-l] Copyright issues of wikimedia projects

Sun May 30 11:03:00 UTC 2004

Daniel-
> It also happens in actual articles. [[Nick Berg]] and [[Abu Ghraib prisoner
> abuse reports]] for example. But I do see a distinction: Articles on
> Wikipedia would synthesize the whole topic while Wikinews articles would
> just cover the breaking news on that topic. However I'm skeptical that the
> average user will grep this distinction, resulting in a content fork for
> recent history.

News articles provide as much information as possible on a single event,  
even a single event in a larger story such as the Abu Ghraib scandal.  
There would be no Wikinews article called [[Abu Ghraib scandal]]. There  
would be a Wikinews article called [[Abu Ghraib contractor faces new  
inquiry - May 28, 2004]]. After its release, there would be a fixed time  
limit within which we can publish updates to a Wikinews article. Once that  
time has passed, any new developments would have to be covered in a new  
article, just like in regular news media.

Such articles are intended for someone who is familiar with the big  
picture of a situation, but wants to know the most recent development in  
as much detail as possible. The half-life of an individual article would  
be rather low; after a couple of weeks, it would turn into a source  
document, referenced by other Wikinews articles ("Recent stories on this  
topic:") and by Wikipedia.

Such developments, in Wikipedia, would often be summarized down to one  
paragraph or even one sentence ("Mufti Nizamuddin Shamzai was killed by  
unknown assailants on May 30, 2004 [1]"), where the Wikinews article would  
contain as much detail as possible, quotes from all relevant parties  
(which we can freely copy from outside sources), speculation about the  
future impact etc.

Furthermore, I anticipate that original reporting will begin as soon as  
the Wikinews project is launched, and that is something which Wikipedia  
can never do.

> Wikipedia's ability to be up-to-date is one of the most
> cited good things about us and we should be "very" careful not to jeopardize
> that.

I see no such jeopardy. Wikinews and Wikipedia are highly complementary.  
Where they do similar things - synthesize information from outside sources  
- they do it with a different level of detail and with a different goal.  
Wikipedia will benefit greatly from Wikinews just like Wikipedia benefits  
from [[current events]] - additional links to non-existent or short  
articles will lead to additional detail being added. The articles  
themselves are very different in nature, style and scope.

> If and when we get a
> problem with people giving first hand reports, then we can seriously think
> about starting Wikinews, IMO.

That would be a fatal mistake, based on the assumption that people will do  
stupid things which are clearly not allowed and not wanted on an  
encyclopedia. We should not just start projects to deal with problems on  
Wikipedia, we should also start them when there is a clear, separate goal  
which is within the mission of the Wikimedia Foundation.

An "If and when we get a problem" mentality for starting new projects is  
very, very dangerous. Please think about whether that is really the  
position you want to adopt.

> I'm sure that will happen someday, but I don't
> think that will be a problem until Wikipedia.org is in the top 100 or so
> list of websites on the Internet.

That may well be the case, if you set the limit "We will start Wikinews  
when people start reporting original news on Wikipedia". That would,  
however, be a very silly thing to do. In fact, if that is the limit we set  
for starting Wikinews, it may very well never come to pass, because any  
attempts at original reporting on Wikipedia will be stopped so quickly  
that no significant trend to do so can ever develop.

The question we should ask is of course not "When will people start  
original reporting on Wikipedia?" but "Can we find enough people already  
to do useful original reporting on Wikinews?". I believe we can. Take a  
look at Indymedia, which, like Wikipedia, is an international project. It  
has a far higher traffic ranking than Wikipedia - about 4000 - yet it  
manages to be present at major world events.

Indymedia has an extremely leftist bent and this is reflected by its  
choice of topics (protests, scandals, conspiracy theories etc.), but other  
than that, it is a good example that grass-roots reporting can work. This  
would be our primary competitor when working on Wikinews, not Wikipedia.  
And unlike Indymedia, Wikinews would have a Top-500 website that can  
infect it with volunteers to do the reporting, and a constant flow of free  
images and content that can be used to do so. The summarizing of outside  
sources would be primarily a service to cover the things which we do not  
cover ourselves.

