Milos writes:
Again, personally, I don't have problems with it. However, I think that the present construction of the attribution issue is far from well defined and that it leaves WMF projects in extremely vulnerable position. Just a small group of malicious persons may make a real mess with attribution conditions. And probability for that is huge.
I disagree that there is a "huge" probability of legal exposure with regard to this question. I follow moral-rights jurisprudence reasonably closely, and I have yet to see any reason to believe that the risk of legal action against the Wikimedia Foundation (or anyone else) is going to be increased by the new licensing scheme. (Indeed, if the probability were huge, we would already have seen such cases, since the GFDL prescriptions are more exacting than the CC-BY-SA prescriptions.)
If something is not legally valid, it may not pass as an argument at
the court.
I'm known to be risk-averse with regard to legal exposure for the Foundation, but I am not terribly troubled by the prospect you raise here. Moral-rights doctrine is grounded in assumptions about authorship that don't map well to massive collaborative enterprises like Wikipedia. Explaining to a court how we do handle attribution in the context of such enterprises -- including the fact that we make an effort to do reasonable attribution -- should convince most reasonable courts that the purposes for which moral-rights doctrine was invented are being served. I lose sleep over other kinds of legal issues relating to Wikimedia projects, but not this one.
--Mike
On Fri, Mar 20, 2009 at 10:41 PM, Mike Godwin mnemonic@gmail.com wrote:
I disagree that there is a "huge" probability of legal exposure with regard to this question. I follow moral-rights jurisprudence reasonably closely, and I have yet to see any reason to believe that the risk of legal action against the Wikimedia Foundation (or anyone else) is going to be increased by the new licensing scheme. (Indeed, if the probability were huge, we would already have seen such cases, since the GFDL prescriptions are more exacting than the CC-BY-SA prescriptions.)
Free content became important legal form of content since Wikipedia. And we are leaving the time when Wikipedia was a miracle with extreme amount of good faith around it. Also, all contributors were us and we don't want to sue us.
As Wikipedia is becoming more and more a regular part of our civilization, we may expect more and more regular behavior. We already had malicious legal attacks in UK, Germany and France (I remember those three issues). We had a number of informal settlements, even in Serbia (a relatively famous person, a daughter of famous Yugoslavian director and translator, wanted money for some quotes at Bosnian Wikiquote).
By building a position with significant holes, and attribution issue is still a significant hole, we are making unsustainable construction. If we have, let's say, 10.000.000 of contributors and 1% of them (100.000) is not happy with Wikipedia because of any reason and 1% of them (1000) want to sue WMF or whoever and 1% of them can do it, we'll have 10 big problems. We may fail in just 10% of the cases and we'll suffer from significant consequences.
I'm known to be risk-averse with regard to legal exposure for the Foundation, but I am not terribly troubled by the prospect you raise here. Moral-rights doctrine is grounded in assumptions about authorship that don't map well to massive collaborative enterprises like Wikipedia. Explaining to a court how we do handle attribution in the context of such enterprises -- including the fact that we make an effort to do reasonable attribution -- should convince most reasonable courts that the purposes for which moral-rights doctrine was invented are being served. I lose sleep over other kinds of legal issues relating to Wikimedia projects, but not this one.
Note that the whole idea related to the license migration and relaxing attribution is because of content reusing, and just in small part because of Wikipedia itself (while I see benefits for Wikibooks and Wikiversity, I don't see any significant benefit for Wikipedia). Also, Wikipedia and other Wikimedia projects are attributing all contributors.
So, reusers may be sued for plagiarism ("Copyright Me and Wikipedia authors", even "Me" contributed much less significant material than 20 Wikipedia authors). After the first couple of such processes Wikipedia recommendations would loose any credibility. (And you should know better if it would be possible that WMF would be sued for misleading recommendations.)
BTW, it won't be an issue if we are living in a world without bourgeois egotism :) But, we aren't living in such world.
Technically, our biggest problem may be removing 100.000 edits. And it is possible to be done without endangering the project.
