On 10/29/07, Charlotte Webb &lt;charlottethewebb(a)gmail.com&gt; wrote: <snip> Both GFDL and CC-BY prohibit derivative works from removing "copyright notices" appearing in the original. The licenses are ambiguous about whether one can validly relocate such notices that are embedded in the image to the description page. Standing practice on Commons is to remove watermarks when possible, but also to allow the uploader a free hand to request deletion of the image if they object to that removal. -Robert Rohde

On 29/10/2007, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: We are not a newspaper and at random The Independent 26 October page 19 no credit on the images -- geni

On 30/10/2007, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: I could but until you stop making absolute statements I don't have to. Still same date pages 1,2,3,6,7. And we know how newspapers credit free images: http://en.wikipedia.org/wiki/Image:DNN_Wikipe-tan_2007-06-15.jpg -- geni

On 29/10/2007, White Cat &lt;wikipedia.kawaii.neko(a)gmail.com&gt; wrote: Not really - just include the information from the image page for all images in the current version. The problem would be to present it usefully. (For who?) - d.

I note with interest that in the "bleeding edge" mediawiki, there is a "credits" action which attempts to give some credit information - I don't see any links to it in the GUI though. Apparently it comes from WikiTravel. Anyone know anything about it? Or has this been around a while and doesn't do much useful? Steve

On 10/29/07, Gwern Branwen &lt;gwern0(a)gmail.com&gt; wrote: It's mostly about what people expect, I'd say. No one expects to see a bunch of author names splashed across the sidebar showing who wrote what; in cases where books have many editors and authors we're used to seeing a page or three listing them--separately from the text. In the case of photographs, though, people are used to seeing credit in the captions: newspapers, magazines, books, and even encyclopedias often do this. True, some leave all image credits for a separate section of the book, so our method isn't without precedent, but neither is the credit-in-the-caption method. In this case, 'excessive fair use' is probably secret code for 'so many photos we can't really claim fair use anymore'. Taken in that context, this happens all the time: for example, I often find myself removing song lyrics from articles about songs. In some cases the complete lyrics have been listed, in others a large part of them. It can certainly be appropriate to quote some of them for commentary, but quoting the whole song is 'excessive fair use'. Note, though, that I'm not taking a position on the issue of where to give credit on photos. I think that it's a good idea to make sure people know where to find the credits/license, but I'm not sure that putting credits in the caption is necessarily a good way to go either. A lot of arguments for and against it were brought up last time we had this discussion, so I won't rehash them; suffice it to say that I wasn't convinced of either position. I hope we can come to a good conclusion this time through. Tracy Poff

On 10/29/07, Gwern Branwen &lt;gwern0(a)gmail.com&gt; wrote: http://searchengineland.com/070911-083723.php How else are we to ensure maximum participation from marketers? ;)

Gwern Branwen wrote: It's a valid point. We generate a whole history on a separate page for text. #1 could do the same for photographers and other graphic people.Many books accumulate credits on one page at the back of the book. Ec Ec

Gwern Branwen wrote: <snip> errm yea:- http://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Non-free_content&a… The "official" silence on this issue was/is still deafening

Andrew Gray schreef: Not really, 99% of our text is written by Wikipedians, who have no expectation to be prominently attributed; we have taken many of our best pictures from outsiders, many of whom expect more prominent credit to be given. They don't care about, and haven't ever explicitly agreed to Wikipedia's internal rules on attribution. Eugene

On 29/10/2007, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: Text on Wikipedia is almost universally released under GFDL specifically to be used on Wikipedia. We rarely take text from anywhere else except in specific quotes, which we would always attribute (no point quoting someone without saying who you're quoting), and are used under fair use rather than being released under a certain license, anyway. We often take images from other places where they have been released under a variety of licenses, often we no intention for them to be used on Wikipedia. There is a big difference. (Whether or not it's a legally significant difference, I don't know, but it is a big difference.)

