Sent off-list by accident:
---------- Forwarded message ---------- From: Thomas Dalton thomas.dalton@gmail.com Date: 2008/10/24 Subject: Re: [WikiEN-l] SOS Children Wikipedia Selection 2008/9 BitTorrent link up To: Andrew Cates Andrew@soschildren.org
2008/10/24 Andrew Cates Andrew@soschildren.org:
Hi Guys,
I have left a fairly full reply to this on the WMF blog awaiting approval from Jay and also it is discussed on the project pages on Wikipedia, over several years. There is a gap between the wording of licenses and urban myths circulating about what they say.
Broadly the GFDL demands that authors are "credited" but does not include anything on how you idenitify them (unlike the creative commons licenses). Either this means to comply with GFDL you need to carry a local copy of the edit history (which provides the only local way of identifying authors, albeit in a tedious fashion) including all
10,000 versions of the Global Warming article complete with every
piece of obscene vandalism etc.) or this means you have to credit authors providing a theoretically possible route to identify them. Nothing in between this is any better than the second option since to find an author for a piece of text you still have to go to the page history on Wikipedia. The German DVD which carries an author list copy locally which may be better for egos but is not more compliant than us: to get the author who wrote xyz is still a long trip through WP page histories. There is no different in license terms between a link back and any other way given of directing the reader to the page histories in Wikipedia. GFDL does not mention "link" (see Text_of_the_GNU_Free_Documentation_License)
Some of the CC licenses include carrying a local copy of an artists preferred name (we do this as we carry the image pages).
Sure, it's been discussed constantly for the last 7 years and there has been no conclusion, so I would strongly advise erring on the side of caution. I see no reason to include each revision, the GFDL only talks about listing authors, not listing who wrote each bit. That means you just need the history page, not all the pages it links to. An obscure and non-explicit reference to a page that may well not be accessible to the reader, hidden away on a page I doubt more than 1% of users will read (or even be able to find if they wanted to know who wrote an article) is hardly in the spirit of the license, is it? All you need to do is include a list of names (and/or pseudonyms) with each article, that isn't hard.
If you want to encourage the use of free content, you need to do it properly. That you can convince yourself that you're just about complying with the license if you stand on your head and squint as you look at it is not enough.
This is the third time these questions have turned around just on this project. We ARE erring on the side of caution which is why we have all the image pages (unlike others...).
I wouldn't mind doing more but I would question the value of useless gestures. For example if you say who wrote what is irrelevant I can get a list of the 60,000 editors including IP addresses who have edited any of these articles and add it as plain text at the foot of the license page (similar to the German DVD model) but does that really help? Legally I think its irrelevant but I would do it if it helped feelgood for contributers which is fair enough as an object. But isn't the point that Wikipedia should be credited and their fame comes from contributing to Wikipedia, whose greater glory we are supporting? We also have to consider the issue of "names which give offence". etc. If we do it this year I am prepared to bet the contributing list will include minor edits by user "Mr X is a ***" which userid was only deleted after an article was modified by them. As mentioned in the blog post I am already struggling with the number of unhelpful image descriptions and contributers names from the images, not to mention ones hoping "this *!?"!* upload works better than the last one". Try running a rude word checker over the image pages. While you are at it I notice there are a lot of obscene redirects around presumable from people who want to tell their mates "if your search for c*** on Wikipedia it returns an article on person X" so rude word checking the redirect database would be a good thing to set up too. Sigh.
There is a whole series of questions on what is a compliation and is Wikipedia a publisher etc. but my reading of the license is you have to provide a means of knowing you did what. Doing so locally is impossible (and not explicitly asked for) doing so by giving directions to find this elsewhere is what we do. Anything in between we could do but its fluff. Why don't you write 400 words on how Wikipedia is put together and by whom and I'll include that. That would achieve much more in terms of children understanding what is black and white does not fall out of the sky.
Andrew
On Fri, Oct 24, 2008 at 12:16 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
Sent off-list by accident:
---------- Forwarded message ---------- From: Thomas Dalton thomas.dalton@gmail.com Date: 2008/10/24 Subject: Re: [WikiEN-l] SOS Children Wikipedia Selection 2008/9 BitTorrent link up To: Andrew Cates Andrew@soschildren.org
2008/10/24 Andrew Cates Andrew@soschildren.org:
Hi Guys,
I have left a fairly full reply to this on the WMF blog awaiting approval from Jay and also it is discussed on the project pages on Wikipedia, over several years. There is a gap between the wording of licenses and urban myths circulating about what they say.