But, we may suffer from harder consequences: If someone insists to be attributed and we say that that person has to be attributed, others will follow that example. This is especially true for projects out of English language ones. English Wikipedia, for example, has large basis of contributors who are contributing there because of free knowledge. Other projects may have significantly different contributor basis. Around a year ago around five very active contributors (of ~40 of very active contributors in that time, if I remember well) left Serbian Wikipedia. They made another project and they chose CC-BY-NC-ND license there. If they would be at the project at the time when the first contributor demands and gets attribution, I guarantee that those five contributors would be the first who would follow that contributor. In that situation I know the next ~10 contributors who would demand attribution just because previous five got... And we would end with one of two options: (1) general rule that all authors have to be attributed or (2) that 90% of authors say that they have to be attributed. So, again, useless recommendations with transitional mess.
Note, again, that this is not an issue directly related to Wiki[pm]edia, but to reusing of Wiki[pm]edia content.
And about moral rights doctrine: If it is not scaled well to massive collaborative enterprises, the answer is not not to address them at all, but to scale it. (I am talking again something which I don't want... I want moral rights not to be addressed, but are you sure that it will work?)
BUT, if you think that there is no reasonable threat to be sued for "misleading recommendations", it doesn't cost a lot to try that way. Fixing credibility is much less dangerous than loosing two years budget.
On Fri, Mar 20, 2009 at 4:16 PM, Milos Rancic millosh@gmail.com wrote:
As Wikipedia is becoming more and more a regular part of our civilization, we may expect more and more regular behavior. We already had malicious legal attacks in UK, Germany and France (I remember those three issues).
I'm aware of these, and other legal threats as well.
By building a position with significant holes, and attribution issue
is still a significant hole, we are making unsustainable construction.
My assessment is different from yours.
If we have, let's say, 10.000.000 of contributors and 1% of them (100.000) is not happy with Wikipedia because of any reason and 1% of them (1000) want to sue WMF or whoever and 1% of them can do it, we'll have 10 big problems. We may fail in just 10% of the cases and we'll suffer from significant consequences.
This is a version of Pascal's Wager. I don't really believe, however, the risk is even as high as you suggest here. We'll be fine.
After the first couple of such processes Wikipedia recommendations would loose any credibility.
I don't consider this a significant risk.
BUT, if you think that there is no reasonable threat to be sued for "misleading recommendations", it doesn't cost a lot to try that way. Fixing credibility is much less dangerous than loosing two years budget.
I don't think there's any reasonable threat of this sort.
--Mike
2009/3/20 Mike Godwin mnemonic@gmail.com:
If we have, let's say, 10.000.000 of contributors and 1% of them (100.000) is not happy with Wikipedia because of any reason and 1% of them (1000) want to sue WMF or whoever and 1% of them can do it, we'll have 10 big problems. We may fail in just 10% of the cases and we'll suffer from significant consequences.
This is a version of Pascal's Wager. I don't really believe, however, the risk is even as high as you suggest here. We'll be fine.
Pascal's Wager involves infinite gain/loss - this is just basic risk analysis and has nothing at all to do with Pascal's Wager.
I think the percentages given as plausible, but do we really have 10 million contributors? The English Wikipedia apparently has 9,237,657 registered users, but I believe a very large proportion of them have never made an edit, an even larger proportion won't have any edits which still exist in articles. I find it very unlikely that there are 10 million contributors, even across all Wikimedia projects, that have copyrightable contributions. (Of course, I'm ignoring anons - I don't see how they can realistically sue for copyright infringement.) So I think the expected number of problematic cases is significantly less than 1, but it certainly isn't 0.
On Fri, Mar 20, 2009 at 5:15 PM, Thomas Dalton thomas.dalton@gmail.comwrote:
Pascal's Wager involves infinite gain/loss - this is just basic risk analysis and has nothing at all to do with Pascal's Wager.