Robert Rohde wrote: Another way of reading that is to take the Main Page as the title page. The five listed authors could then be the five editors (other than bots) who have the highest number of edits. :-) Ec

On Oct 29, 2007 7:51 PM, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: Since when does one example turn into 99%, many, and many? In that particular case the writer was primarily upset that *no attribution at all was provided*... which was due to a momentary database glitch. [snip] For the most part the 1911 notice was useful for explaining why the article used arcane English and had an occasional racial slur... And I prefer that we not create an easy way for people to advertise or self promote directly inside articles. I'd prefer that our encyclopedia articles not be littered with names, sometimes quite offensive or insulting names. I think it's utterly essential, legally and ethically, that we be as consistent as possible in our attribution of images, so crediting just images from third parties isn't really in option in my eyes. We seem to be at an impasse. :) I am in favor of improving our 'one click away' attribution. I've advocated we create a credits tab multiple times. There is clearly a lot that we can do to improve attribution without shoving it inline. I would not be opposed to including it appended to the bottom of the pages, as mediawiki does for text authors on Wikis that haven't disabled that feature. Is there room to compromise here? Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: The GFDL requires us to list *at least* the five main. We list all of them. We're a little mixed up with the section naming, but we in terms of actual substance we are in full complaince. (except when people paste in stuff without attributing it, of course)

On 10/30/07, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: It serves two purposes in one go. Nice, huh? Crediting images in the same way as crediting text is "utterly essential, legally and ethically"? No, that's a huge stretch. I agree we don't necessarily want to open the door to "advertising", but there surely safeguards we could put on that. IMHO we have already strayed into a grey area with our rampant use of US military propag^H^H^H^H^H photos. Hell yeah. Any of these would be fine, IMHO, for text: - Box with names of prominent contributors - Prominent link to "credits" or "authors" with clearly spelled out list of main contributors (rather than the "history" tab which is really for something else) For images: - Name of author* shown in small print next to the caption - Prominent link to an obvious "more information" type icon on the image (like the example you made) Slightly less good: - List of contributors of images displayed somewhere on the page - Contributors of images shown on same page as authors. In either case we will need to actually store this metadata properly. I don't think we have a reliable way of determining who the author is. you to a filing cabinet and tell you all the information is in there. Can you tell from a history page who the five main contributors are? How long will it take you to tell? Steve *Author - got a better term for the person who made an image, if it isn't necessarily a photo?

On 10/30/07, RLS &lt;evendell(a)gmail.com&gt; wrote: author is neutral in that way. Steve

On 10/29/07, Matthew Brown &lt;morven(a)gmail.com&gt; wrote: There was quite a bit of work done, especially on the commons list, but no one pulled the trigger to make it so. http://en.wikipedia.org/wiki/User:Thebainer/thumbtest

On Oct 29, 2007 7:14 PM, David Gerard &lt;dgerard(a)gmail.com&gt; wrote: So... I don't much care what the outcome is, but the i gets significant negative points in my book for not being maximally linguistically neutral. The *best* way to make this change is to change the monobook behavior in mediawiki because consistency across our projects is important, and they all have the same issue of clicking not being obvious. However, this demands a language-neutral solution. To that end what I would prefer is that some brilliant artist would come along and make a nice clear monobook feel version of the (C) and magnifying glass icon which is in the lower left corner. ... If wishes were horses..

On Oct 29, 2007 8:12 PM, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: It's true. Your suggestion is consistent enough. But now you've now taken it out of my ability to impliment. Replacing a single icon is easy. Setting icons for 209 languages is much harder. ;) Feel free to edit the page to propose other styles. It's not hard to do better than what we have today. :) So long as people can get behind one version we can change it. :)

On Oct 29, 2007 8:26 PM, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: 0_o Is a solution which no one will bother implementing really best? The RED (C) probably does, I agree. But the pale blue icon, the example in the lower left corner? The images are, in most cases, copyrighted. When you copy them you have certain obligations. If your entire understanding of your obligations is going to come from a 15px wide icon, then I'd rather you think you can't copy the images. How can we worry about respecting the authors on one hand but then worry that we might shatter a reusers misconceptions about his copyright obligations? ;) We don't, for the most part, provide inline credit in such cases. Most of the EB1911 notices have long since been removed. They are useful for certain informational purposes, but it's not really a standard practice.

On Oct 29, 2007 8:58 PM, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: Holy crap, so you'd propose we wrap every thumbnail/bordered image with a template. Shoot me now. Thats a very very bad idea. Well, any change is okay with me. No. I shouldn't say that. Proposing we start wraping all captioned images with a template is not okay with me. Thats not at all in strict conformance with the GFDL. (Nor is just mentioning at the bottom though its much better).

On Oct 29, 2007 9:58 PM, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: You have terrible taste. Prepare to die. *stab* *stab* Thou shall not code the autorship data redundantly. The data you want is on the image page. Do not copy it into the page text. Multiple copies invite bitrot, vandaism, human error, and increase work.