Broadly the GFDL demands that authors are "credited" but does not include anything on how you idenitify them (unlike the creative commons licenses). Either this means to comply with GFDL you need to carry a local copy of the edit history (which provides the only local way of identifying authors, albeit in a tedious fashion) including all
10,000 versions of the Global Warming article complete with every
piece of obscene vandalism etc.) or this means you have to credit authors providing a theoretically possible route to identify them. Nothing in between this is any better than the second option since to find an author for a piece of text you still have to go to the page history on Wikipedia. The German DVD which carries an author list copy locally which may be better for egos but is not more compliant than us: to get the author who wrote xyz is still a long trip through WP page histories. There is no different in license terms between a link back and any other way given of directing the reader to the page histories in Wikipedia. GFDL does not mention "link" (see Text_of_the_GNU_Free_Documentation_License)
Some of the CC licenses include carrying a local copy of an artists preferred name (we do this as we carry the image pages).
Sure, it's been discussed constantly for the last 7 years and there has been no conclusion, so I would strongly advise erring on the side of caution. I see no reason to include each revision, the GFDL only talks about listing authors, not listing who wrote each bit. That means you just need the history page, not all the pages it links to. An obscure and non-explicit reference to a page that may well not be accessible to the reader, hidden away on a page I doubt more than 1% of users will read (or even be able to find if they wanted to know who wrote an article) is hardly in the spirit of the license, is it? All you need to do is include a list of names (and/or pseudonyms) with each article, that isn't hard.
If you want to encourage the use of free content, you need to do it properly. That you can convince yourself that you're just about complying with the license if you stand on your head and squint as you look at it is not enough.
Wikimedia UK mailing list wikimediauk-l@wikimedia.org http://meta.wikimedia.org/wiki/Wikimedia_UK http://mail.wikimedia.org/mailman/listinfo/wikimediauk-l
2008/10/24 Andrew Cates Andrew@soschildren.org:
This is the third time these questions have turned around just on this project. We ARE erring on the side of caution which is why we have all the image pages (unlike others...).
I wouldn't mind doing more but I would question the value of useless gestures. For example if you say who wrote what is irrelevant I can get a list of the 60,000 editors including IP addresses who have edited any of these articles and add it as plain text at the foot of the license page (similar to the German DVD model) but does that really help?
It would be far better (and no harder) to include a separate list for each article. That way you avoid worrying about the "Is Wikipedia a single GFDL work or a collection of GFDL works?" debate.
The legalities of the GFDL are so confusing and ambiguous that they are best ignored for the most part. Just try and follow the spirit of the license, which includes that attribution is required. Telling somebody how to find the list of authors in a place they can't access is not attribution by any reasonable definition.
Thomas,
I have read your note twice. I have reread the GNU licence three times. I have read the Wikipedia copyright statements (again).
It is still completely unclear to me if Wikipedia counts as a publisher (which is looks like it is but it says it isn't). Still though I really don't think the "Is Wikipedia a single GFDL work or a collection of GFDL works?" is relevant since as far as I can possibly see the current set up already complies in good faith with all the licenses, at least for as long as Wikipedia itself exists online.
Your argument about "reasonable definition" I don't buy. 95% of authors only give pseudo-names, which are often real name of other people whom they are not. The identification of most of the authors requires their user page as well, or to refer to "the user who used the pseudonym of John Smith on Wikipedia" in which case the references are back to Wikipedia and we might as well send everything there. It is a reasonable definition these days to tell people where they can find the information they require online. Countless instruction manuals and safety notices do this now. We say there are authors who deserve credit and where to find them. I think thats better than pretending a list of names and IPs means much to anyone.
For transparency ref comments on "censoring" we could give the URL to the exact version number of the article we used, as an article history.
Beyond this I don't think further effort increases the compliance. But I repeat my comment I am serious about wanting feelgood for contributers. Perhaps there is a way of doing this sensibly.
Andrew
On Fri, Oct 24, 2008 at 12:56 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
2008/10/24 Andrew Cates Andrew@soschildren.org:
This is the third time these questions have turned around just on this project. We ARE erring on the side of caution which is why we have all the image pages (unlike others...).