It's true that Pascal's own version of Pascal's Wager involves the risk of infinite loss, but it's commonly used among modern philosophers (and lawyers and others) to refer to scenarios regarding small probabilities of very large losses (since in the real world we humans mostly don't deal with infinite losses or gains). If I tell you there's only a 1 percent change of a 10-trillion-dollar personal loss (for example), the same paralyzing logic of Pascal's Wager applies, even though 10 trillion dollars isn't an infinite number of dollars. (And that's all I'll say here about Pascal's Wager on this list -- if you want to discuss it privately, I'm happy to continue that off-topic discussion, since I've been interested in the subject for almost three decades.)
So I
think the expected number of problematic cases is significantly less than 1, but it certainly isn't 0.
Fortunately for legal risk assessment, there's no need for risks to be zero in order for them to be judged insignificant.
--Mike
On Sat, Mar 21, 2009 at 1:15 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
I think the percentages given as plausible, but do we really have 10 million contributors? The English Wikipedia apparently has 9,237,657 registered users, but I believe a very large proportion of them have never made an edit, an even larger proportion won't have any edits which still exist in articles. I find it very unlikely that there are 10 million contributors, even across all Wikimedia projects, that have copyrightable contributions. (Of course, I'm ignoring anons - I don't see how they can realistically sue for copyright infringement.) So I think the expected number of problematic cases is significantly less than 1, but it certainly isn't 0.
We'll have. If you start with just 100.000 contributors and raise percentage to 10% (which may be reasonable too), you'll end with 100 cases.
But, it is reasonable to suppose that Mike's legal predictions are more relevant than mine :) So, legal part is no issue anymore for me.
The only issue which stays is related to users which declare that they want to be attributed: Would we allow that? If yes, is there any plan ("yes" is good answer enough) how to deal with making attribution recommendations useless? If no, is it possible it legally?
2009/3/21 Milos Rancic millosh@gmail.com:
On Sat, Mar 21, 2009 at 1:15 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
I think the percentages given as plausible, but do we really have 10 million contributors? The English Wikipedia apparently has 9,237,657 registered users, but I believe a very large proportion of them have never made an edit, an even larger proportion won't have any edits which still exist in articles. I find it very unlikely that there are 10 million contributors, even across all Wikimedia projects, that have copyrightable contributions. (Of course, I'm ignoring anons - I don't see how they can realistically sue for copyright infringement.) So I think the expected number of problematic cases is significantly less than 1, but it certainly isn't 0.
We'll have. If you start with just 100.000 contributors and raise percentage to 10% (which may be reasonable too), you'll end with 100 cases.
10% doesn't sound at all reasonable to me.
But, it is reasonable to suppose that Mike's legal predictions are more relevant than mine :) So, legal part is no issue anymore for me.
Only the last percentage is really a legal prediction, the rest are more predictions of human behaviour. That said, the last percentage is probably the only one that anyone can give anything more than a wild guess at (which is why I didn't choose any stronger words than "plausible" to describe them).
On Sat, Mar 21, 2009 at 1:57 AM, Thomas Dalton thomas.dalton@gmail.com wrote:
10% doesn't sound at all reasonable to me.
In one of the previous emails I described that 5 very active contributors were not happy with the situation at sr.wp at the time when there were ~40 very active contributors. I don't think that sr.wp had much more than 100 important contributors and I may count 10 unhappy. So, the first 10% have some empirical grounds. Actually, even the next 10% have empirical grounds: I know the person who would sue sr.wp community if they can. Fortunately, the next 10% are 0.1.
But, it is reasonable to suppose that Mike's legal predictions are more relevant than mine :) So, legal part is no issue anymore for me.
Only the last percentage is really a legal prediction, the rest are more predictions of human behaviour. That said, the last percentage is probably the only one that anyone can give anything more than a wild guess at (which is why I didn't choose any stronger words than "plausible" to describe them).
I was not EFF lawyer and Serbia is an irrelevant place for copyright and patent issues: we have 1 (one) person with PhD in copyright :)
wikimedia-l@lists.wikimedia.org