On Oct 30, 2007 12:53 AM, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: Yes, it's true. All the more reason to fix it once. Most images on commons use the information template which does mostly do the right thing. ;) There are several enhancement requests in bugzilla with basically amount to "read data out of the image page so we can do things with it in the articles". It's a pretty reasonable request, no one has gotten around to implementing it yet. Perhaps I'll do it sometime. I completely approve of tooltips. It would nice to make the tooltip Caption &mdash; author / License. I also approve of creating a nice credits tab. And I'm not strongly opposed to appending credits information at the bottom of the page. It's just the always visible inline stuff that I think is a problem.

On Oct 30, 2007 1:13 AM, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: Lets go one at a time: This one is easiest. Let consensus behind a style, then I'll work out the details and make sure we get it implemented. Commons has the data source side of this nearly covered with information template. Basically some kind of magic word would be added to the mediawiki markup that indicates the surrounded post-transclusion text should be copied out into a separate column in the database. Since commons already has structured data for authorship, once the software is done this would require only a few template changes. What do we do on Enwp? ... I'm thinking that an effort to move all free images to commons would actually take no more work than trying to fixup the descriptions on enwp. This is an utterly trivial software change (on the order of a few minutes) after the data is extracted and available in the database. If we ignore text it's like the above, but ignoring text is ignoring the elephant in the room. Traversing 30k+ revisions to get all the user names who have edited and article with a huge history is not reasonable. One possibility would be creating a editable credits tab (like a talk page). The bottom of the page is auto populated with image credits, basically the extracted image data plus a thumbnail for each image. The top would be a editable Wikipage. We could bot populate them with a list of all people who edited, or a filtered version, and then invite people who add new content to add themselves to the page. Don't want to be credited? don't list yourself. I fear that this last one is complicated and significant enough of a change that its dead on arrivial, but it's the only way that I think we could stand to seriously improve our text attribution. :(

On Oct 30, 2007 2:47 AM, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: Sounds good. Yes. AFAIK only because of things like Category:GFDL_Presumed and other mostly rubbish stuff. Fair enough. I can perform it and commit it myself. Lead time 'depends'. If there are changes in SVN which require a database schema change we'll lag further behind SVN. Right now we're about 3.5 weeks behind SVN. The bigger issue here is that this feature would depend on the metadata magic word stuff above which is a bigger change which will likely require a schema change.

User:Carcharoth wanted me to pass this on: I agree with you that photo credits should be easier to access - not so much for individual photographers (though that would be nice), but the vast amount of, say, Library of Congress, NASA, and other US government/public institution pictures. Astronomical pics, for example, should always say in the caption what instrument took them, and so on. Steve

On 10/30/07, David Gerard &lt;dgerard(a)gmail.com&gt; wrote: sourcing it. War photos you would probably want to know which side took the photo, for instance. So anyway I think we have reached some compromise somewhere between viewpoint #1 and #2: mere attribution on the image page is insufficient (we need better links there), but textual attribution on the article page itself next to the image is excessive. Steve

On 10/29/07, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: I was actually going to suggest something similar to the "TDS Article Contribution Counter" on the tool server would be reasonable for identifying the top all-time contributors to a major article, but the bookmark i have for that is now giving me a 403-forbidden error. Is this for some reason more of a database hog than say, using interiot's tool to look up the edit count of a user who has made about "30k+ revisions" to various pages? And even if it is a discouragingly "expensive" task for high-profile articles, here's one definition of "reasonable": a short list of the top five or so contributors (as suggested in GFDL section 4B), which could be cached at an interval proportionate to the total length of the history, i.e. greener articles would allow the "principal authors" list to be re-cached more frequently, something like [[George W. Bush]] much less so. If in fact I'm hallucinating (must be that gray acid) and meta-data like this wouldn't actually put us any closer to obeying the GFDL, it would at least be interesting for casual research while browsing unfamiliar pages, and doing "&action=history&limit=5000" (or worse, loading one page of 50 at a time) to get an overall response to the mental question "lol, so whose work is this, really?" can be painful. —C.W.

On 10/30/07, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: Yeah, that's why I said "something similar to", but that part was purely speculative as I imagine an edit counter tool to *reliably* measure weight rather than volume would be nigh impossible to code, but at the same time I hope somebody proves me wrong on this. Either way, I'm still wondering about the TDS tool (for slightly unrelated reasons). Anybody know if it's been moved to a different url or willfully disabled? —C.W.