I wouldn't mind doing more but I would question the value of useless gestures. For example if you say who wrote what is irrelevant I can get a list of the 60,000 editors including IP addresses who have edited any of these articles and add it as plain text at the foot of the license page (similar to the German DVD model) but does that really help?
It would be far better (and no harder) to include a separate list for each article. That way you avoid worrying about the "Is Wikipedia a single GFDL work or a collection of GFDL works?" debate.
The legalities of the GFDL are so confusing and ambiguous that they are best ignored for the most part. Just try and follow the spirit of the license, which includes that attribution is required. Telling somebody how to find the list of authors in a place they can't access is not attribution by any reasonable definition.
Wikimedia UK mailing list wikimediauk-l@wikimedia.org http://meta.wikimedia.org/wiki/Wikimedia_UK http://mail.wikimedia.org/mailman/listinfo/wikimediauk-l
It is still completely unclear to me if Wikipedia counts as a publisher (which is looks like it is but it says it isn't).
I'm not sure it's really relevant. The WMF claims to be whatever is required for Section 230 (or whatever it is) protection, although I'm not at all convinced it actually qualifies (although I think there may now be case law saying it does).
Still though I really don't think the "Is Wikipedia a single GFDL work or a collection of GFDL works?" is relevant since as far as I can possibly see the current set up already complies in good faith with all the licenses, at least for as long as Wikipedia itself exists online.
It's only relevant to your suggestion to include one long list of names. Wikipedia includes separate lists for each article, I would suggest you do the same.
Your argument about "reasonable definition" I don't buy. 95% of authors only give pseudo-names, which are often real name of other people whom they are not. The identification of most of the authors requires their user page as well, or to refer to "the user who used the pseudonym of John Smith on Wikipedia" in which case the references are back to Wikipedia and we might as well send everything there.
I think it is reasonable to assume that someone contributing to Wikipedia under a pseudonym is happy to be credited under that pseudonym.
It is a reasonable definition these days to tell people where they can find the information they require online.
Not when one of the major purposes of the DVD is that it can be used when you don't have internet access. (The other major purpose being that it's specially selected and checked for UK school children.)
Countless instruction manuals and safety notices do this now. We say there are authors who deserve credit and where to find them. I think thats better than pretending a list of names and IPs means much to anyone.
It's means what
For transparency ref comments on "censoring" we could give the URL to the exact version number of the article we used, as an article history.
I don't see much point in that, myself.
Beyond this I don't think further effort increases the compliance. But I repeat my comment I am serious about wanting feelgood for contributers. Perhaps there is a way of doing this sensibly.
I think just including a list of names (and maybe IP addresses - there is an argument that contributing anonymously waives your right to attribution) for each article would vastly improve compliance, would make contributors happy and would not be a significant amount of work. You can provide a direct link to the Wikipedia history page so people can get the full information, simply as a courtesy, as well.
By the way I have checked the Wikipedia on DVD (Release Version) series and again they say:
"For a complete list of contributors for a given article, visit the corresponding entry on the English Wikipedia and click on "History" ."
But they also say
"For more details about the license of an image, visit the corresponding entry on the English Wikipedia and click on the picture."
We decided with images that the second didn't work for later CC licenses.
Andrew =====================
On Fri, Oct 24, 2008 at 12:56 PM, Thomas Dalton thomas.dalton@gmail.com wrote:
2008/10/24 Andrew Cates Andrew@soschildren.org:
This is the third time these questions have turned around just on this project. We ARE erring on the side of caution which is why we have all the image pages (unlike others...).
I wouldn't mind doing more but I would question the value of useless gestures. For example if you say who wrote what is irrelevant I can get a list of the 60,000 editors including IP addresses who have edited any of these articles and add it as plain text at the foot of the license page (similar to the German DVD model) but does that really help?
It would be far better (and no harder) to include a separate list for each article. That way you avoid worrying about the "Is Wikipedia a single GFDL work or a collection of GFDL works?" debate.
The legalities of the GFDL are so confusing and ambiguous that they are best ignored for the most part. Just try and follow the spirit of the license, which includes that attribution is required. Telling somebody how to find the list of authors in a place they can't access is not attribution by any reasonable definition.
Wikimedia UK mailing list wikimediauk-l@wikimedia.org http://meta.wikimedia.org/wiki/Wikimedia_UK http://mail.wikimedia.org/mailman/listinfo/wikimediauk-l
wikimediauk-l@lists.wikimedia.org