> We should by then have dozens of
> Wikimedians in each large city of the world and thus have a large potential
> pool of journalists on the ground.

You don't need a large pool of journalists to be useful, a small but  
growing pool is already good enough. If in our first weeks our original  
reporting only covers Linux conferences, kernel releases and baseball  
events, that doesn't matter. That's how Wikipedia started out - we didn't  
wait until we had people from all fields of knowledge.

>> You appear to be under the misconception that a license different from the
>> FDL would automatically mean license incompatibility. While this is, to
>> some extent, true in the direction ''FDL text'' => ''text in another
>> license'' (because the FDL requires that all derivative works are FDL-
>> licensed), it is not necessarily true in the direction ''text in another
>> license'' => ''FDL text''.

> Free content is not free when it cannot be freely copied back and forth
> between the source and the derivative work.

It's not that simple. When dealing with licenses, you're dealing with a  
mix of freedoms which can be balanced and juggled to reach various goals.

>> This specifically addresses your concern of using Wikinews content to
>> update Wikipedia articles. Given that Wikinews is not intended to be an
>> encyclopedia, I think that background information from Wikipedia is best
>> provided using links, which is not a problem.

> This would be different from a great many news articles I read - the good
> ones always include background information.

That's because in a New York Times article, you have no other choice - you  
can't just say "Oh, and if you want to know more about who that cleric is,  
check out Wikipedia" because people can't easily get to Wikipedia from the  
printed newspaper. Wikinews would include background information, but it  
would be typically a summary of Wikipedia - "Nizamuddin Shamzai was a pro- 
Taliban cleric who had called for a holy war against the United States;  
see the Wikipedia article for details." Thus it would be somewhere between  
"no background information" and "the level of background information of a  
good printed newspaper article."

Interestingly enough, online news outlets increasingly include links to  
Wikipedia.

> That is why I am proposing the creation of a GNU Free Content License that
> invariant-free GNU FDL text can be migrated to. Then the only change in the
> GNU FDL 2.0 that would need to be made is a clause stating that:

> "Any document licensed under the GNU Free Documentation License version 2.0
> or any later version that does not contain Invariant Sections, Front-Cover
> Texts, or Back-Cover Texts, can be licensed under the GNU Free Content
> License version 1.0 or any later version."

That's all nice and good, but it doesn't address the main problem - all  
existing documents which meet these conditions would be affected. How many  
of these documents are there? I have no idea. Do you? Unfortunately, there  
is no central registry of FDL texts. I believe that the FSF will only  
agree to such a change if we can demonstrate that no harm to existing  
materials will result.

>> Indeed, and the main question of copyleft vs. public domain / attribution-
>> only licenses is "What is more important to me - enlarging the body of
>> content available under a free license, or making sure that as many people
>> as possbile will be exposed to the content in question?"  The possibility
>> that we may want to answer this question differently for some Wikimedia
>> projects than for others should not be discounted out of hand.

> The point is to expose as many people as possible to the content, yes. But
> that can only be done by having it under a free content license. If we
> release into the public domain, then work on the content gets forked and
> improvements made to the forks can not be back-ported to the original.

That is a prediction which is not necessarily true. Even if it is true,  
you appear to assume that the work in question would have been done at all  
under a copyleft license. That is doubtful, particularly for Wikinews. The  
typical usage example for a Wikinews article would be a daily newspaper  
which wants to have an alternative to the Reuters feed. These people would  
likely be turned away by the FDL (although not necessarily by the FCL, if  
we can make it work), hence no actual contributions would be lost.  
However, the level of exposure to the content would be lower -- depending  
on which newspaper we are talking about, by orders of magnitude.

Even if we start with the FDL or FCL, we may want to have a clause in our  
submission standards that states that content may also be licensed under  
any other free content license (as in free distribution and free  
modification) chosen by the Wikimedia Foundation in the future - a kind of  
"opt-out" clause for the copyleft principle if it can be shown to hurt us  
more than it helps us.

Regards,

Erik