On 10/30/07, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: I don't think that's true. Flickr uses a grey "(c)" (backwards?) for the link to licence information. I think the (C) is a good indicator that you shouldn't just pinch the image without thinking. You should stop, check the licence information, then go ahead. In any case, most Wikipedia images *are* copyrighted. They're also free licenced. Steve

On 10/30/07, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: Think Wiki. You produce one version. Everyone on a Wikipedia where the "i" is inappropriate can produce their own. Steve

On 10/30/07, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: If you ignore the substantial problem of text migrating from one article to another (admittedly a big deal), then I would have to disagree that it can't be done automatically. It shouldn't be done in real time, but a computer with access to the full revision history could figure out how much of the current text each historical contributor was responsible for, and this could be used to identify primary authors and add their names via a bot. It's not a trivial problem but the existing research efforts on the persistance and evolution of text in wiki articles already largely address these issues. -Robert Rohde

On 30/10/2007, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: Indeed. No-one seriously objects to not having the full text of the GFDL right there on every page, rather than linked from somewhere obvious. - d.

On Oct 30, 2007 1:21 PM, Anthony &lt;wikimail(a)inbox.org&gt; wrote: Sure, so the one time vandal gets equal time to the person with thousands of edits. I guess we'll have to disagree on this one. Hm. How about an editable part, then the image credits. Then a big list of all names which excludes anyone listed in the editable part. The big list could have vandal names hidden by sysops (bots?), and people who have opted-out are hidden. Perhaps make the minor edit flag into a per edit opt-out flag. If all your edits are marked minor you are hidden in the credits page by default. Otherwise you can go hide yourself.

On Oct 30, 2007 1:36 PM, Andrew Gray &lt;shimgray(a)gmail.com&gt; wrote: It's unusual for templates to contain really significant copyright bearing text, of course there is layout and code and blah. We should probably make the decision that our templates are all public domain, apply that to the edit screens for that namespace and hope that it will someday be true. ;)

I like this idea -- as a (hidable?) section of the history page, not as a replacement for it... authorship is a significant part of the history; from a GFDL and attribution perspective, one of the most important ones. SJ On Oct 30, 2007 1:14 PM, Anthony &lt;wikimail(a)inbox.org&gt; wrote: -- ++SJ

On 30/10/2007, Gregory Maxwell &lt;gmaxwell(a)gmail.com&gt; wrote: As an aside, someone once suggested (and I think coded up a demonstration) for "hiding" vandal usernames - they'd be quietly removed from history pages, etc, and marked as [vandal] or [blocked user] the like. The main reason to do this was to avoid all the cruft on Special:Listusers, but it could help here... -- - Andrew Gray andrew.gray(a)dunelm.org.uk

No? Software change such that from a certain point on any edit will automatically update a credit page. At the same time, start running something that goes through all the non-deleted edits and update the page as well. Then and only then do one display a link to the credit page. The speed of any program that go through the history of all the articles can be adjusted to run over several weeks / months at a rate the server can cope with. No one is seriously (I hope) suggesting that any change will need to be made overnight. (One will need to take into consideration what happens when pages that was deleted when the program did its run are then undeleted after the credit page goes live.) KTC -- Experience is a good school but the fees are high. - Heinrich Heine

On 11/11/07, Steve Summit &lt;scs(a)eskimo.com&gt; wrote: That's a reasonable point of view. If most people agree, then we should not use those images. We can't have it both ways: use free images that require attribution, then not really attribute them properly. Certainly, we should attribute images where it's convenient to do Other peole have indicated that that's too much, and they'd prefer an aggregated list of credits at the bottom of the page. That's reasonable too. Steve

On 11/11/07, geni &lt;geniice(a)gmail.com&gt; wrote: all URLs should be banned as "attribution"? In any case, what exactly would Wikipedia be doing with photos of electric guitars in showers... Steve

On Sun, 11 Nov 2007 13:56:24 +0000, "Andrew Gray" &lt;shimgray(a)gmail.com&gt; wrote: And what do we do about people who currently upload and add pictures *with* a namecheck, and also add links to their personal website? How do we tell the difference between vanity, trying to get *your* name on the FA, and a genuine desire to help Wikipedia? Guy (JzG) -- http://www.chapmancentral.co.uk http://en.wikipedia.org/wiki/User:JzG

On 11/11/2007, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: We attribute NASA in NASA images even though we don't have to credit them at all, because the source of the image is almost always of actual article interest. Place, time and photographer are frequently going to be things of interest as article content and well worth putting in the caption for that reason. - d.

On 11/11/2007, Steve Bennett &lt;stevagewp(a)gmail.com&gt; wrote: Because we don't need spam links in image captions Nothing. That would just be how the person who provided the image of the presidential candidate asked to be credited. -- geni